WO2024077587A1 - Battery performance prediction method, and battery performance distribution prediction method - Google Patents

Battery performance prediction method, and battery performance distribution prediction method Download PDF

Info

Publication number
WO2024077587A1
WO2024077587A1 PCT/CN2022/125298 CN2022125298W WO2024077587A1 WO 2024077587 A1 WO2024077587 A1 WO 2024077587A1 CN 2022125298 W CN2022125298 W CN 2022125298W WO 2024077587 A1 WO2024077587 A1 WO 2024077587A1
Authority
WO
WIPO (PCT)
Prior art keywords
battery
battery performance
process parameters
manufacturing
performance
Prior art date
Application number
PCT/CN2022/125298
Other languages
French (fr)
Chinese (zh)
Inventor
李晓彤
吴兴远
Original Assignee
宁德时代新能源科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 宁德时代新能源科技股份有限公司 filed Critical 宁德时代新能源科技股份有限公司
Priority to PCT/CN2022/125298 priority Critical patent/WO2024077587A1/en
Publication of WO2024077587A1 publication Critical patent/WO2024077587A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells

Definitions

  • the present application relates to the field of batteries, and specifically to a battery performance prediction method, device, computer equipment, storage medium and computer program product; it also relates to a battery performance distribution prediction method, device, computer equipment, storage medium and computer program product.
  • the present application provides a battery performance prediction method, apparatus, computer equipment, storage medium and computer program product to achieve accurate prediction of battery performance; and provides a battery performance distribution prediction method, apparatus, computer equipment, storage medium and computer program product to achieve accurate prediction of battery performance distribution.
  • the present application provides a battery performance prediction method, the method comprising:
  • the battery performance of the battery is obtained.
  • the process parameters of battery manufacturing are first obtained, and then the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value.
  • the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
  • obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
  • the trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • a trained battery correction value prediction model is used to accurately obtain the impact correction value corresponding to the process parameters.
  • the model fully explores the influence of different process parameters on battery performance, and can obtain the accurate correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery. Therefore, the battery performance of the battery can be accurately predicted in the end.
  • the method of obtaining the battery correction value prediction model includes:
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values, so that the battery performance prediction value prediction model can accurately characterize the influence of the process parameters on the battery performance. Therefore, the battery performance prediction value prediction model finally trained can support the accurate prediction of the correction value of the influence of the process parameters on the battery performance.
  • the method of obtaining the battery correction value prediction model includes:
  • the battery performance of several sample batteries is obtained, and the corresponding relationship between the process parameters and the battery performance is generated;
  • the machine learning model is trained according to the correspondence between process parameters and battery performance to obtain a battery correction value prediction model.
  • sample process parameters of several sample batteries are first obtained, and the corresponding relationship between the battery performance of the sample batteries is obtained according to the sample process parameters, and the corresponding relationship between the process parameters and the battery performance of the sample batteries is constructed.
  • the machine learning model is trained with the constructed corresponding relationship to obtain a battery correction value prediction model.
  • machine learning training is carried out using sample data, and the battery correction value prediction model finally trained can accurately characterize the corresponding relationship between the process parameters and the battery performance, and thus can support the final accurate prediction of the battery performance of the battery.
  • the battery performance of several sample batteries is obtained, and the corresponding relationship between the process parameters and the battery performance is generated, including:
  • the P2D electrochemical-thermal coupling model was used to process several sample process parameters to obtain the battery performance of each sample battery;
  • the sample process parameters and battery performance are associated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
  • the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery, and then the battery performance of each sample battery is associated with the sample process parameters, that is, the corresponding relationship between the process parameters and the battery performance of the sample battery is obtained.
  • the battery performance of the sample battery can be quickly obtained through the P2D electrochemical-thermal coupling model, thereby improving the processing efficiency.
  • the P2D electrochemical-thermal coupling model is obtained by:
  • test process parameters are processed through the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction results corresponding to the test process parameters;
  • the initial P2D electrochemical-thermal coupling model was optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model.
  • the initial P2D electrochemical-thermal coupling model is optimized based on the test cell performance corresponding to the measured test process parameters to obtain a more accurate P2D electrochemical-thermal coupling model that better meets the scenario requirements.
  • the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model, including:
  • the initial P2D electrochemical-thermal coupling model is adjusted according to the correction factor to obtain the P2D electrochemical-thermal coupling model.
  • the test battery performance is compared with the initial battery performance prediction results to obtain battery performance difference data, identify the fluctuating process parameters corresponding to the battery performance difference, and obtain the correction factor corresponding to the indicator.
  • the initial P2D electrochemical-thermal coupling model is adjusted with the correction factor to obtain a more accurate P2D electrochemical-thermal coupling model.
  • obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
  • the P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the P2D electrochemical-thermal coupling model is used to directly process the process parameters. Since the P2D electrochemical-thermal coupling model is a model with stable performance and accurate battery performance prediction, it can accurately obtain the correction value of the impact of the process parameters on the battery performance of the battery.
  • obtaining process parameters for battery manufacturing includes:
  • At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  • the process parameters include indicators of multiple dimensions such as particle size, coating thickness, pole piece size, porosity, compaction density and surface density, and the influence of different process parameters on battery performance is analyzed.
  • the present application provides a method for predicting battery performance distribution, the method comprising:
  • process parameter fluctuation data several groups of process parameters for battery manufacturing are generated.
  • the process parameter fluctuation data of battery manufacturing is first obtained, and several groups of process parameters are generated according to the process parameter fluctuation data of battery manufacturing. This can fully simulate the process parameters that may correspond in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
  • process parameter fluctuation data including:
  • process parameter fluctuation data several groups of process parameters of the manufacturing process are randomly generated.
  • a random generation method is adopted to generate several groups of process parameters.
  • the random generation method enriches the number of process parameters; on the other hand, the random generation method makes the generated process parameters closer to the actual situation of the corresponding process parameters in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
  • the present application further provides a battery performance prediction device, the device comprising:
  • a process parameter acquisition module is used to obtain process parameters for battery manufacturing
  • a processing module used to obtain a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
  • the performance prediction module is used to obtain the battery performance of the battery according to the impact correction value.
  • the present application further provides a battery performance distribution prediction device, the device comprising:
  • a correction module used to obtain correction values of the effects of the manufacturing process on the battery performance of the battery corresponding to several groups of process parameters, and obtain several groups of influence correction values;
  • a battery performance acquisition module used to obtain battery performance of several groups of batteries according to several impact correction values
  • the battery performance distribution processing module is used to generate battery performance distribution prediction results based on the battery performance of several groups of batteries.
  • the present application further provides a computer device.
  • the computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
  • the battery performance of the battery is obtained.
  • the present application further provides a computer device.
  • the computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
  • the present application further provides a computer-readable storage medium.
  • the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
  • the battery performance of the battery is obtained.
  • the present application further provides a computer-readable storage medium.
  • the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
  • the present application further provides a computer program product.
  • the computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
  • the battery performance of the battery is obtained.
  • the present application further provides a computer program product.
  • the computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
  • the process parameters of battery manufacturing are first obtained, and then the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value.
  • the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
  • the battery performance distribution prediction device, computer equipment, storage medium and computer program product obtain several groups of process parameters for battery manufacturing, obtain the corresponding manufacturing process's impact correction values on the battery performance of the battery for the several groups of process parameters, and then obtain the battery performance of several groups of batteries based on the several groups of impact correction values, analyze the battery performance of these batteries, and generate the battery performance distribution prediction results corresponding to the several groups of batteries.
  • the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered respectively, and the battery performance of several groups of batteries can be accurately predicted, so that accurate battery performance distribution prediction can be achieved.
  • FIG1 is a diagram showing an application environment of a battery performance prediction method according to an embodiment
  • FIG2 is a schematic diagram of a flow chart of a battery performance prediction method in one embodiment
  • FIG3 is a schematic flow chart of a battery performance prediction method in another embodiment
  • FIG4 is a schematic flow chart of a battery performance prediction method in yet another embodiment
  • FIG5 is a schematic diagram of a flow chart of a method for predicting battery performance distribution in one embodiment
  • FIG6 is a schematic flow chart of a method for predicting battery performance distribution in another embodiment
  • FIG7 is a schematic diagram of the technical concept of a battery performance distribution prediction method in practical application in one embodiment
  • FIG8 is a structural block diagram of a battery performance prediction device in one embodiment
  • FIG9 is a structural block diagram of a battery performance distribution prediction device in one embodiment
  • FIG10 is a diagram showing the internal structure of a computer device in one embodiment
  • Power batteries are not only used in energy storage power systems such as hydropower, thermal power, wind power and solar power stations, but also widely used in electric vehicles such as electric bicycles, electric motorcycles, electric vehicles, as well as military equipment and aerospace and other fields.
  • electric vehicles such as electric bicycles, electric motorcycles, electric vehicles, as well as military equipment and aerospace and other fields.
  • the market demand is also constantly expanding.
  • people can get more data support and reference when designing batteries to design batteries with higher performance or more in line with the needs of actual applications.
  • some researchers have proposed to optimize battery design and conduct battery research and development by analyzing battery performance, so some scholars have proposed a battery performance test scheme. In traditional schemes, battery performance is tested by experimental testing.
  • the inventor of this application proposes a new battery performance prediction scheme. First, the process parameters of battery manufacturing are obtained, then the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value. In the whole scheme, the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
  • the present application provides a battery performance prediction method, which can be specifically applied to the scenario of Figure 1, where the terminal 102 sends a battery performance prediction request to the server 104, and the server 104 responds to the battery performance prediction request, extracts the process parameters of the battery manufacturing carried in the battery performance prediction request, and then obtains the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery, and finally obtains the battery performance of the battery according to the correction value. Further, the server 104 can send the battery performance of the battery to the terminal 102, and the terminal 102 displays the battery performance of the battery.
  • the above-mentioned battery performance prediction scheme can be directly applied to the terminal, that is, the terminal can complete the performance prediction alone.
  • the user can operate on the terminal side, input the process parameters of battery manufacturing to the terminal to perform battery performance prediction, and the terminal responds to the user's battery performance prediction request, executes the above-mentioned battery performance prediction method, obtains the battery performance prediction result, and then pushes the battery performance prediction result to the user, for example, the battery performance of the battery can be displayed on the terminal display interface.
  • the specific processing process is similar to the above content and will not be repeated here.
  • the present application provides a battery performance prediction method, comprising:
  • Process parameters refer to parameters related to the battery manufacturing process, which may include but are not limited to particle size, coating thickness, pole piece size, porosity, compaction density and surface density, among which particle size refers to the diameter of the granular material in the incoming material; coating thickness refers to the coating thickness in the production of the battery cell; pole piece size refers to the size of the pole piece in the wound battery cell, which specifically includes the width of the pole piece, etc.; porosity refers to the percentage of the pore volume of the block material in the incoming material to the total volume of the material in the natural state; compaction density refers to the ratio of surface density to material thickness in battery production. In battery cell production, compaction density has a great influence on battery performance.
  • Compaction density is not only related to the size and density of the particles, but also to the gradation of the particles. Generally, particles with large compaction density have a good normal distribution. It can be considered that under certain process conditions, the greater the compaction density, the higher the battery capacity; surface density refers to the mass per unit area of a material with a certain thickness in the field of engineering materials.
  • the basic data of process parameters can be obtained through the relevant fluctuations of incoming materials in the battery manufacturing process, the incoming material detection system within the battery manufacturing company, or the incoming material fluctuation data provided by the supplier of incoming materials (raw materials) for battery manufacturing. Since there may be fluctuations in the manufacturing process during the actual manufacturing of the battery, and these fluctuations in the process are likely to affect the final battery performance, it is necessary to first obtain the process parameters for battery manufacturing.
  • Different process parameters will have different corresponding manufacturing processes.
  • the manufacturing process changes (fluctuates), it will affect the battery performance of the final battery produced. That is, in actual battery manufacturing, the battery performance of the battery is not only affected by the battery design parameters (the main influencing factor), but also by the battery manufacturing process.
  • the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the impact correction value of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained.
  • the corresponding relationship between the process parameters and the impact correction values of the battery performance can be constructed based on historical data. When it is necessary to obtain the impact correction values of the battery performance, the impact correction values are directly obtained based on the obtained process parameters and the above-mentioned corresponding relationship.
  • the corresponding relationship between the process parameters and the impact correction values of the battery performance can be a simple corresponding relationship table, or it can be a machine learning model obtained by specific training.
  • the corresponding relationship between the process parameters and the impact correction values of the battery performance is fully explored and characterized by the machine learning model.
  • S400 obtains the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the battery performance of the battery can be obtained, that is, the battery performance of the battery is obtained by comprehensively considering the benchmark battery performance and the correction value.
  • the benchmark battery performance is obtained based on the battery design parameters, which specifically refers to the battery performance corresponding to the battery design parameters under an ideal battery manufacturing environment.
  • Battery design parameters refer to the relevant parameters specified by the battery designer when it is designed. Generally speaking, different types of batteries correspond to different battery design parameters. When designers design and develop a new battery, they will give the specific design parameters corresponding to the new battery.
  • the benchmark battery performance can be obtained based on the battery design parameters.
  • the benchmark battery performance refers to the battery performance of the battery manufactured only under the conditions of the design parameters, or it can be simply understood as the benchmark battery performance of the battery under experimental conditions without considering the influence of the battery manufacturing process.
  • the influence of the process parameters in the battery manufacturing process on the battery performance is further considered, that is, the benchmark battery performance and the influence correction value are combined to obtain the corrected battery performance of the battery.
  • the above-mentioned battery performance prediction method first obtains the process parameters of battery manufacturing, uses the battery performance prediction value prediction model to process the process parameters, obtains the processing results, and then considers the impact of the manufacturing process on the battery performance based on the benchmark battery performance corresponding to the battery design parameters to obtain the battery performance of the battery.
  • the impact of the manufacturing process on the battery performance is considered; on the other hand, the impact of the process parameters on the battery performance is accurately analyzed by using the prediction model based on the battery performance prediction value, and the impact of different process parameters on the battery performance is fully explored through the model; therefore, the whole scheme can obtain accurate battery performance prediction results.
  • S400 includes: using the trained battery correction value prediction model to process the process parameters to obtain a correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the battery correction value prediction model is a model for obtaining the correction value of the impact of the manufacturing process on the battery performance of the battery according to the process parameters.
  • the battery correction value prediction model characterizes the corresponding relationship between the process parameters and the correction value of the impact of the battery performance.
  • the correction value of the impact of the battery performance corresponding to the current process parameters can be obtained based on the trained battery correction value prediction model and process parameters to fully consider the impact of the battery manufacturing process on the battery performance of the battery.
  • the battery correction value prediction model is a pre-trained model, which can be specifically trained based on historical data.
  • the corresponding relationship between the process parameter and the correction value of the impact of the battery performance can be obtained based on the analysis of historical data, and the model can be trained based on the corresponding relationship to obtain the trained battery correction value prediction model. Furthermore, the corresponding relationship between the process parameter and the correction value of the impact of the battery performance can be used to train the machine learning model, give full play to the learning and analysis capabilities of the machine learning model, deeply explore the impact of different process parameters on the final battery performance, and obtain accurate correction values for the impact of the battery performance.
  • a trained battery correction value prediction model is used to accurately obtain the impact correction value corresponding to the process parameters.
  • the model fully explores the influence of different process parameters on battery performance, and can obtain the accurate correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery. Therefore, the battery performance of the battery can be accurately predicted in the end.
  • the method of obtaining the battery correction value prediction model includes:
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values, so that the battery performance prediction value prediction model can accurately characterize the influence of the process parameters on the battery performance. Therefore, the battery performance prediction value prediction model finally trained can support the accurate prediction of the correction value of the influence of the process parameters on the battery performance.
  • the method of obtaining the battery correction value prediction model includes:
  • Sample batteries refer to batteries used to train battery performance prediction models. Generally, a small number of batteries are selected as sample batteries, for example, 100 batteries are randomly selected as sample batteries. The process parameters of these sample batteries in the actual manufacturing process are obtained, that is, the sample process parameters of the sample batteries are obtained. During the manufacturing process of batteries, for batteries of the same model (same design parameters), their process parameters fluctuate within a certain range of values. Specifically, multiple process parameters are randomly selected/composed within the range of process parameter values that can fluctuate as sample process parameters.
  • S340 Obtain battery performance of a plurality of sample batteries according to the sample process parameters, and generate a corresponding relationship between the process parameters and the battery performance.
  • the currently available battery performance prediction method can be used to obtain the corresponding battery performance, and then generate the corresponding relationship between the process parameters and the battery performance.
  • the sample process parameters can be used to design a battery, and then the designed battery can be tested by experimental measurement to obtain the battery performance of the sample battery.
  • a virtual battery can be simulated according to the sample process parameters, and then the battery performance corresponding to the simulated battery can be obtained.
  • the sample process parameters can also be input into the electrochemical model to obtain the battery performance of the sample battery.
  • the correspondence between the process parameters and the battery performance of the sample battery is generated by associating the sample process parameters and the battery performance of the sample battery.
  • the correspondence between the process parameters and the battery performance of the sample battery can be a table of the correspondence between the process parameters and the battery performance.
  • any one of the process parameters can be used as a variable, and the other process parameters can be used as invariants to draw the corresponding battery performance change relationship curve, that is, to obtain a battery performance fluctuation curve with different single process parameters as variables.
  • the correspondence between the process parameters and the battery performance of the sample battery can be obtained by calculation. A set of accurate data sets are used, and this data set is used as the data source for the next step of machine learning model training.
  • the coating thickness fluctuation as an example: when only the coating thickness fluctuation is considered, the corresponding battery performance under the influence of different coating thicknesses is obtained.
  • the different battery performance can be characterized by battery performance distribution, and the battery performance distribution can specifically include capacity distribution, power distribution, or temperature distribution.
  • Machine learning is a multi-disciplinary interdisciplinary subject involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in studying how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their own performance.
  • association rules use rules to describe the relationship between two or more variables, and are a method that objectively reflects the nature of the data itself. It is a large category of machine learning tasks, which can be divided into two stages. First, high-frequency project groups are found from the data set, and then their association rules are studied. The analysis results obtained are a summary of the laws between variables. Here, the correspondence between the process parameters and battery performance of the sample battery is used as training data to train the machine learning model to obtain a battery correction value prediction model. Through machine learning, the correlation between process parameter fluctuations and battery performance fluctuations is fully explored.
  • sample process parameters of several sample batteries are first obtained, and the corresponding relationship between the battery performance of the sample batteries is obtained according to the sample process parameters, and the corresponding relationship between the process parameters and the battery performance of the sample batteries is constructed.
  • the machine learning model is trained with the constructed corresponding relationship to obtain a battery correction value prediction model.
  • machine learning training is carried out using sample data, and the battery correction value prediction model finally trained can accurately characterize the corresponding relationship between the process parameters and the battery performance, and thus can support the final accurate prediction of the battery performance of the battery.
  • S340 includes:
  • S344 Correlate the sample process parameters and the battery performance to obtain a corresponding relationship between the process parameters and the battery performance of the sample battery.
  • the P2D electrochemical-thermal coupling model is a model obtained by coupling the P2D electrochemical model with the thermal model. Since the battery charging and discharging process generates heat, the battery temperature rises, and the increase in battery temperature will change the battery performance. If this phenomenon is ignored, it will inevitably lead to inaccurate prediction of battery performance. Therefore, it is necessary to couple the P2D electrochemical model with the battery thermal model to more accurately predict the battery performance.
  • the P2D electrochemical model is based on the theory of concentrated solution and porous electrode theory.
  • the grid (a discrete method) is divided according to the finite element idea, and then the partial differential equations in the electrochemical process are solved to obtain electrochemical properties such as electrode potential, electrolyte potential, and electrolyte concentration; the heat calculated by the electrochemical model is coupled into the thermal model as a heat source, causing the temperature change in the 3D thermal model, and the temperature change is fed back to the electrochemical model, causing the temperature-related parameters in the electrochemical model to change.
  • the changes in these parameters further trigger the changes in the heat source, thereby realizing the coupling of the electrochemical model and the thermal model, and ultimately affecting the electrothermal performance of the battery.
  • the sample process parameters are input into the P2D electrochemical-thermal coupling model to obtain the battery performance of the sample battery. It is precisely because the P2D electrochemical-thermal coupling model has the above-mentioned performance advantages and functions that the battery performance can be obtained after the sample process parameters are input into the P2D electrochemical-thermal coupling model.
  • the P2D electrochemical-thermal coupling model can obtain the battery performance of the sample battery based on the input sample process parameters, on the one hand, the processing speed of the P2D electrochemical-thermal coupling model is still relatively slow, and it takes a certain amount of time to process the output battery performance, which is acceptable for processing a small number of sample battery data, but it can no longer achieve the best performance for batch and large-scale battery performance prediction; on the other hand, the P2D electrochemical-thermal coupling model cannot deeply analyze and explore the relationship between fluctuations in different process parameters and fluctuations in battery performance, and it does not have a learning function.
  • the battery performance parameters directly obtained by the P2D electrochemical-thermal coupling model have limited guiding effect on subsequent battery design and development.
  • the P2D electrochemical-thermal coupling model is first used to process the sample process parameters of a small number of sample batteries, and the corresponding relationship between the sample process parameters of the sample batteries and the battery performance is established.
  • the corresponding relationship between the sample process parameters of the sample batteries and the battery performance is then used as training data to train the machine learning model, and the relationship between the fluctuations of different process parameters and the fluctuations of battery performance is mined and analyzed through the machine learning model.
  • Associating sample process parameters and battery performance refers to associating and recording the corresponding battery performance under different sample process parameters.
  • a data table can be set to record the battery performance corresponding to different sample process parameters, or a graphical curve can be used to record the battery performance corresponding to different sample process parameters.
  • the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery, and then the battery performance of each sample battery is associated with the sample process parameters, that is, the corresponding relationship between the process parameters and the battery performance of the sample battery is obtained.
  • the battery performance of the sample battery can be quickly obtained through the P2D electrochemical-thermal coupling model, thereby improving the processing efficiency.
  • the P2D electrochemical-thermal coupling model is obtained by:
  • test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
  • the initial P2D electrochemical-thermal coupling model refers to the conventional and general P2D electrochemical-thermal coupling model.
  • the test process parameters refer to the process parameters of the battery corresponding to the battery performance tested in the experimental test state. For example, here we choose to use the experimental test method to obtain a very small amount of battery process parameters and battery performance corresponding data. This very small number of batteries can be understood as test batteries.
  • the process parameters corresponding to this part of the test batteries are the test process parameters.
  • a complete and rigorous experimental test method is used to test the corresponding test battery performance.
  • the test battery performance has a very high accuracy, and the error between it and the battery's true battery performance is basically negligible, which can accurately characterize the true performance of the test battery.
  • the test process parameters are input into the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction result, which is the result of performance prediction of the test battery and belongs to the predicted value; the test battery performance and the initial battery performance prediction result are compared, and the difference between the two is analyzed. According to the difference between the actual value and the predicted value, the initial P2D electrochemical-thermal coupling model is optimized and corrected to obtain the P2D electrochemical-thermal coupling model, so as to obtain a more accurate P2D electrochemical-thermal coupling model.
  • the initial P2D electrochemical-thermal coupling model is optimized based on the test cell performance corresponding to the measured test process parameters to obtain a more accurate P2D electrochemical-thermal coupling model that better meets the scenario requirements.
  • the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model, including:
  • Some process parameters have a high influence on battery performance, while some process parameters have a low influence on battery performance.
  • Different weight values or correlation coefficients have been assigned to different process parameters in the entire model.
  • the weights/coefficients or constant values related to these process parameters in the P2D electrochemical-thermal coupling model need to be corrected. Specifically, the correction factors are directly used.
  • the correction factors corresponding to different process parameters can be obtained based on historical data analysis, or they may be obtained through feedback adjustment, that is, by selecting different correction factors, the P2D electrochemical-thermal coupling model is updated so that the final predicted value is infinitely close to the actual measured value, so as to obtain a more accurate P2D electrochemical-thermal coupling model.
  • the test battery can be obtained specifically in the following two ways: one is to artificially manufacture batteries of different designs, and specifically test performance such as capacity/DCR, etc., and simulate the performance of different designs respectively, compare them with actual measurements, and analyze the accuracy of the influence of each parameter of the model on the results; the other is to directly select from the batteries in the upper and lower warehouses of the production line. Specifically, by taking the incoming material/design/process data of the battery that has been manufactured and tested in the lower warehouse, n groups of design input models are generated according to the fluctuation data for calculation, and the calculated battery cell performance is compared with the performance of this batch of batteries after actual storage, and the accuracy of the influence of each parameter of the model on the results and the influence of the model are analyzed.
  • the test battery performance is compared with the initial battery performance prediction results to obtain battery performance difference data, identify the fluctuating process parameters corresponding to the battery performance difference, and obtain the correction factor corresponding to the indicator.
  • the initial P2D electrochemical-thermal coupling model is adjusted with the correction factor to obtain a more accurate P2D electrochemical-thermal coupling model.
  • training a machine learning model based on the corresponding relationship between the process parameters of a sample battery and the battery performance to obtain a battery correction value prediction model includes:
  • An initial Gaussian process regression model is obtained; the initial Gaussian process regression model is trained using the corresponding relationship between the process parameters of the sample battery and the battery performance to obtain a battery correction value prediction model.
  • GPR Gaussian Process Regression
  • GP Gaussian Process
  • the model assumptions of GPR include noise (regression residual) and Gaussian process priors, and its solution is based on Bayesian inference. If the form of the kernel function is not restricted, GPR is theoretically a universal approximator for any continuous function in a compact space. Based on the convenient properties of Gaussian processes and their kernel functions, GPR has been applied to problems in the fields of time series analysis, image processing, and automatic control. GPR is an algorithm with high computational overhead and is usually used for regression problems with low dimensions and small samples, but there are also extended algorithms suitable for large samples and high dimensions.
  • the initial Gaussian process regression model is trained based on the correspondence between the process parameters and battery performance of the sample battery.
  • part of the data can be used as a training set, and the other part of the data can be used as a test set.
  • the initial Gaussian process regression model is first trained with the training set to obtain the trained model, and then the trained model is tested with the test set to verify whether the trained model is qualified. If it is unqualified, the iterative training and optimization of the model are continued until it is finally qualified.
  • the training set can be composed of more data, and the test set can be composed of relatively less data.
  • the data volume ratio of the training set and the test set can be 8:2 to achieve a balance between training and testing.
  • the initial Gaussian process regression model can specifically be the Gaussian process regression model of ARD.
  • the Gaussian process regression model based on association rules can further accurately explore the correlation between process parameters and battery performance, that is, it can ultimately obtain a more accurate battery performance distribution result.
  • a Gaussian process regression model is used as the basic model, and the initial Gaussian process regression model is trained using the correspondence between the process parameters of the sample battery and the battery performance to obtain a battery correction value prediction model. Since the Gaussian process regression model can realize non-parametric regression of a stationary random field, the battery correction value prediction model finally obtained can support stable and accurate battery performance prediction.
  • obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
  • the P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the P2D electrochemical-thermal coupling model is used to directly process the process parameters. Since the P2D electrochemical-thermal coupling model is a model with stable performance and accurate battery performance prediction, it can accurately obtain the correction value of the impact of the process parameters on the battery performance of the battery.
  • obtaining process parameters for battery manufacturing includes:
  • At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  • Particle size refers to the diameter of the particulate matter in the incoming material
  • coating thickness refers to the thickness of the coating in the production of battery cells
  • pole piece size refers to the size of the pole piece in the wound battery cell, which specifically includes the width of the pole piece, etc.
  • porosity refers to the percentage of the pore volume of the block material in the incoming material to the total volume of the material in the natural state
  • compaction density refers to the ratio of surface density to material thickness in battery production. In battery cell production, compaction density has a great influence on battery performance. Compaction density is not only related to the size and density of the particles, but also to the grading of the particles. Generally, particles with large compaction density have a good normal distribution. It can be considered that under certain process conditions, the greater the compaction density, the higher the capacity of the battery; surface density refers to the mass per unit area of a material of a certain thickness in the field of engineering materials.
  • the process parameters include indicators of multiple dimensions such as particle size, coating thickness, pole piece size, porosity, compaction density and surface density, and the influence of different process parameters on battery performance is analyzed.
  • the inventor of the present application further realized that the performance prediction can be performed on batch batteries to obtain the performance prediction results of batch batteries, and then the performance prediction results of batch batteries can be further distributed analyzed to finally obtain the distribution results of battery performance.
  • Designers can intuitively understand the performance distribution of batch batteries based on the distribution analysis results of battery performance, which is helpful for designers to design batteries and shorten the battery research and development cycle.
  • the direct measurement method for batch battery performance research will also have the defect of inaccuracy.
  • the battery performance distribution prediction scheme several groups of process parameters for battery manufacturing are obtained, and the above-mentioned battery performance prediction method is used for processing to obtain the corrected battery performance of several groups of batteries, and the battery performance distribution results are analyzed with the corrected battery performance of several groups of batteries.
  • the above-mentioned battery performance prediction method is used for battery performance prediction, it is possible to obtain accurate corrected battery performance of several groups of batteries, and therefore, the battery performance distribution results of the battery can be accurately predicted in the end.
  • the battery performance distribution prediction scheme provided by the present application can also be applied to the application scenario shown in FIG. 1.
  • the terminal 102 sends a battery performance distribution prediction request to the server 104.
  • the server 104 responds to the battery performance distribution prediction request and obtains several groups of process parameters for battery manufacturing; obtains several groups of process parameters corresponding to the manufacturing process and correction values of the battery performance of the battery, and obtains several groups of correction values; obtains the battery performance of several groups of batteries according to the correction values; and generates the battery performance distribution prediction results based on the battery performance of several groups of batteries.
  • the server 104 sends the battery performance distribution results to the terminal 102, and the terminal 102 displays the battery performance distribution results, so that the R&D personnel can intuitively understand the battery performance of the batch battery and find the correlation between the battery performance and the electrochemical parameters, which is conducive to shortening the battery R&D cycle.
  • the above-mentioned battery performance distribution prediction solution can be directly applied to the terminal, that is, the terminal completes the performance distribution prediction alone, and its specific processing process is similar to the above-mentioned content, which will not be repeated here.
  • the present application provides a method for predicting battery performance distribution, the method comprising:
  • S820 Acquire several groups of process parameters for battery manufacturing.
  • S880 Generate a battery performance distribution prediction result based on the battery performance of several groups of batteries.
  • S820 includes:
  • S824 Generate several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
  • the process fluctuation data of battery manufacturing refers to the corresponding process fluctuation data in the battery manufacturing process, which can be obtained through the relevant fluctuations of incoming materials in the battery manufacturing process, the incoming material detection system within the battery manufacturing company, or the incoming material fluctuation data provided by the battery manufacturing incoming material (raw material) supplier.
  • the process parameter fluctuation data here can be understood as a fluctuation range value. Taking the coating thickness as an example, it can fluctuate within the range of (a, b).
  • the process parameter fluctuation data several groups of process parameters of the manufacturing process are combined and generated. For example, several groups of process parameters of the manufacturing process can be generated by random extraction and combination to simulate the diversified process parameter changes in real battery manufacturing.
  • the process parameter fluctuation data of battery manufacturing is first obtained, and several groups of process parameters are generated according to the process parameter fluctuation data of battery manufacturing. This can fully simulate the process parameters that may correspond in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
  • process parameter fluctuation data including:
  • process parameter fluctuation data several groups of process parameters of the manufacturing process are randomly generated.
  • a random generation algorithm can be used to generate several groups of process parameters of the manufacturing process, virtual batteries can be generated based on these process parameters, and then the battery performance corresponding to these virtual batteries can be obtained by the above-mentioned battery performance prediction method to generate battery performance distribution results.
  • the battery performance distribution result can be a normal distribution of battery performance, which can accurately characterize the distribution of battery performance corresponding to the fluctuation of process parameters.
  • random generation algorithms include Monte Carlo algorithms. Monte Carlo method, also known as statistical simulation method, is an important numerical calculation method in the mid-1940s.
  • the random generation algorithm uses the Monte Carlo algorithm, which can realize accurate data random generation processing and support the subsequent accurate battery performance distribution.
  • a random generation method is adopted to generate several groups of process parameters.
  • the random generation method enriches the number of process parameters; on the other hand, the random generation method makes the generated process parameters closer to the actual situation of the corresponding process parameters in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
  • the present application also provides a battery performance prediction device, the device comprising:
  • a process parameter acquisition module 200 is used to acquire process parameters for battery manufacturing
  • the processing module 400 is used to obtain a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
  • the performance prediction module 600 is used to obtain the battery performance of the battery according to the impact correction value.
  • the battery performance prediction device of the present application first obtains the process parameters of battery manufacturing, uses the battery performance prediction value prediction model to process the process parameters, obtains the processing results, and then considers the impact of the manufacturing process on the battery performance based on the benchmark battery performance corresponding to the battery design parameters to obtain the battery performance of the battery.
  • the impact of the manufacturing process on the battery performance is considered; on the other hand, the impact of the process parameters on the battery performance is accurately analyzed by using the prediction model based on the battery performance prediction value, and the impact of different process parameters on the battery performance is fully explored through the model; therefore, the whole scheme can obtain accurate battery performance prediction results.
  • the processing module 400 is further used to process the process parameters using the trained battery correction value prediction model to obtain the correction value of the impact of the manufacturing process corresponding to the process parameter on the battery performance of the battery.
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values.
  • the processing module 400 is also used to obtain sample process parameters of several sample batteries; based on the sample process parameters, obtain the battery performance of several sample batteries, and generate a correspondence between the process parameters and the battery performance; train a machine learning model based on the correspondence between the process parameters and the battery performance to obtain a battery correction value prediction model.
  • the processing module 400 is also used to use the P2D electrochemical-thermal coupling model to process several sample process parameters respectively to obtain the battery performance of each sample battery; associate the sample process parameters and battery performance to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
  • the processing module 400 is also used to obtain the test battery performance corresponding to the measured test process parameters; process the test process parameters through the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction results corresponding to the test process parameters; optimize the initial P2D electrochemical-thermal coupling model according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model.
  • the processing module 400 is also used to compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
  • the processing module 400 is further used to process the process parameters using a P2D electrochemical-thermal coupling model to obtain a correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the process parameter acquisition module 200 is also used to obtain at least one process parameter of the battery manufacturing process including particle size, coating thickness, pole piece size, porosity, compaction density and surface density.
  • the present application also provides a battery performance distribution prediction device, the device comprising:
  • a correction module 840 is used to obtain correction values of the effects of the manufacturing process on the battery performance of the battery corresponding to several groups of process parameters, and obtain several groups of effect correction values;
  • a battery performance acquisition module 860 for obtaining battery performance of a plurality of battery groups according to a plurality of impact correction values
  • the battery performance distribution processing module 880 is used to generate a battery performance distribution prediction result based on the battery performance of several groups of batteries.
  • the battery performance distribution prediction device obtains several groups of process parameters for battery manufacturing, processes them using the battery performance prediction method, obtains the corrected battery performance of several groups of batteries, and analyzes the battery performance distribution results using the corrected battery performance of several groups of batteries.
  • the battery performance prediction method since the battery performance prediction method is used to predict the battery performance, it can obtain accurate corrected battery performance of several groups of batteries, and therefore, it can finally accurately predict the battery performance distribution results of the battery.
  • the several groups of process parameter acquisition modules 820 are also used to acquire process parameter fluctuation data for battery manufacturing; and generate several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
  • the several groups of process parameter acquisition modules 820 are further used to randomly generate several groups of process parameters of the manufacturing process according to the process parameter fluctuation data.
  • Each module in the above-mentioned battery performance prediction device can be implemented in whole or in part by software, hardware and a combination thereof.
  • the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
  • a computer device which may be a server, and its internal structure diagram may be as shown in FIG10.
  • the computer device includes a processor, a memory, and a network interface connected via a system bus.
  • the processor of the computer device is used to provide computing and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system, a computer program, and a database.
  • the internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium.
  • the database of the computer device is used to store data related to the trained model.
  • the network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a battery performance prediction method is implemented.
  • FIG. 10 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied.
  • the specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.
  • a computer device including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
  • the battery performance of the battery is obtained.
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • the trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
  • the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • the P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  • a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
  • the battery performance of the battery is obtained.
  • the trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
  • the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
  • test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
  • the P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  • a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
  • the battery performance of the battery is obtained.
  • the trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
  • the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
  • test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
  • the P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  • At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  • a computer device including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • the processor when the processor executes the computer program, the processor further implements the following steps:
  • process parameter fluctuation data several groups of process parameters of the manufacturing process are randomly generated.
  • a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
  • process parameter fluctuation data several groups of process parameters of the manufacturing process are randomly generated.
  • a computer program product comprising a computer program, which, when executed by a processor, implements the following steps:
  • process parameter fluctuation data several groups of process parameters of the manufacturing process are randomly generated.
  • Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Manufacturing & Machinery (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Electrochemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Secondary Cells (AREA)

Abstract

A battery performance prediction method and apparatus, a computer device, a storage medium, and a computer program product. The battery performance prediction method comprises: first acquiring a process parameter of battery manufacturing, then acquiring a correction value of an impact of a manufacturing process corresponding to the process parameter on the battery performance of a battery, and finally obtaining the battery performance of the battery according to the correction value of the impact. The whole solution takes into consideration the impact of the manufacturing process corresponding to the process parameter on the battery performance, so that the battery performance of the battery can be accurately predicted. The battery performance distribution prediction method and apparatus, the computer device, the storage medium, and the computer program product can achieve accurate prediction of battery performance distribution.

Description

电池性能预测方法与电池性能分布预测方法Battery performance prediction method and battery performance distribution prediction method 技术领域Technical Field
本申请涉及电池领域，具体涉及一种电池性能预测方法、装置、计算机设备、存储介质以及计算机程序产品；还涉及一种电池性能分布预测方法、装置、计算机设备、存储介质以及计算机程序产品。The present application relates to the field of batteries, and specifically to a battery performance prediction method, device, computer equipment, storage medium and computer program product; it also relates to a battery performance distribution prediction method, device, computer equipment, storage medium and computer program product.
背景技术Background technique
随着科学技术的进步与发展，目前电池已经广泛应用于人们生产与生活中，为了更好的使用电池常常需要对电池性能进行预测。With the progress and development of science and technology, batteries have been widely used in people's production and life. In order to better use batteries, it is often necessary to predict battery performance.
传统对电池性能预测一般是基于固定电池设计参数进行的。然而，电池的电池性能在实际应用中受到多种因素的影响，因此传统的基于固定电池设计参数的电池性能预测方案无法准确预测出实际制造电池的电池性能。Traditional battery performance prediction is generally based on fixed battery design parameters. However, the battery performance of a battery is affected by many factors in actual applications. Therefore, the traditional battery performance prediction scheme based on fixed battery design parameters cannot accurately predict the battery performance of the actual manufactured battery.
可见，目前亟需一种电池性能预测方案，以实现对电池性能的准确预测。It can be seen that there is an urgent need for a battery performance prediction solution to achieve accurate prediction of battery performance.
发明内容Summary of the invention
鉴于上述问题，本申请提供一种电池性能预测方法、装置、计算机设备、存储介质以及计算机程序产品，以实现对电池性能的准确预测；以及提供一种电池性能分布预测方法、装置、计算机设备、存储介质以及计算机程序产品，以实现对电池性能分布的准确预测。In view of the above problems, the present application provides a battery performance prediction method, apparatus, computer equipment, storage medium and computer program product to achieve accurate prediction of battery performance; and provides a battery performance distribution prediction method, apparatus, computer equipment, storage medium and computer program product to achieve accurate prediction of battery performance distribution.
第一方面，本申请提供一种电池性能预测方法，方法包括：In a first aspect, the present application provides a battery performance prediction method, the method comprising:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
本申请实施例的技术方案中，先获取电池制造的工艺参数，再获取工艺参数对应的制造工艺对电池的电池性能的影响修正值，最后根据影响修正值的得到电池的电池性能。整个方案中，考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测电池的电池性能。In the technical solution of the embodiment of the present application, the process parameters of battery manufacturing are first obtained, and then the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value. In the whole solution, the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
在一些实施例中，获取工艺参数对应的制造工艺对电池的电池性能的影响修正值包括：In some embodiments, obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺参数对应的制造工艺对电池的电池性能的影响修正值。The trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
本申请实施例的技术方案中，采用已训练的电池修正值预测模型准确得到工艺参数对应的影响修正值，通过模型充分挖掘不同工艺参数对电池性能的影响，能够得到准确的工艺参数对应的制造工艺对电池的电池性能的影响修正值，因此最终可以准确预测电池的电池性能。In the technical solution of the embodiment of the present application, a trained battery correction value prediction model is used to accurately obtain the impact correction value corresponding to the process parameters. The model fully explores the influence of different process parameters on battery performance, and can obtain the accurate correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery. Therefore, the battery performance of the battery can be accurately predicted in the end.
在一些实施例中，获得电池修正值预测模型的方式包括：In some embodiments, the method of obtaining the battery correction value prediction model includes:
电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
本申请实施例的技术方案中，电池性能预测值预测模型是基于电池制造的工艺参数与电池性能预测值对应关系训练得到，这样电池性能预测值预测模型可以准确表征工艺参数对电池性能的影响，因此，最终训练得到的电池性能预测值预测模型可以支持准确预测工艺参数对电池性能的影响修正值。In the technical solution of the embodiment of the present application, the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values, so that the battery performance prediction value prediction model can accurately characterize the influence of the process parameters on the battery performance. Therefore, the battery performance prediction value prediction model finally trained can support the accurate prediction of the correction value of the influence of the process parameters on the battery performance.
在一些实施例中，获得电池修正值预测模型的方式包括：In some embodiments, the method of obtaining the battery correction value prediction model includes:
获取若干个样本电池的样本工艺参数；Obtaining sample process parameters of several sample batteries;
根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系；According to the sample process parameters, the battery performance of several sample batteries is obtained, and the corresponding relationship between the process parameters and the battery performance is generated;
根据工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。The machine learning model is trained according to the correspondence between process parameters and battery performance to obtain a battery correction value prediction model.
本申请实施例的技术方案中，先获取若干个样本电池的样本工艺参数，根据样本工艺参数获取样本电池的电池性能对应关系，构建样本电池的工艺参数与电池性能对应关系，以该构建的对应关系训练训练机器学习模型，得到电池修正值预测模型。在这里，采用样本数据的方式来进行机器学习的训练，最终训练得到的电池修正值预测模型能够准确表征工艺参数与电池性能之间的对应关系，因此可以支持最终准确预测电池的电池性能。In the technical solution of the embodiment of the present application, sample process parameters of several sample batteries are first obtained, and the corresponding relationship between the battery performance of the sample batteries is obtained according to the sample process parameters, and the corresponding relationship between the process parameters and the battery performance of the sample batteries is constructed. The machine learning model is trained with the constructed corresponding relationship to obtain a battery correction value prediction model. Here, machine learning training is carried out using sample data, and the battery correction value prediction model finally trained can accurately characterize the corresponding relationship between the process parameters and the battery performance, and thus can support the final accurate prediction of the battery performance of the battery.
在一些实施例中，根据样本工艺参数，获取若干个样本电池的电池性 能，生成工艺参数与电池性能的对应关系包括：In some embodiments, according to the sample process parameters, the battery performance of several sample batteries is obtained, and the corresponding relationship between the process parameters and the battery performance is generated, including:
采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；The P2D electrochemical-thermal coupling model was used to process several sample process parameters to obtain the battery performance of each sample battery;
关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。The sample process parameters and battery performance are associated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
本申请实施例的技术方案中，采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能，再将各样本电池的电池性能与样本工艺参数关联，即得到样本电池的工艺参数与电池性能对应关系。整个方案中，通过P2D电化学-热耦合模型可以快速得到样本电池的电池性能，提高处理效率。In the technical solution of the embodiment of the present application, the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery, and then the battery performance of each sample battery is associated with the sample process parameters, that is, the corresponding relationship between the process parameters and the battery performance of the sample battery is obtained. In the whole solution, the battery performance of the sample battery can be quickly obtained through the P2D electrochemical-thermal coupling model, thereby improving the processing efficiency.
在一些实施例中，P2D电化学-热耦合模型的获得方式包括：In some embodiments, the P2D electrochemical-thermal coupling model is obtained by:
获取测量得到的测试工艺参数对应的测试电池性能；Obtaining the test battery performance corresponding to the measured test process parameters;
通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；The test process parameters are processed through the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction results corresponding to the test process parameters;
根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The initial P2D electrochemical-thermal coupling model was optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model.
本申请实施例的技术方案中，以测量得到的测试工艺参数对应的测试电池性能对初始P2D电化学-热耦合模型进行优化，以得到更加符合场景需求、更加准确的P2D电化学-热耦合模型。In the technical solution of the embodiment of the present application, the initial P2D electrochemical-thermal coupling model is optimized based on the test cell performance corresponding to the measured test process parameters to obtain a more accurate P2D electrochemical-thermal coupling model that better meets the scenario requirements.
在一些实施例中，根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型，包括：In some embodiments, the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model, including:
对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；Compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data;
识别电池性能差异数据对应的波动工艺参数；Identify fluctuating process parameters corresponding to battery performance difference data;
获取波动工艺参数对应的修正因子；Obtaining correction factors corresponding to fluctuating process parameters;
根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。The initial P2D electrochemical-thermal coupling model is adjusted according to the correction factor to obtain the P2D electrochemical-thermal coupling model.
本申请实施例的技术方案中，对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据，识别出产生电池性能差异对应的波动工艺 参数，并获取该指标对应的修正因子，以该修正因子调整初始P2D电化学-热耦合模型，得到更加准确的P2D电化学-热耦合模型。In the technical solution of the embodiment of the present application, the test battery performance is compared with the initial battery performance prediction results to obtain battery performance difference data, identify the fluctuating process parameters corresponding to the battery performance difference, and obtain the correction factor corresponding to the indicator. The initial P2D electrochemical-thermal coupling model is adjusted with the correction factor to obtain a more accurate P2D electrochemical-thermal coupling model.
在一些实施例中，获取工艺参数对应的制造工艺对电池的电池性能的影响修正值包括：In some embodiments, obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。The P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
本申请实施例的技术方案中，采用P2D电化学-热耦合模型来直接对工艺参数进行处理，由于P2D电化学-热耦合模型是一种性能稳定、且电池性能预测准确的模型，其能够准确得到工艺参数对电池的电池性能的影响修正值。In the technical solution of the embodiment of the present application, the P2D electrochemical-thermal coupling model is used to directly process the process parameters. Since the P2D electrochemical-thermal coupling model is a model with stable performance and accurate battery performance prediction, it can accurately obtain the correction value of the impact of the process parameters on the battery performance of the battery.
在一些实施例中，获取电池制造的工艺参数包括：In some embodiments, obtaining process parameters for battery manufacturing includes:
获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
本申请实施例的技术方案中，本申请实施例的技术方案中，工艺参数包括粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度多个维度的指标，分析不同工艺参数对电池性能的影响。In the technical solution of the embodiment of the present application, in the technical solution of the embodiment of the present application, the process parameters include indicators of multiple dimensions such as particle size, coating thickness, pole piece size, porosity, compaction density and surface density, and the influence of different process parameters on battery performance is analyzed.
第二方面，本申请提供一种电池性能分布预测方法，方法包括：In a second aspect, the present application provides a method for predicting battery performance distribution, the method comprising:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
本申请实施例的技术方案中，获取电池制造的若干组工艺参数，针对该若干组工艺参数分别获取对应的制造工艺对电池的电池性能的影响修正值，再基于若干组影响修正值，得到若干组电池的电池性能，分析这些电池的电池性能，生成若干组电池对应的电池性能分布预测结果。整个方案中，针对电池的若干组工艺参数，分别考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测若干组电池的电池性能，因此，可以实现准确的电池性能分布预测。In the technical solution of the embodiment of the present application, several groups of process parameters for battery manufacturing are obtained, and the corresponding manufacturing process's influence correction values on the battery performance of the battery are obtained for the several groups of process parameters, and then the battery performance of several groups of batteries is obtained based on the several groups of influence correction values, and the battery performance of these batteries is analyzed to generate the battery performance distribution prediction results corresponding to the several groups of batteries. In the whole solution, for several groups of process parameters of the battery, the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered respectively, and the battery performance of several groups of batteries can be accurately predicted, so that accurate battery performance distribution prediction can be achieved.
在一些实施例中，获取电池制造的若干组工艺参数，包括：In some embodiments, several sets of process parameters for battery manufacturing are obtained, including:
获取电池制造的工艺参数波动数据；Obtain process parameter fluctuation data for battery manufacturing;
根据工艺参数波动数据，生成电池制造的若干组工艺参数。According to the process parameter fluctuation data, several groups of process parameters for battery manufacturing are generated.
本申请实施例的技术方案中，先获取电池制造的工艺参数波动数据，根据电池制造的工艺参数波动数据来生成若干组工艺参数，能够充分仿真在实际电池制造中可能对应的工艺参数，从而使后续得到的电池性能分布预测结果更加符合真实情况。In the technical solution of the embodiment of the present application, the process parameter fluctuation data of battery manufacturing is first obtained, and several groups of process parameters are generated according to the process parameter fluctuation data of battery manufacturing. This can fully simulate the process parameters that may correspond in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
在一些实施例中，根据工艺参数波动数据，生成电池制造的若干组工艺参数，包括：In some embodiments, several sets of process parameters for battery manufacturing are generated based on the process parameter fluctuation data, including:
根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
本申请实施例的技术方案中，采用随机生成的方式来生成若干组工艺参数，随机生成的方式一方面丰富了工艺参数的数量；另一方面，通过随机生成的方式使得生成的工艺参数更加贴近实际电池制造中对应工艺参数的真实情况，从而使后续得到的电池性能分布预测结果更加符合真实情况。In the technical solution of the embodiment of the present application, a random generation method is adopted to generate several groups of process parameters. On the one hand, the random generation method enriches the number of process parameters; on the other hand, the random generation method makes the generated process parameters closer to the actual situation of the corresponding process parameters in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
第三方面，本申请还提供一种电池性能预测装置，装置包括：In a third aspect, the present application further provides a battery performance prediction device, the device comprising:
工艺参数获取模块，用于获取电池制造的工艺参数；A process parameter acquisition module is used to obtain process parameters for battery manufacturing;
处理模块，用于获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；A processing module, used to obtain a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
性能预测模块，用于根据影响修正值，得到电池的电池性能。The performance prediction module is used to obtain the battery performance of the battery according to the impact correction value.
第四方面，本申请还提供一种电池性能分布预测装置，装置包括：In a fourth aspect, the present application further provides a battery performance distribution prediction device, the device comprising:
若干组工艺参数获取模块，用于获取电池制造的若干组工艺参数；Several groups of process parameter acquisition modules, used to acquire several groups of process parameters for battery manufacturing;
修正模块，用于获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；A correction module, used to obtain correction values of the effects of the manufacturing process on the battery performance of the battery corresponding to several groups of process parameters, and obtain several groups of influence correction values;
电池性能获取模块，用于根据若干影响修正值，得到若干组电池的电池性能；A battery performance acquisition module, used to obtain battery performance of several groups of batteries according to several impact correction values;
电池性能分布处理模块，用于基于若干组电池的电池性能，生成电池性能分布预测结果。The battery performance distribution processing module is used to generate battery performance distribution prediction results based on the battery performance of several groups of batteries.
第五方面，本申请还提供了一种计算机设备。计算机设备包括存储器和处理器，存储器存储有计算机程序，处理器执行计算机程序时实现以下步骤：In a fifth aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
第六方面，本申请还提供了一种计算机设备。计算机设备包括存储器和处理器，存储器存储有计算机程序，处理器执行计算机程序时实现以下步骤：In a sixth aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
第七方面，本申请还提供了一种计算机可读存储介质。计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现以下步骤：In a seventh aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
第八方面，本申请还提供了一种计算机可读存储介质。计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现以下步骤：In an eighth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
第九方面，本申请还提供了一种计算机程序产品。计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现以下步骤：In a ninth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
第十方面，本申请还提供了一种计算机程序产品。计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现以下步骤：In a tenth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
上述电池性能预测装置、计算机设备、存储介质以及计算机程序产品中，先获取电池制造的工艺参数，再获取工艺参数对应的制造工艺对电池的电池性能的影响修正值，最后根据影响修正值的得到电池的电池性能。整个方案中，考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测电池的电池性能。In the above-mentioned battery performance prediction device, computer equipment, storage medium and computer program product, the process parameters of battery manufacturing are first obtained, and then the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value. In the whole scheme, the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
上述电池性能分布预测装置、计算机设备、存储介质以及计算机程序产品，获取电池制造的若干组工艺参数，针对该若干组工艺参数分别获取对应的制造工艺对电池的电池性能的影响修正值，再基于若干组影响修正值，得到若干组电池的电池性能，分析这些电池的电池性能，生成若干组电池对应的电池性能分布预测结果。整个方案中，针对电池的若干组工艺参数，分别考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测若干组电池的电池性能，因此，可以实现准确的电池性能分布预测。The battery performance distribution prediction device, computer equipment, storage medium and computer program product obtain several groups of process parameters for battery manufacturing, obtain the corresponding manufacturing process's impact correction values on the battery performance of the battery for the several groups of process parameters, and then obtain the battery performance of several groups of batteries based on the several groups of impact correction values, analyze the battery performance of these batteries, and generate the battery performance distribution prediction results corresponding to the several groups of batteries. In the whole scheme, for several groups of process parameters of the battery, the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered respectively, and the battery performance of several groups of batteries can be accurately predicted, so that accurate battery performance distribution prediction can be achieved.
上述说明仅是本申请技术方案的概述，为了能够更清楚了解本申请的技术手段，而可依照说明书的内容予以实施，并且为了让本申请的上述和其它目的、特征和优点能够更明显易懂，以下特举本申请的具体实施方式。The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
通过阅读对下文优选实施方式的详细描述，各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。附图仅用于示出优选实施方式的目的，而并不认为是对本申请的限制。而且在全部附图中，用相同的附图标号表示相同的部件。在附图中：Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Moreover, the same reference numerals are used throughout the drawings to represent the same components. In the drawings:
图1为一个实施例中电池性能预测方法的应用环境图；FIG1 is a diagram showing an application environment of a battery performance prediction method according to an embodiment;
图2为一个实施例中电池性能预测方法的流程示意图；FIG2 is a schematic diagram of a flow chart of a battery performance prediction method in one embodiment;
图3为另一个实施例中电池性能预测方法的流程示意图；FIG3 is a schematic flow chart of a battery performance prediction method in another embodiment;
图4为又一个实施例中电池性能预测方法的流程示意图；FIG4 is a schematic flow chart of a battery performance prediction method in yet another embodiment;
图5为一个实施例中电池性能分布预测方法的流程示意图；FIG5 is a schematic diagram of a flow chart of a method for predicting battery performance distribution in one embodiment;
图6为另一个实施例中电池性能分布预测方法的流程示意图；FIG6 is a schematic flow chart of a method for predicting battery performance distribution in another embodiment;
图7为一个实施例中电池性能分布预测方法在实际应用中技术构思示意图；FIG7 is a schematic diagram of the technical concept of a battery performance distribution prediction method in practical application in one embodiment;
图8为一个实施例中电池性能预测装置的结构框图；FIG8 is a structural block diagram of a battery performance prediction device in one embodiment;
图9为一个实施例中电池性能分布预测装置的结构框图；FIG9 is a structural block diagram of a battery performance distribution prediction device in one embodiment;
图10为一个实施例中计算机设备的内部结构图；FIG10 is a diagram showing the internal structure of a computer device in one embodiment;
具体实施方式Detailed ways
下面将结合附图对本申请技术方案的实施例进行详细的描述。以下实施例仅用于更加清楚地说明本申请的技术方案，因此只作为示例，而不能以此来限制本申请的保护范围。The following embodiments of the technical solution of the present application will be described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.
除非另有定义，本文所使用的所有的技术和科学术语与属于本申请的技术领域的技术人员通常理解的含义相同；本文中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本申请；本申请的说明书和权利要求书及上述附图说明中的术语“包括”和“具有”以及它们的任何变形，意图在于覆盖不排他的包含。Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.
在本文中提及“实施例”意味着，结合实施例描述的特定特征、结构或特性可以包含在本申请的至少一个实施例中。在说明书中的各个位置出现该短语并不一定均是指相同的实施例，也不是与其它实施例互斥的独立的或备选的实施例。本领域技术人员显式地和隐式地理解的是，本文所描述的实施例可以与其它实施例相结合。Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
目前，从市场形势的发展来看，动力电池的应用越加广泛。动力电池不仅被应用于水力、火力、风力和太阳能电站等储能电源***，而且还被广泛应用于电动自行车、电动摩托车、电动汽车等电动交通工具，以及军事装备和航空航天等多个领域。随着动力电池应用领域的不断扩大，其市场的需求量也在不断地扩增。与此同时，人们在设计电池时能够得到更多 的数据支持和参考来设计出更高性能或者更加符合实际应用需要的电池。对于此，有研究人员提出从分析电池性能入手来优化电池设计以及进行电池研发，于是就有学者提出了电池性能测试的方案，传统的方案中都是采用实验测试的方式测试电池性能，这种方式虽然可以测试得到电池的性能，但是一方面电池在实际生产过程时，从材料导入、搅拌、涂布到最后化成，经过多种工艺流程，每一步工艺流程均会带来工艺参数的变动，对最后的电池性能(例如电池容量)产生影响；另一方面，传统技术中对单体电池电性能的预测主要停留在依靠一组参数预测同一组电池的电性能。这种点状的性能预测无法提供批次电池性能的预测，无法使设计人员直观了解批次电池性能情况。At present, from the perspective of the development of the market situation, the application of power batteries is becoming more and more extensive. Power batteries are not only used in energy storage power systems such as hydropower, thermal power, wind power and solar power stations, but also widely used in electric vehicles such as electric bicycles, electric motorcycles, electric vehicles, as well as military equipment and aerospace and other fields. With the continuous expansion of the application field of power batteries, the market demand is also constantly expanding. At the same time, people can get more data support and reference when designing batteries to design batteries with higher performance or more in line with the needs of actual applications. For this, some researchers have proposed to optimize battery design and conduct battery research and development by analyzing battery performance, so some scholars have proposed a battery performance test scheme. In traditional schemes, battery performance is tested by experimental testing. Although this method can test the performance of the battery, on the one hand, in the actual production process of the battery, from material introduction, stirring, coating to final formation, through a variety of process flows, each step of the process will bring about changes in process parameters, which will affect the final battery performance (such as battery capacity); on the other hand, the prediction of the electrical performance of single cells in traditional technology mainly relies on a set of parameters to predict the electrical performance of the same group of batteries. This point-based performance prediction cannot provide a prediction of batch battery performance, and cannot enable designers to intuitively understand the batch battery performance.
本申请发明人注意到除了电池设计影响最终电池性能之外，电池在制造过程中工艺上相关参数的波动同样会影响最终制造出电池的电池性能，因此，有必要考虑电池制造的工艺参数对电池性能的影响。The inventors of the present application have noticed that in addition to the battery design affecting the final battery performance, fluctuations in process-related parameters during the battery manufacturing process will also affect the battery performance of the final battery. Therefore, it is necessary to consider the impact of battery manufacturing process parameters on battery performance.
本申请发明人提出全新的电池性能预测方案。先获取电池制造的工艺参数，再获取工艺参数对应的制造工艺对电池的电池性能的影响修正值，最后根据影响修正值的得到电池的电池性能。整个方案中，考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测电池的电池性能。The inventor of this application proposes a new battery performance prediction scheme. First, the process parameters of battery manufacturing are obtained, then the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained, and finally the battery performance of the battery is obtained according to the correction value. In the whole scheme, the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the battery performance of the battery can be accurately predicted.
本申请提供一种电池性能预测方法，其具体可以应用于图1的场景中，终端102发送电池性能预测请求至服务器104，服务器104响应该电池性能预测请求，提取电池性能预测请求中携带的电池制造的工艺参数，再获取工艺参数对应的制造工艺对电池的电池性能的影响修正值，最后根据影响修正值的得到电池的电池性能。进一步的，服务器104可以将电池的电池性能发送至终端102，终端102显示电池的电池性能。The present application provides a battery performance prediction method, which can be specifically applied to the scenario of Figure 1, where the terminal 102 sends a battery performance prediction request to the server 104, and the server 104 responds to the battery performance prediction request, extracts the process parameters of the battery manufacturing carried in the battery performance prediction request, and then obtains the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery, and finally obtains the battery performance of the battery according to the correction value. Further, the server 104 can send the battery performance of the battery to the terminal 102, and the terminal 102 displays the battery performance of the battery.
可以理解的是，在上述电池性能预测方案可以直接应用于终端，即由终端来独自完成性能预测，具体可以是用户在终端侧操作，输入电池制造的工艺参数至终端以进行电池性能预测，终端响应用户的电池性能预测请求，执行上述电池性能预测方法，得到电池性能预测结果，再将电池性能预测结果推送给到用户，例如可以在终端显示界面上显示电池的电池性能。其具体的处理过程与上述内容类似，在此不再赘述。It is understandable that the above-mentioned battery performance prediction scheme can be directly applied to the terminal, that is, the terminal can complete the performance prediction alone. Specifically, the user can operate on the terminal side, input the process parameters of battery manufacturing to the terminal to perform battery performance prediction, and the terminal responds to the user's battery performance prediction request, executes the above-mentioned battery performance prediction method, obtains the battery performance prediction result, and then pushes the battery performance prediction result to the user, for example, the battery performance of the battery can be displayed on the terminal display interface. The specific processing process is similar to the above content and will not be repeated here.
如图2所示，本申请提供一种电池性能预测方法，包括：As shown in FIG2 , the present application provides a battery performance prediction method, comprising:
S200：获取电池制造的工艺参数。S200: Obtaining process parameters for battery manufacturing.
工艺参数是指在电池制造工艺相关的参数，其具体可以包括但不限于粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度等，其中粒径是指来料中颗粒物质的直径；涂布厚度是指电芯生产中涂布的厚度；极片尺寸是指卷绕电芯中极片的尺寸，其具体包括极片的宽度等；孔隙率是指来料中块状材料的孔隙体积与材料在自然状态下总体积的百分比；压实密度是指电池生产中面密度与材料的厚度比值，在电芯生产中压实密度对电池性能有较大的影响，压实密度不光和颗粒的大小、密度有关系，还和粒子的级配有关系，压实密度大的一般都有很好的粒子正态分布，可以认为，工艺条件一定的条件下，压实密度越大，电池的容量越高；面密度是指工程材料领域中定厚度的物质单位面积的质量。工艺参数的基础数据可以通过电池制造过程中来料的相关波动、电池制造公司内部的来料检测***、或根据电池制造来料(原料)供应商提供的来料波动数据获取。由于电池实际制造过程中可能存在制造工艺上的波动，并且这些工艺上的波动很有可能会影响最终电池性能，因此，在这里需要先获取电池制造的工艺参数。Process parameters refer to parameters related to the battery manufacturing process, which may include but are not limited to particle size, coating thickness, pole piece size, porosity, compaction density and surface density, among which particle size refers to the diameter of the granular material in the incoming material; coating thickness refers to the coating thickness in the production of the battery cell; pole piece size refers to the size of the pole piece in the wound battery cell, which specifically includes the width of the pole piece, etc.; porosity refers to the percentage of the pore volume of the block material in the incoming material to the total volume of the material in the natural state; compaction density refers to the ratio of surface density to material thickness in battery production. In battery cell production, compaction density has a great influence on battery performance. Compaction density is not only related to the size and density of the particles, but also to the gradation of the particles. Generally, particles with large compaction density have a good normal distribution. It can be considered that under certain process conditions, the greater the compaction density, the higher the battery capacity; surface density refers to the mass per unit area of a material with a certain thickness in the field of engineering materials. The basic data of process parameters can be obtained through the relevant fluctuations of incoming materials in the battery manufacturing process, the incoming material detection system within the battery manufacturing company, or the incoming material fluctuation data provided by the supplier of incoming materials (raw materials) for battery manufacturing. Since there may be fluctuations in the manufacturing process during the actual manufacturing of the battery, and these fluctuations in the process are likely to affect the final battery performance, it is necessary to first obtain the process parameters for battery manufacturing.
S400：获取工艺参数对应的制造工艺对电池的电池性能的影响修正值。S400: Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery.
工艺参数不同其对应的制造工艺也会存在区别，在制造工艺发生变化(波动)时，会影响最终生产出电池的电池性能。即在实际电池制造中，电池的电池性能除了受到电池设计参数的影响之外(主要影响因素)，还受到电池制造工艺的影响。在这里，考虑工艺参数对应的制造工艺对电池性能影响，获取工艺参数对应的制造工艺对电池的电池性能的影响修正值。具体来说，可以基于历史数据构建出工艺参数与电池性能的影响修正值之间对应关系，在需要获取电池性能的影响修正值时，基于获取的工艺参数以及上述对应关系，直接得到影响修正值。进一步的，该工艺参数与电池性能的影响修正值之间对应关系具体可以是简单的对应关系表，其还可以是具体训练得到的机器学习模型，通过机器学习模型的方式来充分挖掘、表征工艺参数与电池性能的影响修正值之间对应关系。Different process parameters will have different corresponding manufacturing processes. When the manufacturing process changes (fluctuates), it will affect the battery performance of the final battery produced. That is, in actual battery manufacturing, the battery performance of the battery is not only affected by the battery design parameters (the main influencing factor), but also by the battery manufacturing process. Here, the impact of the manufacturing process corresponding to the process parameters on the battery performance is considered, and the impact correction value of the manufacturing process corresponding to the process parameters on the battery performance of the battery is obtained. Specifically, the corresponding relationship between the process parameters and the impact correction values of the battery performance can be constructed based on historical data. When it is necessary to obtain the impact correction values of the battery performance, the impact correction values are directly obtained based on the obtained process parameters and the above-mentioned corresponding relationship. Furthermore, the corresponding relationship between the process parameters and the impact correction values of the battery performance can be a simple corresponding relationship table, or it can be a machine learning model obtained by specific training. The corresponding relationship between the process parameters and the impact correction values of the battery performance is fully explored and characterized by the machine learning model.
S600：根据影响修正值，得到电池的电池性能。S600: Obtaining battery performance of the battery according to the impact correction value.
S400得到工艺参数对应的制造工艺对电池的电池性能的影响修正值，在此基础上，再基于电池设计参数对应的基准电池性能，即可得到电池的电池性能，即通过综合考虑基准电池性能以及影响修正值，得到电池的电池性能。S400 obtains the correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery. On this basis, based on the benchmark battery performance corresponding to the battery design parameters, the battery performance of the battery can be obtained, that is, the battery performance of the battery is obtained by comprehensively considering the benchmark battery performance and the correction value.
基准电池性能是基于电池设计参数得到的，其具体是指在理想电池制造环境下电池设计参数对应的电池性能。电池设计参数是指在电池在设计人员设计下规定的相关参数，一般来说，不同型号的电池其对应不同的电池设计参数，设计人员在设计、开发一款新电池时，都会给出该款新电池对应的具体设计参数。基于电池设计参数即可获取到基准电池性能，基准电池性能是指仅考虑设计参数条件的情况下制造的电池的电池性能，或者可以简单理解为在实验状态下、不考虑电池制造工艺影响下电池的基准电池性能。在这里，在常规仅考虑电池设计参数对应的基准电池性能基础上，进一步还考虑电池制造工艺中工艺参数对电池性能的影响，即综合基准电池性能和影响修正值，得到电池的修正电池性能。The benchmark battery performance is obtained based on the battery design parameters, which specifically refers to the battery performance corresponding to the battery design parameters under an ideal battery manufacturing environment. Battery design parameters refer to the relevant parameters specified by the battery designer when it is designed. Generally speaking, different types of batteries correspond to different battery design parameters. When designers design and develop a new battery, they will give the specific design parameters corresponding to the new battery. The benchmark battery performance can be obtained based on the battery design parameters. The benchmark battery performance refers to the battery performance of the battery manufactured only under the conditions of the design parameters, or it can be simply understood as the benchmark battery performance of the battery under experimental conditions without considering the influence of the battery manufacturing process. Here, on the basis of the conventional consideration of only the benchmark battery performance corresponding to the battery design parameters, the influence of the process parameters in the battery manufacturing process on the battery performance is further considered, that is, the benchmark battery performance and the influence correction value are combined to obtain the corrected battery performance of the battery.
上述电池性能预测方法，先获取电池制造的工艺参数，采用电池性能预测值预测模型对工艺参数进行处理，得到处理结果，再在电池设计参数对应的基准电池性能基础上考虑制造工艺对电池性能的影响，得到电池的电池性能。整个方案中，一方面考虑制造工艺对电池性能的影响；另一方面，采用基于电池性能预测值预测模型准确分析工艺参数对电池性能的影响，通过模型充分挖掘不同工艺参数对电池性能的影响；因此，整个方案可以得到准确的电池性能预测结果。The above-mentioned battery performance prediction method first obtains the process parameters of battery manufacturing, uses the battery performance prediction value prediction model to process the process parameters, obtains the processing results, and then considers the impact of the manufacturing process on the battery performance based on the benchmark battery performance corresponding to the battery design parameters to obtain the battery performance of the battery. In the whole scheme, on the one hand, the impact of the manufacturing process on the battery performance is considered; on the other hand, the impact of the process parameters on the battery performance is accurately analyzed by using the prediction model based on the battery performance prediction value, and the impact of different process parameters on the battery performance is fully explored through the model; therefore, the whole scheme can obtain accurate battery performance prediction results.
在一些实施例中，如图3所示，S400包括：采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺参数对应的制造工艺对电池的电池性能的影响修正值。In some embodiments, as shown in FIG. 3 , S400 includes: using the trained battery correction value prediction model to process the process parameters to obtain a correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
电池修正值预测模型是用于根据工艺参数得到制造工艺对电池的电池性能的影响修正值的模型，简单来说，电池修正值预测模型表征工艺参数-电池性能的影响修正值对应关系，在实际应用中，可以基于该训练得到的电池修正值预测模型以及工艺参数，来获取在当前工艺参数下对应的电池性能的影响修正值，以充分考虑电池制造工艺对电池的电池性能影响。在这里，电池修正值预测模型是预先训练得到的模型，其具体可以基于历史 数据训练得到。例如可以是基于历史数据分析得到工艺参数-电池性能的影响修正值的对应关系，基于该对应关系对模型进行训练，从而得到该已训练的电池修正值预测模型。更进一步来说，可以是采用工艺参数-电池性能的影响修正值的对应关系对机器学习模型进行训练，充分发挥机器学习模型的学习和分析能力，深度挖掘不同工艺参数对最终电池性能的影响，得到可以准确的电池性能的影响修正值。The battery correction value prediction model is a model for obtaining the correction value of the impact of the manufacturing process on the battery performance of the battery according to the process parameters. In short, the battery correction value prediction model characterizes the corresponding relationship between the process parameters and the correction value of the impact of the battery performance. In practical applications, the correction value of the impact of the battery performance corresponding to the current process parameters can be obtained based on the trained battery correction value prediction model and process parameters to fully consider the impact of the battery manufacturing process on the battery performance of the battery. Here, the battery correction value prediction model is a pre-trained model, which can be specifically trained based on historical data. For example, the corresponding relationship between the process parameter and the correction value of the impact of the battery performance can be obtained based on the analysis of historical data, and the model can be trained based on the corresponding relationship to obtain the trained battery correction value prediction model. Furthermore, the corresponding relationship between the process parameter and the correction value of the impact of the battery performance can be used to train the machine learning model, give full play to the learning and analysis capabilities of the machine learning model, deeply explore the impact of different process parameters on the final battery performance, and obtain accurate correction values for the impact of the battery performance.
本申请实施例的技术方案中，采用已训练的电池修正值预测模型准确得到工艺参数对应的影响修正值，通过模型充分挖掘不同工艺参数对电池性能的影响，能够得到准确的工艺参数对应的制造工艺对电池的电池性能的影响修正值，因此最终可以准确预测电池的电池性能。In the technical solution of the embodiment of the present application, a trained battery correction value prediction model is used to accurately obtain the impact correction value corresponding to the process parameters. The model fully explores the influence of different process parameters on battery performance, and can obtain the accurate correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery. Therefore, the battery performance of the battery can be accurately predicted in the end.
在一些实施例中，获得电池修正值预测模型的方式包括：In some embodiments, the method of obtaining the battery correction value prediction model includes:
电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
本申请实施例的技术方案中，电池性能预测值预测模型是基于电池制造的工艺参数与电池性能预测值对应关系训练得到，这样电池性能预测值预测模型可以准确表征工艺参数对电池性能的影响，因此，最终训练得到的电池性能预测值预测模型可以支持准确预测工艺参数对电池性能的影响修正值。In the technical solution of the embodiment of the present application, the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values, so that the battery performance prediction value prediction model can accurately characterize the influence of the process parameters on the battery performance. Therefore, the battery performance prediction value prediction model finally trained can support the accurate prediction of the correction value of the influence of the process parameters on the battery performance.
在一些实施例中，如图3所示，获得电池修正值预测模型的方式包括：In some embodiments, as shown in FIG3 , the method of obtaining the battery correction value prediction model includes:
S320：获取若干个样本电池的样本工艺参数。S320: Obtain sample process parameters of several sample batteries.
样本电池是指用于训练电池性能预测值预测模型的电池，一般选择数量较少的电池作为样本电池，例如通过随机的方式抽取100个电池作为样本电池。获取这些样本电池在实际制造过程中的工艺参数，即得到样本电池的样本工艺参数。电池在制造过程中，针对相同型号(相同设计参数)的电池，其工艺参数是在一定范围值内波动的，具体可以波动的工艺参数范围值内随机选择/组成多个工艺参数作为样本工艺参数。Sample batteries refer to batteries used to train battery performance prediction models. Generally, a small number of batteries are selected as sample batteries, for example, 100 batteries are randomly selected as sample batteries. The process parameters of these sample batteries in the actual manufacturing process are obtained, that is, the sample process parameters of the sample batteries are obtained. During the manufacturing process of batteries, for batteries of the same model (same design parameters), their process parameters fluctuate within a certain range of values. Specifically, multiple process parameters are randomly selected/composed within the range of process parameter values that can fluctuate as sample process parameters.
S340：根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系。S340: Obtain battery performance of a plurality of sample batteries according to the sample process parameters, and generate a corresponding relationship between the process parameters and the battery performance.
根据样本工艺参数，可以采取目前已有的电池性能预测方法来获取对 应的电池性能，进而生成工艺参数与电池性能的对应关系。例如，可以采用样本工艺参数来设计电池，再采用实验测量方式对设计电池进行测试，得到样本电池的电池性能。再例如，可以是根据样本工艺参数仿真出虚拟电池，再获取仿真电池对应的电池性能。再例如，还可以是将样本工艺参数输入至电化学模型，得到样本电池的电池性能。According to the sample process parameters, the currently available battery performance prediction method can be used to obtain the corresponding battery performance, and then generate the corresponding relationship between the process parameters and the battery performance. For example, the sample process parameters can be used to design a battery, and then the designed battery can be tested by experimental measurement to obtain the battery performance of the sample battery. For another example, a virtual battery can be simulated according to the sample process parameters, and then the battery performance corresponding to the simulated battery can be obtained. For another example, the sample process parameters can also be input into the electrochemical model to obtain the battery performance of the sample battery.
获得样本电池的电池性能后，通过关联样本工艺参数和样本电池的电池性能，生成样本电池的工艺参数与电池性能对应关系。一些实施例中，样本电池的工艺参数与电池性能对应关系，可以为工艺参数与电池性能对应关系表。例如可以分别以任意一种工艺参数作为变量，其他工艺参数作为不变量来绘制对应的电池性能变化关系曲线，即得到不同单个工艺参数作为变量的电池性能波动曲线。样本电池的工艺参数与电池性能对应关系可以采用计算方式来得到的一组精准数据集，此数据集作为下一步机器学习模型训练的数据源。以涂布厚度波动为例：仅考虑涂布厚度波动时，获取不同的涂布厚度影响下对应的电池性能，进一步，该不同电池性能可以采用电池性能分布方式表征，电池性能分布具体可以包括容量分布、功率分布、或温度分布等。After obtaining the battery performance of the sample battery, the correspondence between the process parameters and the battery performance of the sample battery is generated by associating the sample process parameters and the battery performance of the sample battery. In some embodiments, the correspondence between the process parameters and the battery performance of the sample battery can be a table of the correspondence between the process parameters and the battery performance. For example, any one of the process parameters can be used as a variable, and the other process parameters can be used as invariants to draw the corresponding battery performance change relationship curve, that is, to obtain a battery performance fluctuation curve with different single process parameters as variables. The correspondence between the process parameters and the battery performance of the sample battery can be obtained by calculation. A set of accurate data sets are used, and this data set is used as the data source for the next step of machine learning model training. Take the coating thickness fluctuation as an example: when only the coating thickness fluctuation is considered, the corresponding battery performance under the influence of different coating thicknesses is obtained. Further, the different battery performance can be characterized by battery performance distribution, and the battery performance distribution can specifically include capacity distribution, power distribution, or temperature distribution.
S360：根据工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。S360: Train a machine learning model based on the correspondence between process parameters and battery performance to obtain a battery correction value prediction model.
在实际应用中，本申请发明人发现同一个电池设计表中，影响电性能的参数众多，每一个设计参数都存在一个分布范围，此设计在制造过程又会引入众多的工艺参数，要把所有的可能性都考虑在内，计算量就会变得非常庞大，为了实现准确、高效的计算在这里引入机器学习的方式来进行计算。机器学习是一门多领域交叉学科，涉及概率论、统计学、逼近论、凸分析、算法复杂度理论等多门学科。专门研究计算机怎样模拟或实现人类的学习行为，以获取新的知识或技能，重新组织已有的知识结构使之不断改善自身的性能。In practical applications, the inventors of this application found that in the same battery design table, there are many parameters that affect the electrical performance. Each design parameter has a distribution range. This design will introduce many process parameters in the manufacturing process. If all possibilities are taken into account, the amount of calculation will become very large. In order to achieve accurate and efficient calculations, machine learning is introduced here to perform calculations. Machine learning is a multi-disciplinary interdisciplinary subject involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in studying how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their own performance.
在本申请中主要利用基于关联规则算法的机器学习，关联规则是用规则去描述两个变量或多个变量之间的关系，是客观反映数据本身性质的方法。它是机器学习的一大类任务，可分为两个阶段，先从资料集中找到高频项目组，再去研究它们的关联规则。其得到的分析结果即是对变量间规 律的总结。在这里，以样本电池的工艺参数与电池性能对应关系作为训练数据来训练机器学习模型，得到电池修正值预测模型。通过机器学习方式，充分挖掘工艺参数波动与电池性能波动之间的相关性。In this application, machine learning based on association rule algorithms is mainly used. Association rules use rules to describe the relationship between two or more variables, and are a method that objectively reflects the nature of the data itself. It is a large category of machine learning tasks, which can be divided into two stages. First, high-frequency project groups are found from the data set, and then their association rules are studied. The analysis results obtained are a summary of the laws between variables. Here, the correspondence between the process parameters and battery performance of the sample battery is used as training data to train the machine learning model to obtain a battery correction value prediction model. Through machine learning, the correlation between process parameter fluctuations and battery performance fluctuations is fully explored.
本申请实施例的技术方案中，先获取若干个样本电池的样本工艺参数，根据样本工艺参数获取样本电池的电池性能对应关系，构建样本电池的工艺参数与电池性能对应关系，以该构建的对应关系训练训练机器学习模型，得到电池修正值预测模型。在这里，采用样本数据的方式来进行机器学习的训练，最终训练得到的电池修正值预测模型能够准确表征工艺参数与电池性能之间的对应关系，因此可以支持最终准确预测电池的电池性能。In the technical solution of the embodiment of the present application, sample process parameters of several sample batteries are first obtained, and the corresponding relationship between the battery performance of the sample batteries is obtained according to the sample process parameters, and the corresponding relationship between the process parameters and the battery performance of the sample batteries is constructed. The machine learning model is trained with the constructed corresponding relationship to obtain a battery correction value prediction model. Here, machine learning training is carried out using sample data, and the battery correction value prediction model finally trained can accurately characterize the corresponding relationship between the process parameters and the battery performance, and thus can support the final accurate prediction of the battery performance of the battery.
在一些实施例中，如图4所示，S340包括：In some embodiments, as shown in FIG. 4 , S340 includes:
S342：采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；S342: using a P2D electrochemical-thermal coupling model to process several sample process parameters respectively to obtain the battery performance of each sample battery;
S344：关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。S344: Correlate the sample process parameters and the battery performance to obtain a corresponding relationship between the process parameters and the battery performance of the sample battery.
P2D电化学-热耦合模型是将P2D电化学模型与热模型进行耦合之后的得到的模型。由于电池充放电过程会产生热量，使电池温度升高，而电池温度的升高会给电池性能带来变化，若忽视这一现象，必然会导致电池性能预测不准确，因此，需要将P2D电化学模型与电池热模型进行耦合，以更加准确预测出电池性能。P2D电化学模型以浓溶液理论和多孔电极理论为理论支撑，依据有限元思想对网格(一种离散方法)进行剖分，然后求解电化学过程中的偏微分方程，可以得到电极电位、电解质电位、电解质浓度等电化学性能；将电化学模型计算得到的热量作为热源整体耦合进热模型中，引起3D热模型中温度的变化，而温度的变化又反馈到电化学模型中，引起电化学模型中与温度相关参数的变化，这些参数的变化进一步引发热源的变化，以此来实现电化学模型与热模型的耦合，最终影响电池的电热性能。The P2D electrochemical-thermal coupling model is a model obtained by coupling the P2D electrochemical model with the thermal model. Since the battery charging and discharging process generates heat, the battery temperature rises, and the increase in battery temperature will change the battery performance. If this phenomenon is ignored, it will inevitably lead to inaccurate prediction of battery performance. Therefore, it is necessary to couple the P2D electrochemical model with the battery thermal model to more accurately predict the battery performance. The P2D electrochemical model is based on the theory of concentrated solution and porous electrode theory. The grid (a discrete method) is divided according to the finite element idea, and then the partial differential equations in the electrochemical process are solved to obtain electrochemical properties such as electrode potential, electrolyte potential, and electrolyte concentration; the heat calculated by the electrochemical model is coupled into the thermal model as a heat source, causing the temperature change in the 3D thermal model, and the temperature change is fed back to the electrochemical model, causing the temperature-related parameters in the electrochemical model to change. The changes in these parameters further trigger the changes in the heat source, thereby realizing the coupling of the electrochemical model and the thermal model, and ultimately affecting the electrothermal performance of the battery.
将样本工艺参数输入至P2D电化学-热耦合模型得到样本电池的电池性能。正是由于P2D电化学-热耦合模型具有上述的性能优势与功能，在将样本工艺参数输入至P2D电化学-热耦合模型之后即可得到电池性能。另 外，需要指出的是虽然P2D电化学-热耦合模型能够基于输入的样本工艺参数得到样本电池的电池性能，但是一方面，P2D电化学-热耦合模型处理速度还是相对较慢，其需要一定的时间才能处理输出电池性能，这在针对数量较少的样本电池数据处理时尚可接受，但是在针对批量、大量电池性能预测时，其已经无法做到最优；另一方面，P2D电化学-热耦合模型无法深度剖析、挖掘不同工艺参数波动与电池性能波动之间的关系，其不具备学习功能，因此，直接采用P2D电化学-热耦合模型得到的电池性能参数对后续电池设计与研发指导作用有限。在本申请中先采用P2D电化学-热耦合模型对少量的样本电池的样本工艺参数进行处理，建立起样本电池的样本工艺参数与电池性能对应关系，再以该样本电池的样本工艺参数与电池性能对应关系作为训练数据训练机器学习模型，通过机器学习模型来挖掘、和分析不同工艺参数波动与电池性能波动之间的关系。关联样本工艺参数以及电池性能，是指将不同样本工艺参数下对应的电池性能进行关联记录。例如可以设置数据表记录下不同样本工艺参数对应电池性能，还可以是采用图形曲线的方式记录下不同样本工艺参数对应电池性能。The sample process parameters are input into the P2D electrochemical-thermal coupling model to obtain the battery performance of the sample battery. It is precisely because the P2D electrochemical-thermal coupling model has the above-mentioned performance advantages and functions that the battery performance can be obtained after the sample process parameters are input into the P2D electrochemical-thermal coupling model. In addition, it should be pointed out that although the P2D electrochemical-thermal coupling model can obtain the battery performance of the sample battery based on the input sample process parameters, on the one hand, the processing speed of the P2D electrochemical-thermal coupling model is still relatively slow, and it takes a certain amount of time to process the output battery performance, which is acceptable for processing a small number of sample battery data, but it can no longer achieve the best performance for batch and large-scale battery performance prediction; on the other hand, the P2D electrochemical-thermal coupling model cannot deeply analyze and explore the relationship between fluctuations in different process parameters and fluctuations in battery performance, and it does not have a learning function. Therefore, the battery performance parameters directly obtained by the P2D electrochemical-thermal coupling model have limited guiding effect on subsequent battery design and development. In this application, the P2D electrochemical-thermal coupling model is first used to process the sample process parameters of a small number of sample batteries, and the corresponding relationship between the sample process parameters of the sample batteries and the battery performance is established. The corresponding relationship between the sample process parameters of the sample batteries and the battery performance is then used as training data to train the machine learning model, and the relationship between the fluctuations of different process parameters and the fluctuations of battery performance is mined and analyzed through the machine learning model. Associating sample process parameters and battery performance refers to associating and recording the corresponding battery performance under different sample process parameters. For example, a data table can be set to record the battery performance corresponding to different sample process parameters, or a graphical curve can be used to record the battery performance corresponding to different sample process parameters.
本申请实施例的技术方案中，采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能，再将各样本电池的电池性能与样本工艺参数关联，即得到样本电池的工艺参数与电池性能对应关系。整个方案中，通过P2D电化学-热耦合模型可以快速得到样本电池的电池性能，提高处理效率。In the technical solution of the embodiment of the present application, the P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery, and then the battery performance of each sample battery is associated with the sample process parameters, that is, the corresponding relationship between the process parameters and the battery performance of the sample battery is obtained. In the whole solution, the battery performance of the sample battery can be quickly obtained through the P2D electrochemical-thermal coupling model, thereby improving the processing efficiency.
在一些实施例中，P2D电化学-热耦合模型的获得方式包括：In some embodiments, the P2D electrochemical-thermal coupling model is obtained by:
获取测量得到的测试工艺参数对应的测试电池性能；通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
初始P2D电化学-热耦合模型是指常规通用的P2D电化学-热耦合模型。测试工艺参数是指用于在实验测试状态测试电池性能对应电池的工艺参数，例如在这里选择采用实验测试的方式来获取非常少量的电池工艺参数与电池性能对应数据，这部分非常少量的电池可以理解为测试电池，这部分测试电池对应的工艺参数即为测试工艺参数，针对这部分测试电池采 用完整、严谨的实验测试方法测试其对应的测试电池性能，该测试电池性能准确度非常高、其与电池真实性电池性能之间的误差基本可以忽略不计，其能够准确表征测试电池的真实性能。同时，将测试工艺参数输入至初始P2D电化学-热耦合模型，得到初始电池性能预测结果，该初始电池性能预测结果是对测试电池进行性能预测得到的结果，属于预测值；对比测试电池性能以及初始电池性能预测结果，分析两者之间差异，根据真实值和预测值的差异对初始P2D电化学-热耦合模型进行优化与修正，得到P2D电化学-热耦合模型，以得到更加准确的P2D电化学-热耦合模型。The initial P2D electrochemical-thermal coupling model refers to the conventional and general P2D electrochemical-thermal coupling model. The test process parameters refer to the process parameters of the battery corresponding to the battery performance tested in the experimental test state. For example, here we choose to use the experimental test method to obtain a very small amount of battery process parameters and battery performance corresponding data. This very small number of batteries can be understood as test batteries. The process parameters corresponding to this part of the test batteries are the test process parameters. For this part of the test batteries, a complete and rigorous experimental test method is used to test the corresponding test battery performance. The test battery performance has a very high accuracy, and the error between it and the battery's true battery performance is basically negligible, which can accurately characterize the true performance of the test battery. At the same time, the test process parameters are input into the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction result, which is the result of performance prediction of the test battery and belongs to the predicted value; the test battery performance and the initial battery performance prediction result are compared, and the difference between the two is analyzed. According to the difference between the actual value and the predicted value, the initial P2D electrochemical-thermal coupling model is optimized and corrected to obtain the P2D electrochemical-thermal coupling model, so as to obtain a more accurate P2D electrochemical-thermal coupling model.
本申请实施例的技术方案中，以测量得到的测试工艺参数对应的测试电池性能对初始P2D电化学-热耦合模型进行优化，以得到更加符合场景需求、更加准确的P2D电化学-热耦合模型。In the technical solution of the embodiment of the present application, the initial P2D electrochemical-thermal coupling model is optimized based on the test cell performance corresponding to the measured test process parameters to obtain a more accurate P2D electrochemical-thermal coupling model that better meets the scenario requirements.
在一些实施例中，根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型，包括：In some embodiments, the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model, including:
对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；识别电池性能差异数据对应的波动工艺参数；获取波动工艺参数对应的修正因子；根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。Compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
对比测试电池性能与初始电池性能预测结果，得到电池性能真实值与电池性能预测值之间的差异数据，分析电池性能差异数据识别导致这些差异对应的波动工艺参数，例如涂布厚度波动对应的电池性能真实值与预测值存在较大差异，或极片尺寸波动对应的电池性能真实值与预测值存在较大差异，识别这些工艺参数对应的修正因子，该修正因子是指不同工艺参数对应电池性能影响程度的修正因子。在P2D电化学-热耦合模型(初始P2D电化学-热耦合模型)中表征有不同工艺参数对电池性能的影响程度，有的工艺参数对电池性能的影响程度高、有的工艺参数对电池性能的影响程度低，在整个模型中已经针对不同工艺参数赋予了不同的权重值或相关系数，在这里，当前电池性能真实值和预测值存在差异时，就表明需要对P2D电化学-热耦合模型中这些工艺参数相关的权重/系数或者常量值进行修正，具体采用修正因子的方式直接进行，具体不同工艺参数对应的修正因 子可以基于历史数据分析得到，也可能通过反馈调节的方式得到，即通过选择不同的修正因子，更新P2D电化学-热耦合模型，以使最终的预测值无限接近真实测量值，以得到更加准确的P2D电化学-热耦合模型。Compare the test battery performance with the initial battery performance prediction results to obtain the difference data between the actual battery performance and the predicted battery performance. Analyze the battery performance difference data to identify the fluctuating process parameters that cause these differences. For example, there is a large difference between the actual battery performance value and the predicted value corresponding to the coating thickness fluctuation, or there is a large difference between the actual battery performance value and the predicted value corresponding to the electrode size fluctuation. Identify the correction factors corresponding to these process parameters, which refer to the correction factors that correspond to the degree of influence of different process parameters on battery performance. In the P2D electrochemical-thermal coupling model (initial P2D electrochemical-thermal coupling model), the influence of different process parameters on battery performance is characterized. Some process parameters have a high influence on battery performance, while some process parameters have a low influence on battery performance. Different weight values or correlation coefficients have been assigned to different process parameters in the entire model. Here, when there is a difference between the actual value and the predicted value of the current battery performance, it indicates that the weights/coefficients or constant values related to these process parameters in the P2D electrochemical-thermal coupling model need to be corrected. Specifically, the correction factors are directly used. The correction factors corresponding to different process parameters can be obtained based on historical data analysis, or they may be obtained through feedback adjustment, that is, by selecting different correction factors, the P2D electrochemical-thermal coupling model is updated so that the final predicted value is infinitely close to the actual measured value, so as to obtain a more accurate P2D electrochemical-thermal coupling model.
在上述实施例中，测试电池具体可以采用以下两种方式获得：一种是通过人为制造不同设计的电池，其具体测试如容量/DCR等性能，分别仿真不同设计的性能，与实测进行对比，分析模型各个参数对结果影响程度及影响力的准确性；另一种，是从生产线上下仓电池中直接选取，具体的，通过捞取已经制造完成的下仓并测试的电池来料/设计/制程数据，根据波动数据生成n组设计输入模型进行计算，计算出的电芯性能与此批次电池实际下仓后测试的性能进行对比，分析模型各个参数对结果影响程度及影响力的准确性。In the above embodiment, the test battery can be obtained specifically in the following two ways: one is to artificially manufacture batteries of different designs, and specifically test performance such as capacity/DCR, etc., and simulate the performance of different designs respectively, compare them with actual measurements, and analyze the accuracy of the influence of each parameter of the model on the results; the other is to directly select from the batteries in the upper and lower warehouses of the production line. Specifically, by taking the incoming material/design/process data of the battery that has been manufactured and tested in the lower warehouse, n groups of design input models are generated according to the fluctuation data for calculation, and the calculated battery cell performance is compared with the performance of this batch of batteries after actual storage, and the accuracy of the influence of each parameter of the model on the results and the influence of the model are analyzed.
本申请实施例的技术方案中，对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据，识别出产生电池性能差异对应的波动工艺参数，并获取该指标对应的修正因子，以该修正因子调整初始P2D电化学-热耦合模型，得到更加准确的P2D电化学-热耦合模型。In the technical solution of the embodiment of the present application, the test battery performance is compared with the initial battery performance prediction results to obtain battery performance difference data, identify the fluctuating process parameters corresponding to the battery performance difference, and obtain the correction factor corresponding to the indicator. The initial P2D electrochemical-thermal coupling model is adjusted with the correction factor to obtain a more accurate P2D electrochemical-thermal coupling model.
在一些实施例中，根据样本电池的工艺参数与电池性能对应关系训练机器学习模型，得到电池修正值预测模型包括：In some embodiments, training a machine learning model based on the corresponding relationship between the process parameters of a sample battery and the battery performance to obtain a battery correction value prediction model includes:
获取初始高斯过程回归模型；采用样本电池的工艺参数与电池性能对应关系训练初始高斯过程回归模型，得到电池修正值预测模型。An initial Gaussian process regression model is obtained; the initial Gaussian process regression model is trained using the corresponding relationship between the process parameters of the sample battery and the battery performance to obtain a battery correction value prediction model.
高斯过程回归(Gaussian Process Regression,GPR)是使用高斯过程(Gaussian Process,GP)先验对数据进行回归分析的非参数模型(non-parameteric model)。GPR的模型假设包括噪声(回归残差)和高斯过程先验两部分，其求解按贝叶斯推断(Bayesian inference)进行。若不限制核函数的形式，GPR在理论上是紧致空间(compact space)内任意连续函数的通用近似(universal approximator)。基于高斯过程及其核函数所具有的便利性质，GPR在时间序列分析、图像处理和自动控制等领域的问题中有得到应用。GPR是计算开销较大的算法，通常被用于低维和小样本的回归问题，但也有适用于大样本和高维情形的扩展算法。Gaussian Process Regression (GPR) is a non-parameteric model that uses Gaussian Process (GP) priors to perform regression analysis on data. The model assumptions of GPR include noise (regression residual) and Gaussian process priors, and its solution is based on Bayesian inference. If the form of the kernel function is not restricted, GPR is theoretically a universal approximator for any continuous function in a compact space. Based on the convenient properties of Gaussian processes and their kernel functions, GPR has been applied to problems in the fields of time series analysis, image processing, and automatic control. GPR is an algorithm with high computational overhead and is usually used for regression problems with low dimensions and small samples, but there are also extended algorithms suitable for large samples and high dimensions.
以样本电池的工艺参数与电池性能对应关系对初始高斯过程回归模型进行训练，在训练过程中可以将一部分数据作为训练集，另一部分数据作 为测试集，先采用训练集对初始高斯过程回归模型进行训练，得到训练后的模型，再采用测试集对训练后的模型进行测试，验证训练得到的模型是否合格，若不合格，则继续保持对模型的迭代训练与优化，直至最终合格。进一步来说，训练集可以采用较多的数据组成，测试集可以采用相对较少的数据组建，训练集和测试集的数据量比例可以为8:2，以实现训练和测试的兼顾。初始高斯过程回归模型具体可以为ARD的高斯过程回归模型，基于关联规则的高斯过程回归模型能够更进一步准确挖掘工艺参数与电池性能之间的关联关系，即最终能够得到更加准确的电池性能分布结果。The initial Gaussian process regression model is trained based on the correspondence between the process parameters and battery performance of the sample battery. During the training process, part of the data can be used as a training set, and the other part of the data can be used as a test set. The initial Gaussian process regression model is first trained with the training set to obtain the trained model, and then the trained model is tested with the test set to verify whether the trained model is qualified. If it is unqualified, the iterative training and optimization of the model are continued until it is finally qualified. Further, the training set can be composed of more data, and the test set can be composed of relatively less data. The data volume ratio of the training set and the test set can be 8:2 to achieve a balance between training and testing. The initial Gaussian process regression model can specifically be the Gaussian process regression model of ARD. The Gaussian process regression model based on association rules can further accurately explore the correlation between process parameters and battery performance, that is, it can ultimately obtain a more accurate battery performance distribution result.
本申请实施例的技术方案中，以高斯过程回归模型作为基础模型，采用样本电池的工艺参数与电池性能对应关系训练初始高斯过程回归模型，得到电池修正值预测模型，由于高斯过程回归模型能够实现平稳随机场的非参数回归，最终得到的电池修正值预测模型能够支持稳定、准确的电池性能预测。In the technical solution of the embodiment of the present application, a Gaussian process regression model is used as the basic model, and the initial Gaussian process regression model is trained using the correspondence between the process parameters of the sample battery and the battery performance to obtain a battery correction value prediction model. Since the Gaussian process regression model can realize non-parametric regression of a stationary random field, the battery correction value prediction model finally obtained can support stable and accurate battery performance prediction.
在一些实施例中，获取工艺参数对应的制造工艺对电池的电池性能的影响修正值包括：In some embodiments, obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery includes:
采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。The P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
本申请实施例的技术方案中，采用P2D电化学-热耦合模型来直接对工艺参数进行处理，由于P2D电化学-热耦合模型是一种性能稳定、且电池性能预测准确的模型，其能够准确得到工艺参数对所述电池的电池性能的影响修正值。In the technical solution of the embodiment of the present application, the P2D electrochemical-thermal coupling model is used to directly process the process parameters. Since the P2D electrochemical-thermal coupling model is a model with stable performance and accurate battery performance prediction, it can accurately obtain the correction value of the impact of the process parameters on the battery performance of the battery.
在一些实施例中，获取电池制造的工艺参数包括：In some embodiments, obtaining process parameters for battery manufacturing includes:
获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
粒径是指来料中颗粒物质的直径；涂布厚度是指电芯生产中涂布的厚度；极片尺寸是指卷绕电芯中极片的尺寸，其具体包括极片的宽度等；孔隙率是指来料中块状材料的孔隙体积与材料在自然状态下总体积的百分比；压实密度是指电池生产中面密度与材料的厚度比值，在电芯生产中压实密度对电池性能有较大的影响，压实密度不光和颗粒的大小、密度有关 系，还和粒子的级配有关系，压实密度大的一般都有很好的粒子正态分布，可以认为，工艺条件一定的条件下，压实密度越大，电池的容量越高；面密度是指工程材料领域中定厚度的物质单位面积的质量。Particle size refers to the diameter of the particulate matter in the incoming material; coating thickness refers to the thickness of the coating in the production of battery cells; pole piece size refers to the size of the pole piece in the wound battery cell, which specifically includes the width of the pole piece, etc.; porosity refers to the percentage of the pore volume of the block material in the incoming material to the total volume of the material in the natural state; compaction density refers to the ratio of surface density to material thickness in battery production. In battery cell production, compaction density has a great influence on battery performance. Compaction density is not only related to the size and density of the particles, but also to the grading of the particles. Generally, particles with large compaction density have a good normal distribution. It can be considered that under certain process conditions, the greater the compaction density, the higher the capacity of the battery; surface density refers to the mass per unit area of a material of a certain thickness in the field of engineering materials.
本申请实施例的技术方案中，本申请实施例的技术方案中，工艺参数包括粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度多个维度的指标，分析不同工艺参数对电池性能的影响。In the technical solution of the embodiment of the present application, in the technical solution of the embodiment of the present application, the process parameters include indicators of multiple dimensions such as particle size, coating thickness, pole piece size, porosity, compaction density and surface density, and the influence of different process parameters on battery performance is analyzed.
在上述电池性能预测方法的基础上，本申请发明人还进一步意识到，可以针对批次电池进行性能预测，得到批次电池的性能预测结果，再将批次电池的性能预测结果进行进一步的分布分析，最终得到电池性能的分布结果，设计人员可以基于电池性能的分布分析结果直观了解到批次电池的性能分布情况，有助于设计人员进行电池设计发，缩短电池研发周期。另外，在获取批次电池的性能过程中，若还是采取简单直接测试的方式显然需要耗费大量的时间，并且影响电池性能变化的参数多样、且不同参数之间可能还存在相互影响和相互作用的可能性，因此采取直接测量的方式针对批次电池性能研究同样还还会存在不准确的缺陷。On the basis of the above-mentioned battery performance prediction method, the inventor of the present application further realized that the performance prediction can be performed on batch batteries to obtain the performance prediction results of batch batteries, and then the performance prediction results of batch batteries can be further distributed analyzed to finally obtain the distribution results of battery performance. Designers can intuitively understand the performance distribution of batch batteries based on the distribution analysis results of battery performance, which is helpful for designers to design batteries and shorten the battery research and development cycle. In addition, in the process of obtaining the performance of batch batteries, if a simple and direct test method is still adopted, it will obviously take a lot of time, and the parameters that affect the change of battery performance are diverse, and there may be the possibility of mutual influence and interaction between different parameters. Therefore, the direct measurement method for batch battery performance research will also have the defect of inaccuracy.
电池性能分布预测方案中，获取电池制造的若干组工艺参数，采用上述的电池性能预测方法进行处理，得到若干组电池的修正电池性能，以若干组电池的修正电池性能分析电池性能分布结果。整个方案中，由于采用上述的电池性能预测方法进行电池性能预测，其能够得到准确的若干组电池的修正电池性能，因此，最终能够准确预测电池的电池性能分布结果。In the battery performance distribution prediction scheme, several groups of process parameters for battery manufacturing are obtained, and the above-mentioned battery performance prediction method is used for processing to obtain the corrected battery performance of several groups of batteries, and the battery performance distribution results are analyzed with the corrected battery performance of several groups of batteries. In the whole scheme, since the above-mentioned battery performance prediction method is used for battery performance prediction, it is possible to obtain accurate corrected battery performance of several groups of batteries, and therefore, the battery performance distribution results of the battery can be accurately predicted in the end.
另外，本申请提供的电池性能分布预测方案同样可以应用于图1所示的应用场景中，终端102发送电池性能分布预测请求至服务器104，服务器104响应该电池性能分布预测请求，获取电池制造的若干组工艺参数；获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；根据若干影响修正值，得到若干组电池的电池性能；基于若干组电池的电池性能，生成电池性能分布预测结果。进一步的，服务器104将电池性能分布结果发送至终端102，终端102显示电池性能分布结果，这样研发人员可以直观了解到批次电池的电池性能，发现电池的性能与电化学参数之间关联关系，有利于研发人员缩短电池的研发 周期。In addition, the battery performance distribution prediction scheme provided by the present application can also be applied to the application scenario shown in FIG. 1. The terminal 102 sends a battery performance distribution prediction request to the server 104. The server 104 responds to the battery performance distribution prediction request and obtains several groups of process parameters for battery manufacturing; obtains several groups of process parameters corresponding to the manufacturing process and correction values of the battery performance of the battery, and obtains several groups of correction values; obtains the battery performance of several groups of batteries according to the correction values; and generates the battery performance distribution prediction results based on the battery performance of several groups of batteries. Further, the server 104 sends the battery performance distribution results to the terminal 102, and the terminal 102 displays the battery performance distribution results, so that the R&D personnel can intuitively understand the battery performance of the batch battery and find the correlation between the battery performance and the electrochemical parameters, which is conducive to shortening the battery R&D cycle.
可以理解的是，在上述电池性能分布预测方案可以直接应用于终端，即由终端来独自完成性能分布预测，其具体的处理过程与上述内容类似，在此不再赘述。It can be understood that the above-mentioned battery performance distribution prediction solution can be directly applied to the terminal, that is, the terminal completes the performance distribution prediction alone, and its specific processing process is similar to the above-mentioned content, which will not be repeated here.
如图5所示，本申请提供一种电池性能分布预测方法，方法包括：As shown in FIG5 , the present application provides a method for predicting battery performance distribution, the method comprising:
S820：获取电池制造的若干组工艺参数。S820: Acquire several groups of process parameters for battery manufacturing.
S840：获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；S840: Obtaining correction values of the impact of the manufacturing process on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of impact correction values;
S860：根据若干影响修正值，得到若干组电池的电池性能；S860: Obtaining battery performance of a plurality of battery groups according to a plurality of impact correction values;
S880：基于若干组电池的电池性能，生成电池性能分布预测结果。S880: Generate a battery performance distribution prediction result based on the battery performance of several groups of batteries.
在本申请实施例的技术方案中，先获取电池制造的若干组工艺参数，针对每组工艺参数采取上述电池性能方法来得到若干组电池的电池性能，基于若干组电池的电池性能，生成电池性能分布预测结果。具体来说，在这里获取电池制造的若干组制造工艺的工艺参数，采用上述的电池性能预测方法对这些若干组的工艺参数分别进行处理，得到若干组电池性能预测值，分析这些若干组电池性能预测值，得到电池性能分布结果。这里的若干组是相对数量较多组，例如可以为1万组工艺参数、10万组工艺参数，通过大量预测不同工艺参数对应的电池性能预测值可以得到更加准确的相同型号/批次电池对应的电池性能分布情况，以给电池设计人员提供更加准确、完整的数据支持。In the technical solution of the embodiment of the present application, several groups of process parameters for battery manufacturing are first obtained, and the above-mentioned battery performance method is adopted for each group of process parameters to obtain the battery performance of several groups of batteries, and the battery performance distribution prediction results are generated based on the battery performance of several groups of batteries. Specifically, several groups of process parameters of the manufacturing process of battery manufacturing are obtained here, and the above-mentioned battery performance prediction method is adopted to process these several groups of process parameters respectively to obtain several groups of battery performance prediction values, and these several groups of battery performance prediction values are analyzed to obtain battery performance distribution results. The several groups here are relatively large groups, for example, 10,000 groups of process parameters and 100,000 groups of process parameters. By predicting a large number of battery performance prediction values corresponding to different process parameters, a more accurate battery performance distribution corresponding to the same model/batch of batteries can be obtained, so as to provide battery designers with more accurate and complete data support.
本申请实施例的技术方案中，获取电池制造的若干组工艺参数，针对该若干组工艺参数分别获取对应的制造工艺对电池的电池性能的影响修正值，再基于若干组影响修正值，得到若干组电池的电池性能，分析这些电池的电池性能，生成若干组电池对应的电池性能分布预测结果。整个方案中，针对电池的若干组工艺参数，分别考虑工艺参数对应的制造工艺对电池性能影响，能够准确预测若干组电池的电池性能，因此，可以实现准确的电池性能分布预测。In the technical solution of the embodiment of the present application, several groups of process parameters for battery manufacturing are obtained, and the corresponding manufacturing process's influence correction values on the battery performance of the battery are obtained for the several groups of process parameters, and then the battery performance of several groups of batteries is obtained based on the several groups of influence correction values, and the battery performance of these batteries is analyzed to generate the battery performance distribution prediction results corresponding to the several groups of batteries. In the whole solution, for several groups of process parameters of the battery, the influence of the manufacturing process corresponding to the process parameters on the battery performance is considered respectively, and the battery performance of several groups of batteries can be accurately predicted, so that accurate battery performance distribution prediction can be achieved.
在一些实施例中，如图6所示，S820包括：In some embodiments, as shown in FIG6 , S820 includes:
S822：获取电池制造的工艺参数波动数据；S822: Obtaining process parameter fluctuation data of battery manufacturing;
S824：根据工艺参数波动数据，生成电池制造的若干组工艺参数。S824: Generate several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
电池制造的工艺波动数据是指电池制造工艺过程中对应的工艺波动数据，其具体可以通过电池制造过程中来料的相关波动、电池制造公司内部的来料检测***、或根据电池制造来料(原料)供应商提供的来料波动数据获取，这里的工艺参数波动数据可以理解为一个波动的范围值，以涂布厚度为例，其可以在(a，b)范围内波动。根据工艺参数波动数据，从中组合生成若干组制造工艺的工艺参数。例如可以采用随机抽取、组合的方式来生成若干组制造工艺的工艺参数，以模拟在真实电池制造中多样化的工艺参数变动。The process fluctuation data of battery manufacturing refers to the corresponding process fluctuation data in the battery manufacturing process, which can be obtained through the relevant fluctuations of incoming materials in the battery manufacturing process, the incoming material detection system within the battery manufacturing company, or the incoming material fluctuation data provided by the battery manufacturing incoming material (raw material) supplier. The process parameter fluctuation data here can be understood as a fluctuation range value. Taking the coating thickness as an example, it can fluctuate within the range of (a, b). According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are combined and generated. For example, several groups of process parameters of the manufacturing process can be generated by random extraction and combination to simulate the diversified process parameter changes in real battery manufacturing.
本申请实施例的技术方案中，先获取电池制造的工艺参数波动数据，根据电池制造的工艺参数波动数据来生成若干组工艺参数，能够充分仿真在实际电池制造中可能对应的工艺参数，从而使后续得到的电池性能分布预测结果更加符合真实情况。In the technical solution of the embodiment of the present application, the process parameter fluctuation data of battery manufacturing is first obtained, and several groups of process parameters are generated according to the process parameter fluctuation data of battery manufacturing. This can fully simulate the process parameters that may correspond in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
在一些实施例中，根据工艺参数波动数据，生成电池制造的若干组工艺参数，包括：In some embodiments, several sets of process parameters for battery manufacturing are generated based on the process parameter fluctuation data, including:
根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
在这里采用随机的方式生成若干组制造工艺的工艺参数。例如可以采用随机生成算法生成若干组制造工艺的工艺参数，基于这些工艺参数生成虚拟电池，再采用上述电池性能预测方法得到这些虚拟电池对应的电池性能，产生电池性能分布结果。电池性能分布结果可以是电池性能的正态分布，其能够准确表征随着工艺参数的波动对应电池性能的分布情况。在实际应用中，随机生成算法包括蒙特卡洛算法。蒙特·卡罗方法(Monte Carlo method)，也称统计模拟方法，是二十世纪四十年代中期重要的数值计算方法。是指使用随机数(或更常见的伪随机数)来解决很由于科学技术的发展和电子计算机的发明，而被提出的一种以概率统计理论为指导的一类非常多计算问题的方法。本申请实施例的技术方案中，随机生成算法选用蒙特卡洛算法，其可以实现准确的数据随机生成处理，支持后续得到准确的电池性能分布。Here, several groups of process parameters of the manufacturing process are generated in a random manner. For example, a random generation algorithm can be used to generate several groups of process parameters of the manufacturing process, virtual batteries can be generated based on these process parameters, and then the battery performance corresponding to these virtual batteries can be obtained by the above-mentioned battery performance prediction method to generate battery performance distribution results. The battery performance distribution result can be a normal distribution of battery performance, which can accurately characterize the distribution of battery performance corresponding to the fluctuation of process parameters. In practical applications, random generation algorithms include Monte Carlo algorithms. Monte Carlo method, also known as statistical simulation method, is an important numerical calculation method in the mid-1940s. It refers to the use of random numbers (or more common pseudo-random numbers) to solve a class of very many computational problems guided by probability statistics theory, which is proposed due to the development of science and technology and the invention of electronic computers. In the technical solution of the embodiment of the present application, the random generation algorithm uses the Monte Carlo algorithm, which can realize accurate data random generation processing and support the subsequent accurate battery performance distribution.
本申请实施例的技术方案中，采用随机生成的方式来生成若干组工艺参数，随机生成的方式一方面丰富了工艺参数的数量；另一方面，通过随机生成的方式使得生成的工艺参数更加贴近实际电池制造中对应工艺参数 的真实情况，从而使后续得到的电池性能分布预测结果更加符合真实情况。In the technical solution of the embodiment of the present application, a random generation method is adopted to generate several groups of process parameters. On the one hand, the random generation method enriches the number of process parameters; on the other hand, the random generation method makes the generated process parameters closer to the actual situation of the corresponding process parameters in actual battery manufacturing, so that the subsequent battery performance distribution prediction results are more in line with the actual situation.
为详细说明本申请的技术方案及其效果，下面将采用具体应用实例展开描述，下面将采用具体应用实例展开描述，在具体应用实例中，如图7所示，整个方法包括以下步骤：In order to explain the technical solution and its effect in detail, a specific application example will be used to describe the invention. In the specific application example, as shown in FIG7 , the whole method includes the following steps:
1、统计电池的工艺参数，得到工艺参数的公差分布；1. Count the process parameters of the battery and obtain the tolerance distribution of the process parameters;
2、将工艺参数输入至P2D电化学-热耦合模型中，得到工艺参数-电化学参数-性能对应关系表；2. Input the process parameters into the P2D electrochemical-thermal coupling model to obtain the process parameter-electrochemical parameter-performance correspondence table;
3、将得到的工艺参数-电化学参数-性能对应关系表作为训练数据，训练机器学习模型，并且以部分参数作为测试集测试训练后的机器学习模型，待确认测试合格时，得到能够预测电池性能的机器学习模型；3. Use the obtained process parameter-electrochemical parameter-performance correspondence table as training data to train the machine learning model, and use some parameters as the test set to test the trained machine learning model. When the test is confirmed to be qualified, a machine learning model that can predict battery performance is obtained;
4、将批量工艺参数输入至机器学习模型中，得到大量电池性能数据，归集整理成工艺参数分布-性能对应关系表；4. Input batch process parameters into the machine learning model to obtain a large amount of battery performance data, which is then compiled into a process parameter distribution-performance correspondence table;
5、采用蒙特卡洛算法随机生成各种工艺参数，并产生对应的虚拟电池，生成电池性能分布；5. Use the Monte Carlo algorithm to randomly generate various process parameters, generate corresponding virtual batteries, and generate battery performance distribution;
6、将电池性能分布发送至研发人员，以给研发人员设计\研发电池提供有力的数据支持。6. Send the battery performance distribution to R&D personnel to provide strong data support for R&D personnel to design and develop batteries.
应该理解的是，虽然上述各流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，这些步骤可以以其它的顺序执行。而且，上述各流程图中的至少一部分步骤可以包括多个步骤或者多个阶段，这些步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，这些步骤或者阶段的执行顺序也不必然是依次进行，而是可以与其它步骤或者其它步骤中的步骤或者阶段的至少一部分轮流或者交替地执行。It should be understood that, although the steps in the above-mentioned flowcharts are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the above-mentioned flowcharts may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
如图8所示，本申请还提供一种电池性能预测装置，装置包括：As shown in FIG8 , the present application also provides a battery performance prediction device, the device comprising:
工艺参数获取模块200，用于获取电池制造的工艺参数；A process parameter acquisition module 200 is used to acquire process parameters for battery manufacturing;
处理模块400，用于获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；The processing module 400 is used to obtain a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
性能预测模块600，用于根据影响修正值，得到电池的电池性能。The performance prediction module 600 is used to obtain the battery performance of the battery according to the impact correction value.
本申请电池性能预测装置，先获取电池制造的工艺参数，采用电池性能预测值预测模型对工艺参数进行处理，得到处理结果，再在电池设计参数对应的基准电池性能基础上考虑制造工艺对电池性能的影响，得到电池的电池性能。整个方案中，一方面考虑制造工艺对电池性能的影响；另一方面，采用基于电池性能预测值预测模型准确分析工艺参数对电池性能的影响，通过模型充分挖掘不同工艺参数对电池性能的影响；因此，整个方案可以得到准确的电池性能预测结果。The battery performance prediction device of the present application first obtains the process parameters of battery manufacturing, uses the battery performance prediction value prediction model to process the process parameters, obtains the processing results, and then considers the impact of the manufacturing process on the battery performance based on the benchmark battery performance corresponding to the battery design parameters to obtain the battery performance of the battery. In the whole scheme, on the one hand, the impact of the manufacturing process on the battery performance is considered; on the other hand, the impact of the process parameters on the battery performance is accurately analyzed by using the prediction model based on the battery performance prediction value, and the impact of different process parameters on the battery performance is fully explored through the model; therefore, the whole scheme can obtain accurate battery performance prediction results.
在一些实施例中，处理模块400还用于采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺参数对应的制造工艺对电池的电池性能的影响修正值。In some embodiments, the processing module 400 is further used to process the process parameters using the trained battery correction value prediction model to obtain the correction value of the impact of the manufacturing process corresponding to the process parameter on the battery performance of the battery.
在一些实施例中，电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。In some embodiments, the battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction values.
在一些实施例中，处理模块400还用于获取若干个样本电池的样本工艺参数；根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系；根据工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。In some embodiments, the processing module 400 is also used to obtain sample process parameters of several sample batteries; based on the sample process parameters, obtain the battery performance of several sample batteries, and generate a correspondence between the process parameters and the battery performance; train a machine learning model based on the correspondence between the process parameters and the battery performance to obtain a battery correction value prediction model.
在一些实施例中，处理模块400还用于采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。In some embodiments, the processing module 400 is also used to use the P2D electrochemical-thermal coupling model to process several sample process parameters respectively to obtain the battery performance of each sample battery; associate the sample process parameters and battery performance to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
在一些实施例中，处理模块400还用于获取测量得到的测试工艺参数对应的测试电池性能；通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。In some embodiments, the processing module 400 is also used to obtain the test battery performance corresponding to the measured test process parameters; process the test process parameters through the initial P2D electrochemical-thermal coupling model to obtain the initial battery performance prediction results corresponding to the test process parameters; optimize the initial P2D electrochemical-thermal coupling model according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model.
在一些实施例中，处理模块400还用于对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；识别电池性能差异数据对应的波动工艺参数；获取波动工艺参数对应的修正因子；根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。In some embodiments, the processing module 400 is also used to compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
在一些实施例中，处理模块400还用于采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。In some embodiments, the processing module 400 is further used to process the process parameters using a P2D electrochemical-thermal coupling model to obtain a correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，工艺参数获取模块200还用于获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。In some embodiments, the process parameter acquisition module 200 is also used to obtain at least one process parameter of the battery manufacturing process including particle size, coating thickness, pole piece size, porosity, compaction density and surface density.
如图9所示，本申请还提供一种电池性能分布预测装置，装置包括：As shown in FIG9 , the present application also provides a battery performance distribution prediction device, the device comprising:
若干组工艺参数获取模块820，用于获取电池制造的若干组工艺参数；Several groups of process parameter acquisition modules 820, used to acquire several groups of process parameters for battery manufacturing;
修正模块840，用于获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；A correction module 840 is used to obtain correction values of the effects of the manufacturing process on the battery performance of the battery corresponding to several groups of process parameters, and obtain several groups of effect correction values;
电池性能获取模块860，用于根据若干影响修正值，得到若干组电池的电池性能；A battery performance acquisition module 860, for obtaining battery performance of a plurality of battery groups according to a plurality of impact correction values;
电池性能分布处理模块880，用于基于若干组电池的电池性能，生成电池性能分布预测结果。The battery performance distribution processing module 880 is used to generate a battery performance distribution prediction result based on the battery performance of several groups of batteries.
上述电池性能分布预测装置，获取电池制造的若干组工艺参数，采用上述的电池性能预测方法进行处理，得到若干组电池的修正电池性能，以若干组电池的修正电池性能分析电池性能分布结果。整个方案中，由于采用上述的电池性能预测方法进行电池性能预测，其能够得到准确的若干组电池的修正电池性能，因此，最终能够准确预测电池的电池性能分布结果。The battery performance distribution prediction device obtains several groups of process parameters for battery manufacturing, processes them using the battery performance prediction method, obtains the corrected battery performance of several groups of batteries, and analyzes the battery performance distribution results using the corrected battery performance of several groups of batteries. In the whole scheme, since the battery performance prediction method is used to predict the battery performance, it can obtain accurate corrected battery performance of several groups of batteries, and therefore, it can finally accurately predict the battery performance distribution results of the battery.
在一些实施例中，若干组工艺参数获取模块820还用于获取电池制造的工艺参数波动数据；根据工艺参数波动数据，生成电池制造的若干组工艺参数。In some embodiments, the several groups of process parameter acquisition modules 820 are also used to acquire process parameter fluctuation data for battery manufacturing; and generate several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
在一些实施例中，若干组工艺参数获取模块820还用于根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。In some embodiments, the several groups of process parameter acquisition modules 820 are further used to randomly generate several groups of process parameters of the manufacturing process according to the process parameter fluctuation data.
关于电池性能预测装置/电池性能分布预测的具体实施例可以参见上文中对于电池性能预测方法/电池性能分布预测方法的实施例，在此不再赘述。上述电池性能预测装置中的各个模块可全部或部分通过软件、硬件及 其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。For the specific embodiments of the battery performance prediction device/battery performance distribution prediction, please refer to the embodiments of the battery performance prediction method/battery performance distribution prediction method mentioned above, which will not be repeated here. Each module in the above-mentioned battery performance prediction device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
在一些实施例中，提供了一种计算机设备，该计算机设备可以是服务器，其内部结构图可以如图10所示。该计算机设备包括通过***总线连接的处理器、存储器和网络接口。其中，该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作***、计算机程序和数据库。该内存储器为非易失性存储介质中的操作***和计算机程序的运行提供环境。该计算机设备的数据库用于存储已训练模型相关的数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机程序被处理器执行时以实现一种电池性能预测方法。In some embodiments, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in FIG10. The computer device includes a processor, a memory, and a network interface connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data related to the trained model. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a battery performance prediction method is implemented.
本领域技术人员可以理解，图10中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。Those skilled in the art will understand that the structure shown in FIG. 10 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.
在一些实施例中，提供了一种计算机设备，包括存储器和处理器，存储器中存储有计算机程序，该处理器执行计算机程序时实现以下步骤：In some embodiments, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺参数对应的制造工艺对电池的电池性能的影响修正值。The trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：：In some embodiments, when the processor executes the computer program, the following steps are also implemented:
获取若干个样本电池的样本工艺参数；根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系；根据工艺 参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。Obtain sample process parameters of several sample batteries; obtain battery performance of several sample batteries according to the sample process parameters, and generate a corresponding relationship between the process parameters and the battery performance; train a machine learning model according to the corresponding relationship between the process parameters and the battery performance to obtain a battery correction value prediction model.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：：In some embodiments, when the processor executes the computer program, the following steps are also implemented:
采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。The P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
获取测量得到的测试工艺参数对应的测试电池性能；通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；识别电池性能差异数据对应的波动工艺参数；获取波动工艺参数对应的修正因子；根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。Compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。The P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
在一些实施例中，提供了一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现以下步骤：In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺 参数对应的制造工艺对电池的电池性能的影响修正值。The trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取若干个样本电池的样本工艺参数；根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系；根据工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。Obtain sample process parameters of several sample batteries; obtain battery performance of several sample batteries based on the sample process parameters, and generate a corresponding relationship between the process parameters and the battery performance; train a machine learning model based on the corresponding relationship between the process parameters and the battery performance to obtain a battery correction value prediction model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。The P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取测量得到的测试工艺参数对应的测试电池性能；通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；识别电池性能差异数据对应的波动工艺参数；获取波动工艺参数对应的修正因子；根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。Compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。The P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
在一些实施例中，提供了一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现以下步骤：In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
获取工艺参数对应的制造工艺对电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
根据影响修正值，得到电池的电池性能。According to the impact correction value, the battery performance of the battery is obtained.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用训练得到的电池修正值预测模型对工艺参数进行处理，得到工艺参数对应的制造工艺对电池的电池性能的影响修正值。The trained battery correction value prediction model is used to process the process parameters to obtain the correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is trained based on the correspondence between the process parameters of battery manufacturing and the battery performance prediction value.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取若干个样本电池的样本工艺参数；根据样本工艺参数，获取若干个样本电池的电池性能，生成工艺参数与电池性能的对应关系；根据工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。Obtain sample process parameters of several sample batteries; obtain battery performance of several sample batteries based on the sample process parameters, and generate a corresponding relationship between the process parameters and the battery performance; train a machine learning model based on the corresponding relationship between the process parameters and the battery performance to obtain a battery correction value prediction model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用P2D电化学-热耦合模型分别对若干个样本工艺参数进行处理，获得各样本电池的电池性能；关联样本工艺参数以及电池性能，得到样本电池的工艺参数与电池性能的对应关系。The P2D electrochemical-thermal coupling model is used to process several sample process parameters respectively to obtain the battery performance of each sample battery; the sample process parameters and battery performance are correlated to obtain the corresponding relationship between the process parameters and battery performance of the sample battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取测量得到的测试工艺参数对应的测试电池性能；通过初始P2D电化学-热耦合模型对测试工艺参数进行处理，获得测试工艺参数对应的初始电池性能预测结果；根据测试电池性能以及初始电池性能预测结果对初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The test battery performance corresponding to the measured test process parameters is obtained; the test process parameters are processed by an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters; the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain a P2D electrochemical-thermal coupling model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
对比测试电池性能与初始电池性能预测结果，得到电池性能差异数据；识别电池性能差异数据对应的波动工艺参数；获取波动工艺参数对应的修正因子；根据修正因子调整初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。Compare the test battery performance with the initial battery performance prediction results to obtain battery performance difference data; identify the fluctuating process parameters corresponding to the battery performance difference data; obtain the correction factors corresponding to the fluctuating process parameters; adjust the initial P2D electrochemical-thermal coupling model according to the correction factors to obtain the P2D electrochemical-thermal coupling model.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
采用P2D电化学-热耦合模型对工艺参数进行处理，得到工艺参数对应的工艺参数对应的制造工艺对电池的电池性能的影响修正值。The P2D electrochemical-thermal coupling model is used to process the process parameters to obtain correction values of the effects of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
在一些实施例中，提供了一种计算机设备，包括存储器和处理器，存储器中存储有计算机程序，该处理器执行计算机程序时实现以下步骤：In some embodiments, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
获取电池制造的工艺参数波动数据；根据工艺参数波动数据，生成电池制造的若干组工艺参数。Obtaining process parameter fluctuation data for battery manufacturing; generating several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
在一些实施例中，处理器执行计算机程序时还实现以下步骤：In some embodiments, when the processor executes the computer program, the processor further implements the following steps:
根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
在一些实施例中，提供了一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行时实现以下步骤：In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取电池制造的工艺参数波动数据；根据工艺参数波动数据，生成电池制造的若干组工艺参数。Obtaining process parameter fluctuation data for battery manufacturing; generating several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
在一些实施例中，提供了一种计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现以下步骤：In some embodiments, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:
获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
获取若干组工艺参数对应的制造工艺对电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, and obtaining several groups of effect correction values;
根据若干影响修正值，得到若干组电池的电池性能；According to a number of impact correction values, battery performance of a number of battery groups is obtained;
基于若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of several groups of batteries, a battery performance distribution prediction result is generated.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
获取电池制造的工艺参数波动数据；根据工艺参数波动数据，生成电池制造的若干组工艺参数。Obtaining process parameter fluctuation data for battery manufacturing; generating several groups of process parameters for battery manufacturing according to the process parameter fluctuation data.
在一些实施例中，计算机程序被处理器执行时还实现以下步骤：In some embodiments, when the computer program is executed by the processor, the following steps are also implemented:
根据工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，上述的计算机程序可存储于一非易失性计算机可读取存储介质中，该计算机程序在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和易失性存储器中的至少一种。非易失性存储器可包括只读存储器(Read-Only Memory，ROM)、磁带、软盘、闪存或光存储器等。易失性存储器可包括随机存取存储器(Random Access Memory，RAM)或外部高速缓冲存储器。作为说明而非局限，RAM可以是多种形式，比如静态随机存取存储器(Static Random Access Memory，SRAM)或动态随机存取存储器(Dynamic Random Access Memory，DRAM)等。A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program. The above-mentioned computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
最后应说明的是：以上各实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述各实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的范 围，其均应涵盖在本申请的权利要求和说明书的范围当中。尤其是，只要不存在结构冲突，各个实施例中所提到的各项技术特征均可以任意方式组合起来。本申请并不局限于文中公开的特定实施例，而是包括落入权利要求的范围内的所有技术方案。Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should all be included in the scope of the claims and specification of the present application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any way. The present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions that fall within the scope of the claims.

Claims (17)

  1. 一种电池性能预测方法，其特征在于，所述方法包括：A battery performance prediction method, characterized in that the method comprises:
    获取电池制造的工艺参数；Obtain process parameters for battery manufacturing;
    获取所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值；Obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
    根据所述影响修正值，得到电池的电池性能。The battery performance of the battery is obtained according to the influence correction value.
  2. 根据权利要求1所述的方法，其特征在于，所述获取所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值包括：The method according to claim 1, characterized in that the step of obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery comprises:
    采用训练得到的电池修正值预测模型对所述工艺参数进行处理，得到所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值。The process parameters are processed using the trained battery correction value prediction model to obtain a correction value of the impact of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  3. 根据权利要求2所述的方法，其特征在于，获得所述电池修正值预测模型的方式包括：The method according to claim 2, characterized in that the manner of obtaining the battery correction value prediction model comprises:
    所述电池性能预测值预测模型基于电池制造的工艺参数与电池性能预测值对应关系训练得到。The battery performance prediction value prediction model is obtained by training based on the corresponding relationship between the process parameters of battery manufacturing and the battery performance prediction value.
  4. 根据权利要求2所述的方法，其特征在于，获得所述电池修正值预测模型的方式包括：The method according to claim 2, characterized in that the manner of obtaining the battery correction value prediction model comprises:
    获取若干个样本电池的样本工艺参数；Obtaining sample process parameters of several sample batteries;
    根据所述样本工艺参数，获取若干个所述样本电池的电池性能，生成工艺参数与电池性能的对应关系；According to the sample process parameters, obtaining battery performance of a plurality of the sample batteries, and generating a corresponding relationship between the process parameters and the battery performance;
    根据所述工艺参数与电池性能的对应关系训练机器学习模型，得到电池修正值预测模型。The machine learning model is trained according to the corresponding relationship between the process parameters and the battery performance to obtain a battery correction value prediction model.
  5. 根据权利要求4所述的方法，其特征在于，所述根据所述样本工艺参数，获取若干个所述样本电池的电池性能，生成工艺参数与电池性能的对应关系包括：The method according to claim 4, characterized in that the step of obtaining the battery performance of a plurality of the sample batteries according to the sample process parameters and generating a corresponding relationship between the process parameters and the battery performance comprises:
    采用P2D电化学-热耦合模型分别对若干个所述样本工艺参数进行处理，获得各所述样本电池的电池性能；Using the P2D electrochemical-thermal coupling model to process several sample process parameters respectively, to obtain the battery performance of each sample battery;
    关联所述样本工艺参数以及所述电池性能，得到所述样本电池的工艺参数与电池性能的对应关系。The sample process parameters and the battery performance are associated to obtain a corresponding relationship between the process parameters and the battery performance of the sample battery.
  6. 根据权利要求5所述的方法，其特征在于，所述P2D电化学-热耦合模型的获得方式包括：The method according to claim 5, characterized in that the P2D electrochemical-thermal coupling model is obtained by:
    获取测量得到的测试工艺参数对应的测试电池性能；Obtaining the test battery performance corresponding to the measured test process parameters;
    通过初始P2D电化学-热耦合模型对所述测试工艺参数进行处理，获得所述测试工艺参数对应的初始电池性能预测结果；Processing the test process parameters through an initial P2D electrochemical-thermal coupling model to obtain initial battery performance prediction results corresponding to the test process parameters;
    根据测试电池性能以及初始电池性能预测结果对所述初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型。The initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction results to obtain the P2D electrochemical-thermal coupling model.
  7. 根据权利要求6所述的方法，其特征在于，所述根据所述测试电池性能以及所述初始电池性能预测结果对所述初始P2D电化学-热耦合模型进行优化，获得P2D电化学-热耦合模型，包括：The method according to claim 6, characterized in that the initial P2D electrochemical-thermal coupling model is optimized according to the test battery performance and the initial battery performance prediction result to obtain the P2D electrochemical-thermal coupling model, comprising:
    对比所述测试电池性能与所述初始电池性能预测结果，得到电池性能差异数据；Comparing the test battery performance with the initial battery performance prediction result to obtain battery performance difference data;
    识别所述电池性能差异数据对应的波动工艺参数；Identifying the fluctuating process parameters corresponding to the battery performance difference data;
    获取所述波动工艺参数对应的修正因子；Obtaining a correction factor corresponding to the fluctuating process parameter;
    根据所述修正因子调整所述初始P2D电化学-热耦合模型，得到P2D电化学-热耦合模型。The initial P2D electrochemical-thermal coupling model is adjusted according to the correction factor to obtain a P2D electrochemical-thermal coupling model.
  8. 根据权利要求1所述的方法，其特征在于，所述获取所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值包括：The method according to claim 1, characterized in that the step of obtaining a correction value of the effect of the manufacturing process corresponding to the process parameter on the battery performance of the battery comprises:
    采用P2D电化学-热耦合模型对所述工艺参数进行处理，得到所述工艺参数对应的所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值。The process parameters are processed by using a P2D electrochemical-thermal coupling model to obtain a correction value of the influence of the manufacturing process corresponding to the process parameters on the battery performance of the battery.
  9. 根据权利要求1所述的方法，其特征在于，所述获取电池制造的工艺参数包括：The method according to claim 1, characterized in that the obtaining of process parameters for battery manufacturing comprises:
    获取电池制造的粒径、涂布厚度、极片尺寸、孔隙率、压实密度以及面密度中的至少一种工艺参数。At least one process parameter of battery manufacturing including particle size, coating thickness, pole piece size, porosity, compaction density and surface density is obtained.
  10. 一种电池性能分布预测方法，其特征在于，所述方法包括：A method for predicting battery performance distribution, characterized in that the method comprises:
    获取电池制造的若干组工艺参数；Obtaining several sets of process parameters for battery manufacturing;
    获取若干组工艺参数对应的制造工艺对所述电池的电池性能的影响修正值，得到若干组影响修正值；Obtaining correction values of the effects of the manufacturing processes on the battery performance of the battery corresponding to several groups of process parameters, to obtain several groups of correction values;
    根据所述若干影响修正值，得到若干组电池的电池性能；Obtaining battery performance of a plurality of battery groups according to the plurality of impact correction values;
    基于所述若干组电池的电池性能，生成电池性能分布预测结果。Based on the battery performance of the plurality of battery groups, a battery performance distribution prediction result is generated.
  11. 根据权利要求10所述的方法，其特征在于，所述获取电池制造的 若干组工艺参数，包括：The method according to claim 10, characterized in that the step of obtaining a plurality of groups of process parameters for battery manufacturing comprises:
    获取电池制造的工艺参数波动数据；Obtain process parameter fluctuation data for battery manufacturing;
    根据所述工艺参数波动数据，生成电池制造的若干组工艺参数。Several groups of process parameters for battery manufacturing are generated according to the process parameter fluctuation data.
  12. 根据权利要求11所述的方法，其特征在于，所述根据所述工艺参数波动数据，生成电池制造的若干组工艺参数，包括：The method according to claim 11, characterized in that the step of generating a plurality of sets of process parameters for battery manufacturing according to the process parameter fluctuation data comprises:
    根据所述工艺参数波动数据，随机生成若干组制造工艺的工艺参数。According to the process parameter fluctuation data, several groups of process parameters of the manufacturing process are randomly generated.
  13. 一种电池性能预测装置，其特征在于，所述装置包括：A battery performance prediction device, characterized in that the device comprises:
    工艺参数获取模块，用于获取电池制造的工艺参数；A process parameter acquisition module is used to obtain process parameters for battery manufacturing;
    处理模块，用于获取所述工艺参数对应的制造工艺对所述电池的电池性能的影响修正值；A processing module, used for obtaining a correction value of the influence of the manufacturing process corresponding to the process parameter on the battery performance of the battery;
    性能预测模块，用于根据所述影响修正值，得到电池的电池性能。The performance prediction module is used to obtain the battery performance of the battery according to the impact correction value.
  14. 一种电池性能分布预测装置，其特征在于，所述装置包括：A battery performance distribution prediction device, characterized in that the device comprises:
    若干组工艺参数获取模块，用于获取电池制造的若干组工艺参数；Several groups of process parameter acquisition modules, used to acquire several groups of process parameters for battery manufacturing;
    修正模块，用于获取若干组工艺参数对应的制造工艺对所述电池的电池性能的影响修正值，得到若干组影响修正值；A correction module, used to obtain correction values of the effects of the manufacturing processes corresponding to several groups of process parameters on the battery performance of the battery, and obtain several groups of effect correction values;
    电池性能获取模块，用于根据所述若干影响修正值，得到若干组电池的电池性能；A battery performance acquisition module, used to obtain battery performance of several groups of batteries according to the several impact correction values;
    电池性能分布处理模块，用于基于所述若干组电池的电池性能，生成电池性能分布预测结果。The battery performance distribution processing module is used to generate a battery performance distribution prediction result based on the battery performance of the plurality of battery groups.
  15. 一种计算机设备，包括存储器和处理器，所述存储器存储有计算机程序，其特征在于，所述处理器执行所述计算机程序时实现权利要求1至12中任一项所述的方法的步骤。A computer device comprises a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of any one of the methods of claims 1 to 12 when executing the computer program.
  16. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现权利要求1至12中任一项所述的方法的步骤。A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 12 are implemented.
  17. 一种计算机程序产品，包括计算机程序，其特征在于，该计算机程序被处理器执行时实现权利要求1至12中任一项所述的方法的步骤。A computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 12 are implemented.
PCT/CN2022/125298 2022-10-14 2022-10-14 Battery performance prediction method, and battery performance distribution prediction method WO2024077587A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/125298 WO2024077587A1 (en) 2022-10-14 2022-10-14 Battery performance prediction method, and battery performance distribution prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/125298 WO2024077587A1 (en) 2022-10-14 2022-10-14 Battery performance prediction method, and battery performance distribution prediction method

Publications (1)

Publication Number Publication Date
WO2024077587A1 true WO2024077587A1 (en) 2024-04-18

Family

ID=90668475

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/125298 WO2024077587A1 (en) 2022-10-14 2022-10-14 Battery performance prediction method, and battery performance distribution prediction method

Country Status (1)

Country Link
WO (1) WO2024077587A1 (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108009397A (en) * 2017-12-01 2018-05-08 中南大学 Predict emulation mode, device and the equipment of lithium ion battery material chemical property
CN113102516A (en) * 2021-03-05 2021-07-13 东北大学 Hot continuous rolling strip steel head width prediction method integrating rolling mechanism and deep learning
CN113687250A (en) * 2021-08-18 2021-11-23 蜂巢能源科技有限公司 Battery cell capacity prediction method and device, electronic equipment and medium
CN114298091A (en) * 2021-12-13 2022-04-08 国网湖北省电力有限公司电力科学研究院 SF6Gas flow measuring value correction method, device, equipment and storage medium
EP4011834A1 (en) * 2020-12-10 2022-06-15 Basf Se Process for making an electrode active material

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108009397A (en) * 2017-12-01 2018-05-08 中南大学 Predict emulation mode, device and the equipment of lithium ion battery material chemical property
EP4011834A1 (en) * 2020-12-10 2022-06-15 Basf Se Process for making an electrode active material
CN113102516A (en) * 2021-03-05 2021-07-13 东北大学 Hot continuous rolling strip steel head width prediction method integrating rolling mechanism and deep learning
CN113687250A (en) * 2021-08-18 2021-11-23 蜂巢能源科技有限公司 Battery cell capacity prediction method and device, electronic equipment and medium
CN114298091A (en) * 2021-12-13 2022-04-08 国网湖北省电力有限公司电力科学研究院 SF6Gas flow measuring value correction method, device, equipment and storage medium

Similar Documents

Publication Publication Date Title
Barbosa et al. Using performance profiles to analyze the results of the 2006 CEC constrained optimization competition
Macdonald et al. Assessing uncertainty in building simulation
CN112182720B (en) Building energy consumption model evaluation method based on building energy management application scene
CN109891508A (en) Single cell type detection method, device, equipment and storage medium
CN111080155B (en) Air conditioner user frequency modulation capability evaluation method based on generation countermeasure network
Zou et al. The green innovative power of carbon neutrality in China: A perspective of innovation efficiency in China’s high-tech industry based on meta-frontier DEA
JP7411977B2 (en) Machine learning support method and machine learning support device
CN117174899B (en) Preparation method of carbon fluoride anode material
CN114048525A (en) Design system of energy storage battery system and preparation of energy storage
WO2024077587A1 (en) Battery performance prediction method, and battery performance distribution prediction method
CN109635446A (en) It is fitted the method and its application of the open circuit potential curve of combination electrode
CN116562171B (en) Error assessment method for online measurement of temperature and humidity
Xu et al. A hybrid method for lithium-ion batteries state-of-charge estimation based on gated recurrent unit neural network and an adaptive unscented Kalman filter
Liu et al. Surrogate-assisted many-objective optimization of building energy management
CN103793582A (en) Optimization method for cylindrical shell large opening connecting pipe structure
Han et al. Capacity Consistency Prediction and Process Parameter Optimization of Lithium‐Ion Battery based on Neural Network and Particle Swarm Optimization Algorithm
Xie et al. A new solution to the spherical particle surface concentration of lithium-ion battery electrodes
Chen et al. Study on Model Evolution Method Based on the Hybrid Modeling Technology With Support Vector Machine for an SOFC-GT System
CN114492195A (en) CAE model multi-parameter intelligent correction calculation method based on optimization algorithm
Xiong et al. A flow-rate-aware data-driven model of vanadium redox flow battery based on gated recurrent unit neural network
Tingting et al. Evaluation of regional economic innovation ability based on neural network
Shiyun et al. Data analysis and cloud computing of power grid infrastructure projects under the background of informatization
Upadhya et al. A Dispatch Optimization Model for Hybrid Renewable and Battery Systems Incorporating a Battery Degradation Model
CN114970307B (en) General reverse calculation method applied to material design optimization of high-end equipment
CN113420486B (en) Battery anode material integrated design method and system based on multi-scale simulation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22961775

Country of ref document: EP

Kind code of ref document: A1