CN117110892A - Early warning method and system for lithium precipitation of battery and electronic equipment - Google Patents
Early warning method and system for lithium precipitation of battery and electronic equipment Download PDFInfo
- Publication number
- CN117110892A CN117110892A CN202311096169.7A CN202311096169A CN117110892A CN 117110892 A CN117110892 A CN 117110892A CN 202311096169 A CN202311096169 A CN 202311096169A CN 117110892 A CN117110892 A CN 117110892A
- Authority
- CN
- China
- Prior art keywords
- lithium
- battery
- particle
- precipitation
- early warning
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 176
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 176
- 238000001556 precipitation Methods 0.000 title claims abstract description 98
- 238000000034 method Methods 0.000 title claims abstract description 89
- 238000012549 training Methods 0.000 claims abstract description 53
- 238000000926 separation method Methods 0.000 claims abstract description 35
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 16
- 239000002245 particle Substances 0.000 claims description 107
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 claims description 47
- 229910001416 lithium ion Inorganic materials 0.000 claims description 47
- 238000012360 testing method Methods 0.000 claims description 28
- 230000008859 change Effects 0.000 claims description 22
- 238000007600 charging Methods 0.000 claims description 22
- 238000007599 discharging Methods 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 230000001149 cognitive effect Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 3
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims description 3
- 229910052782 aluminium Inorganic materials 0.000 claims description 3
- 238000010277 constant-current charging Methods 0.000 claims description 3
- 238000010281 constant-current constant-voltage charging Methods 0.000 claims description 3
- 229910002804 graphite Inorganic materials 0.000 claims description 3
- 239000010439 graphite Substances 0.000 claims description 3
- 238000011835 investigation Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 abstract description 7
- 238000010801 machine learning Methods 0.000 abstract description 5
- 238000001514 detection method Methods 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000009659 non-destructive testing Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Secondary Cells (AREA)
Abstract
The embodiment of the application discloses a battery lithium precipitation early warning method, a system and electronic equipment, wherein the method comprises the following steps: acquiring first measured data in the charge-discharge cycle process of an experimental battery; preprocessing the first measured data to construct a data sample set; extracting lithium separation characteristics of the lithium separation battery according to the data sample set; inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training, and obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method comprises the steps of carrying out a first treatment on the surface of the Inputting second measured data of the battery to be measured into an XGB model for analysis to obtain lithium precipitation probability of the battery to be measured; for lithium precipitation probability greater than lithium precipitation threshold S 1 The battery to be measured is subjected to early warning. The method provided by the application fuses the lithium battery mechanism model with the machine learning, and early warning is carried out on the lithium battery by the machine learning model, so that the lithium battery can be liftedThe battery with the lithium analysis sign is detected before, the accident occurrence is reduced, the use safety of the battery is improved, the misjudgment rate of the lithium analysis battery is low, and the accuracy of the model is high.
Description
Technical Field
The application relates to the technical field of battery safety early warning, in particular to a battery lithium-precipitation early warning method, a system and electronic equipment.
Background
Currently, lithium ion batteries are increasingly used in digital, energy storage equipment, electric automobiles and the like. The problems of battery aging, lithium precipitation, capacity water jump and the like are also highlighted gradually, and the problem of lithium precipitation in the charging process of the battery is needed to be solved, so that the battery performance is reduced, the service life of the battery is greatly shortened, the quick charge capacity and rate of the battery are limited, and even catastrophic safety accidents such as combustion, explosion and the like are caused.
Disclosure of Invention
The embodiment of the application provides a battery lithium-precipitation early warning method, a system and electronic equipment, which are used for solving the technical problem that the battery lithium-precipitation cannot be early warned in advance in the prior art.
In order to solve the technical problems, the embodiment of the application discloses the following technical scheme:
in a first aspect, a battery lithium precipitation early warning method is provided, the method comprising:
acquiring first measured data in the charge-discharge cycle process of an experimental battery;
preprocessing the first measured data to construct a data sample set;
extracting lithium analysis characteristics of the lithium analysis battery according to the data sample set;
inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training, and obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 ;
Inputting second measured data of the battery to be tested into the XGB model for analysis to obtain lithium precipitation probability of the battery to be tested;
for the lithium separation probability being greater than the lithium separation threshold S 1 And (3) carrying out early warning on the battery to be detected.
With reference to the first aspect, the method for extracting the lithium analysis characteristics of the lithium analysis battery according to the data sample set includes:
drawing a change curve of charge-discharge voltage, charge-discharge current, charge-discharge time or battery capacity of the single lithium-ion battery;
comparing the change curves of a plurality of lithium-ion batteries to obtain lithium-ion characteristics;
the lithium separation characteristic comprises general characteristics and important characteristics, the number of the general characteristics is at least 30, and the number of the important characteristics is at least 10.
In combination with the first aspect, before extracting the lithium separation characteristic of the lithium separation battery, the method further includes distinguishing the lithium separation condition of the experimental battery, and the distinguishing method includes:
disassembling the experimental battery, and judging according to the internal condition of the experimental battery;
the first measured data of the experimental battery with lithium precipitation condition is formed into a data sample set.
With reference to the first aspect, the method for inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training includes:
establishing an XGB model, and dividing a data sample set into a training set and a testing set by adopting hierarchical sampling;
inputting the data in the training set and the lithium analysis characteristics into an XGB model for training;
iterating the XGB model by adopting cross verification, and identifying key parameters of the XGB model by using a particle swarm algorithm at the same time to identify final parameters of the XGB model;
the parameters include one or more of a depth of the tree, a minimum number of samples of leaf nodes, a learning rate, an L1 regularization term, and an L2 regularization term.
With reference to the first aspect, the lithium precipitation threshold S is obtained by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method of (1) comprises:
initializing m particles, wherein each particle is randomly distributed in an n-dimensional space, and the position of the particle in each dimension is within 0-1;
establishing a fitness function: cross entropy loss;
comparing the current fitness of each particle with the historical fitness;
if the current fitness is greater than the historical fitness, updating the current fitness of the particle to the historical fitness which is smaller;
comparing the minimum value of each particle in the particle swarm with the group history minimum value;
if the current fitness is lower, updating the group history best to the current group best fitness;
updating the speed and position of each particle: speed of particle id at k+1st generation:
wherein,representing an inertia part, which consists of inertia weight and particle self-velocity, and represents the trust of the particle on the previous self-motion state; ω represents inertial weight; />The representative cognitive part represents the thinking of the particle, namely the experience part of the particle, and is the distance and direction between the current position of the particle and the self history optimal position;representing a social part, representing information sharing and cooperation among particles, and being the distance and direction between the current position of the particles and the optimal position of the group history; in the formula, k represents the number of iterations; id represents a particle; v represents the particle velocity itself; c 1 Representing individual learning factors; c 2 Representing a population learning factor; r is (r) 1 And r 2 All represent random numbers, interval 0,1]The search randomness is increased; />Representing the historical optimal position of the particle i in the d dimension in the kth iteration, namely, searching the optimal solution obtained by the ith particle after the kth iteration; />After the kth iteration, searching the optimal solution obtained by the ith particle in the representative group; />Representing the position vector of particle i in the d-th dimension in the k-th iteration; />A speed variable representing the d-th dimension of particle i in the kth iteration;
updating the particle speed and the position, and setting the particle position as 0 or 1 of the boundary if the particle position exceeds the range of 0-1;
checking whether the optimal parameter meets the minimum threshold standard at the moment, and if so, taking the parameter;
if not, continuing to update the track until the track is satisfied;
determination of the lithium separation threshold S by multiple cycles 1 And a lithium non-precipitation threshold S 2 。
With reference to the first aspect, the method for preprocessing the first measured data includes:
the method of preprocessing the first measured data includes one or more of data deduplication, decommissioning, or consistent investigation.
With reference to the first aspect, the method for obtaining the first measured data in the charge-discharge cycle process of the experimental battery includes:
charging is started after the charge of the experimental battery is exhausted, and every t 1 Acquiring first measured data for one time until the electric quantity of the experimental battery is full, and standing for more than 30 seconds;
discharging the test cell from full charge to depleted charge and every t 2 Acquiring the first measured data once in a second, and standing for more than 30 seconds;
the charging mode comprises one or more of constant-current charging, constant-current constant-voltage charging and constant-power charging;
the discharging mode comprises one or two of constant-current discharging or constant-power discharging.
With reference to the first aspect, the first measured data includes one or more of a step name, a step number, a cycle number, a charge-discharge current, a charge-discharge voltage, a battery capacity, and a charge-discharge time.
With reference to the first aspect, the experimental battery and the battery to be tested are lithium batteries, and the lithium batteries comprise aluminum shells or soft packages taking graphite as a negative electrode;
the charging multiplying power of the lithium battery is between 0.01 and 5 ℃.
In a second aspect, there is provided a battery lithium-ion early warning system, the system comprising:
the sampling module is used for acquiring first measured data in the charge-discharge cycle process of the experimental battery;
the preprocessing module is used for preprocessing the first measured data and then constructing a data sample set;
the extraction module is used for extracting lithium-precipitation characteristics of the lithium-precipitation battery according to the data sample set;
the training module is used for inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training;
the optimizing module is used for obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 ;
The analysis module is used for inputting second measured data of the battery to be tested into the XGB model for analysis, and obtaining lithium precipitation probability of the battery to be tested;
the early warning module is used for carrying out early warning on the battery to be detected, wherein the lithium separation probability of the battery to be detected is larger than the lithium separation threshold S1.
In a third aspect, there is provided an electronic device comprising a memory storing a computer program and a processor implementing the battery lithium-out warning method of any one of the first aspects when the computer program is executed.
One of the above technical solutions has the following advantages or beneficial effects:
the application provides a battery lithium precipitation early warning method, which comprises the following steps: acquiring first measured data in the charge-discharge cycle process of an experimental battery; for the first actual measurementPreprocessing data to construct a data sample set; extracting lithium separation characteristics of the lithium separation battery according to the data sample set; inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training, and obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method comprises the steps of carrying out a first treatment on the surface of the Inputting second measured data of the battery to be measured into an XGB model for analysis to obtain lithium precipitation probability of the battery to be measured; for lithium precipitation probability greater than lithium precipitation threshold S 1 The battery to be measured is subjected to early warning. According to the method provided by the application, the lithium battery mechanism model is fused with machine learning, so that on one hand, the detection conditions are relaxed, and the purpose of nondestructive detection is realized; on the other hand, the embedding mechanism improves the defect that the traditional machine learning is completely driven by pure data and has no interpretability. And the machine learning model is used for early warning the lithium-ion battery in advance, so that the battery with the lithium-ion sign can be detected in advance, the accident occurrence is reduced, and the use safety of the battery is improved. The method provided by the application has low misjudgment rate on the lithium battery and high accuracy of the model.
Drawings
The technical solution and other advantageous effects of the present application will be made apparent by the following detailed description of the specific embodiments of the present application with reference to the accompanying drawings.
Fig. 1 is a schematic diagram of a battery lithium-precipitation early warning method according to an embodiment of the present application;
fig. 2 is a schematic diagram of a battery lithium-precipitation early warning process according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a verification result provided by an embodiment of the present application;
fig. 4 is a schematic diagram of a battery lithium-ion early warning system according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings are merely for convenience in describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present application. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more of the described features. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
The applicant notes that the current methods of lithium analysis detection in the academia and industry are broadly divided into two types: the main means of the method is to disassemble the battery lower frame, observe the shape and distribution of the lithium precipitation layer on the battery surface through a microscope or a related spectrum instrument so as to judge the lithium precipitation condition of the battery; and secondly, nondestructive testing based on mechanism analysis, namely, constructing a mechanism model to detect whether lithium precipitation occurs by carrying out related processing on battery charge-discharge cycle data, wherein the mechanism model mainly comprises a voltage relaxation curve method, a charge transfer impedance detection method and a negative electrode potential method. The charge transfer impedance detection method and the negative electrode potential method have higher judgment precision, but have strict measurement requirements and are difficult to realize in the whole vehicle environment; although the voltage relaxation curve method can be realized, in many cases, lithium precipitation peaks of a battery core in a relaxation stage, such as a step charge cycle test, silicon doping of a cathode and the like, cannot appear. In addition, the mechanism or disassembly is based on the fact that lithium precipitation is detected after the lithium precipitation is carried out, and the early warning aim is not achieved.
The following further illustrates embodiments of the application by way of example only:
as shown in fig. 1, the embodiment of the application discloses a battery lithium-precipitation early warning method, which comprises the following steps:
s1: and acquiring first measured data in the charge-discharge cycle process of the experimental battery.
Specifically, the acquisition method comprises the following steps:
in the embodiment of the application, the charge is started after the electric quantity of the experimental battery is exhausted, and every t 1 Acquiring first measured data in a second time until the electric quantity of the experimental battery is full, and standing for more than 30 seconds; wherein t is more than or equal to 0.05 1 Less than or equal to 5, and charging the experimental battery by adopting a charging multiplying power of 0.01C-5C after the electric quantity of the experimental battery is exhausted; and the first measured data are collected every 0.05-5 seconds. It should be noted that the charging rate is not always unchanged, the charging rate can be adjusted according to the quantity of the electric quantity of the experimental battery, when the electric quantity of the experimental battery is small, the charging rate is close to 5C, the experimental battery can rapidly accumulate the electric quantity, and when the circuit of the experimental battery is close to saturation, for example, when the circuit reaches 90%, the charging rate is close to 0.01C; meanwhile, in the whole charging process, the charging rate can be changed according to the electric quantity of the experimental battery. The charging mode comprises one or more of constant-current charging, constant-current constant-voltage charging and constant-power charging. In addition, the time interval for collecting the first measured data can also be changed, for example, when the experimental battery just starts to charge, the collecting time is close to 0.05 seconds; when the experimental battery is charged to a stable stage of 20% -80%, the collection time is close to 5 seconds; when the experimental battery is charged to more than 80%, the acquisition time is close to 0.05 seconds. It is contemplated that the acquisition time interval may also be constant from the beginning of charge to the end of charge or proportional or inversely proportional to how much of the charge is. Thus, values for the acquisition time interval may be selected to include 0.05s, 0.08s, 0.1s, 0.15s, 0.2s, 0.25s, 0.3s, 0.5s, 0.8s, 1s, 1.5s, 2s, 2.5s, 3s, 3.5s, 4s, 4.5s, 5s.
In the embodiment of the application, the experimental battery is discharged from full charge to exhaustion, and every t 2 Acquiring first measured data for one second, and standing for more than 30 seconds; wherein t is more than or equal to 0.05 2 The test battery can be discharged at the same or different rate with the charging rate after being charged and is charged at less than or equal to 5, the first measured data are collected at intervals of 0.05-5 seconds in the discharging process, and the battery is charged after the discharging is finishedAnd (5) standing. The discharging mode comprises one or two of constant-current discharging or constant-power discharging.
In an embodiment of the present application, the first measured data includes one or more of a step name, a step number, a number of cycles, a charge-discharge current, a charge-discharge voltage, a battery capacity, and a charge-discharge time. The process step name and the process step serial number are only acquired once, the cycle number is calculated in an accumulated way, and the charge and discharge current, the charge and discharge voltage, the battery capacity and the charge and discharge time are acquired once at regular intervals.
S2: and preprocessing the first measured data to construct a data sample set.
In the embodiment of the application, the specific method comprises the following steps: the method of preprocessing the first measured data includes one or more of data deduplication, decommissioning, or consistent investigation. The data deduplication comprises overlapping data in the acquired first measured data, wherein the overlapping data comprises the same data obtained after the experimental battery is fully charged; the de-pseudo data comprises data which does not accord with the actual situation in the data measurement process, such as the data which is not zero and appears becomes negative number or zero, and then becomes normal value; a consistency survey involves comparing the consistency of the data from front to back.
S3: and extracting the lithium separation characteristic of the lithium separation battery according to the data sample set.
In the embodiment of the application, before extracting the lithium separation characteristic of the lithium separation battery, the method further comprises the step of judging the lithium separation condition of the experimental battery, wherein the judging method comprises the following steps: disassembling the experimental battery, and judging according to the internal condition of the experimental battery; the first measured data of the experimental battery with lithium precipitation condition is formed into a data sample set. And disassembling the tested experimental batteries, observing the internal lithium precipitation condition of the experimental batteries through a high-precision instrument such as a high-power microscope, a related spectrum instrument and the like, judging whether lithium precipitation occurs according to the distribution condition of an internal structure, deleting the experimental batteries and experimental data without lithium precipitation, reserving the experimental batteries and the experimental data with lithium precipitation, and forming a data sample set from the first measured data obtained from the experimental battery data with lithium precipitation.
In the embodiment of the application, the specific method for extracting the lithium analysis characteristics comprises the following steps: drawing a change curve of charge-discharge voltage, charge-discharge current, charge-discharge time or battery capacity of a single lithium-ion battery; comparing the change curves of a plurality of lithium-separating batteries to obtain lithium-separating characteristics; the lithium separation characteristics comprise general characteristics and important characteristics, the number of the general characteristics is at least 30, and the number of the important characteristics is at least 10. After the first measured data of all the lithium-ion battery are obtained, the charge-discharge voltage, the charge-discharge current and the battery capacity in the first measured data are combined with the charge-discharge time to draw a change curve, and the charge-discharge voltage and the charge-discharge time can be drawn to form a voltage-time change curve, or the charge-discharge voltage and the battery capacity can be drawn to form a voltage-capacity change curve, or the charge-discharge voltage and the charge-discharge current can be drawn to form a voltage-current change curve, or the charge-discharge current and the charge-discharge time can be drawn to form a current-time change curve, or the battery capacity can be drawn to form a current-capacity change curve. And searching for the lithium analysis characteristics by combining the change curves of all the lithium analysis batteries, wherein the lithium analysis characteristics comprise important characteristics with obvious change and general characteristics with insignificant change.
S4: inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training, and obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 。
In the embodiment of the application, the specific model training method comprises the following steps: establishing an XGB model, and dividing a data sample set into a training set and a testing set by adopting hierarchical sampling; specifically, the data of the training set and the data of the test set are not overlapped with each other, and the data of the training set and the data set occupy a certain proportion. Typically, the training set is 70% of the original data, and the test set is 30% of the original data. After the data training in the training set is finished, the data in the testing set is input into the XGB model for testing, so that the accuracy of the XGB model is judged. If the accuracy reaches a certain value, the model training is successful, otherwise, the model is not successfully retrained. The proportion of the training set and the test set data can be adjusted according to the training intensity. The training set may occupy 60% -80% and the test set may occupy 20% -40%. It will be appreciated that the more training data in the training set, the higher the accuracy of model training, and the more test data in the test set, the more accurate the results of the model test, both of which can help the model to refine.
In the embodiment of the application, data and lithium analysis characteristics in a training set are input into an XGB model for training; during training, matching the data in the training set with lithium analysis features; adopting a cross verification iteration XGB model, and simultaneously using a particle swarm algorithm to identify key parameters of the XGB model and distinguish final parameters of the XGB model; wherein the identified parameters include one or more of a depth of the tree, a minimum number of samples of the leaf nodes, a learning rate, an L1 regularization term, and an L2 regularization term.
In the embodiment of the application, specifically, the lithium separation threshold S is obtained by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method of (1) comprises: initializing m particles, wherein each particle is randomly distributed in an n-dimensional space, and the position of the particle in each dimension is within 0-1;
establishing a fitness function: cross entropy loss;
comparing the current fitness of each particle with the historical fitness;
if the current fitness is greater than the historical fitness, updating the current fitness of the particle to the historical fitness which is smaller;
comparing the minimum value of each particle in the particle swarm with the group history minimum value;
if the current fitness is lower, updating the group history best to the current group best fitness;
updating the speed and position of each particle: speed of particle id at k+1st generation:
wherein the method comprises the steps of,Representing an inertia part, which consists of inertia weight and particle self-velocity, and represents the trust of the particle on the previous self-motion state; ω represents inertial weight; />The representative cognitive part represents the thinking of the particle, namely the experience part of the particle, and is the distance and direction between the current position of the particle and the self history optimal position;representing a social part, representing information sharing and cooperation among particles, and being the distance and direction between the current position of the particles and the optimal position of the group history; in the formula, k represents the number of iterations; id represents a particle; v represents the particle velocity itself; c 1 Representing individual learning factors; c 2 Representing a population learning factor; r is (r) 1 And r 2 All represent random numbers, interval 0,1]The search randomness is increased; />Representing the historical optimal position of the particle i in the d dimension in the kth iteration, namely, searching the optimal solution obtained by the ith particle after the kth iteration; />After the kth iteration, searching the optimal solution obtained by the ith particle in the representative group; />Representing the position vector of particle i in the d-th dimension in the k-th iteration; />A speed variable representing the d-th dimension of particle i in the kth iteration;
in an embodiment of the application, the particle velocity and position are updatedIf the particle position exceeds the range of 0-1, setting the particle position as 0 or 1 of the boundary; checking whether the optimal parameter meets the minimum threshold standard at the moment, and if so, taking the parameter; if not, continuing to update the track until the track is satisfied; determination of the lithium separation threshold S by multiple cycles 1 And a lithium non-precipitation threshold S 2 。
S5: and inputting the second measured data of the battery to be measured into an XGB model for analysis to obtain the lithium precipitation probability of the battery to be measured.
In the embodiment of the application, first, the second measured data of the battery to be measured needs to be acquired, and it can be understood that the second measured data is one or more data in the first measured data. When the battery to be measured is obtained, after the electric quantity of the battery to be measured is exhausted, the second measured data of the battery to be measured can be obtained respectively by adopting the same charging rate and discharging rate of the experimental battery, or the second measured data of the battery to be measured in the using process can be obtained. After the second measured data are obtained, drawing a change curve according to the second measured data, inputting the change curve into a constructed XGB model, and searching the characteristics in the change curve through the XGB model so as to obtain the lithium precipitation probability of the battery to be tested; or the obtained second measured data is directly input into an XGB model, and the XGB model directly calculates the lithium precipitation probability of the battery to be measured according to the second measured data.
S6: for lithium precipitation probability greater than lithium precipitation threshold S 1 The battery to be measured is subjected to early warning.
In the embodiment of the application, the obtained lithium precipitation probability and the lithium precipitation threshold S 1 Comparing, if the lithium precipitation probability is greater than the lithium precipitation threshold S 1 Indicating that the lithium precipitation condition of the battery to be tested exists; when the lithium separation probability is at the lithium separation threshold S 1 And a lithium non-precipitation threshold S 2 When the lithium ion battery to be detected is in the middle, the situation that lithium ion is not generated, but the lithium ion risk exists is indicated; when the lithium precipitation probability is lower than the lithium non-precipitation threshold S 2 And when the lithium ion battery to be detected is not in the lithium ion condition, the lithium ion risk is avoided.
In the embodiment of the application, the above-mentioned experimental battery and the battery to be tested are lithium batteries, and the lithium batteries comprise an aluminum shell or a soft package taking graphite as a negative electrode. It can be appreciated that lithium ions are filled in a lithium battery, so that prediction can be performed by constructing a model by the method. In some other embodiments, when the battery to be tested is not a lithium battery, the same method steps in the application can be adopted to construct a corresponding model so as to judge the leakage condition of other types of batteries.
As shown in fig. 2, the specific implementation steps of the battery lithium analysis early warning method provided by the embodiment of the application include:
step one: data acquisition and data processing.
The first step: acquiring stored battery data through a database;
and a second step of: a data standardization flow; data extracted from the database including record number, step name, cycle, step number, current, voltage and absolute time fields;
and a third step of: the abnormal extreme data and the missing data are analyzed and preprocessed.
Step two: extracting data characteristics and constructing a sample set;
features are extracted from charge and discharge data in the battery cycle process, and 980 battery actual measurement data are used as a sample set in the scheme.
Step three: and establishing an XGB model, and searching for optimal parameters by using a PSO algorithm.
Specifically, firstly, a sample set is divided into a training set and a test set, wherein the training set accounts for 70% of the data of the sample set, and the test set accounts for 30% of the data of the sample set; and calling an XGB model to perform battery lithium precipitation detection. And carrying out parameter optimal identification by using a PSO algorithm, wherein the particle swarm algorithm super-parameter optimization is based on the historical evaluation result of the objective function and is continuously and iteratively updated. The most important difference between the method and the method is that the method considers the result of the last step when updating the super parameter in each step, can omit the time cost and can better converge and fit.
And training the training set sample, inputting the hyper-parameters to evaluate in the objective function, respectively obtaining a return value and cross entropy loss, and performing loop iteration until the stopping condition is met, thereby completing the training. Table 1 is a partial optimal super parameter set, parameters obtained by optimizing are substituted into an XGB model, a cross verification result is obtained, and as shown in FIG. 3, the accuracy is above 90%, the average accuracy of 10 groups of data is 95.09%, and the stability and accuracy are good.
Table 1 partial best super parameter set
Step four: and after the optimal parameters are determined, the model stability and accuracy are evaluated.
Substituting the optimal super parameters into the model to obtain a model evaluation result, and storing the XGB model.
The data types are shown in Table 2, and the evaluation results are shown in Table 3.
Table 2 training set data types
Training set | Non-precipitation lithium (prediction) | Lithium precipitation (prediction) |
Non-precipitation lithium (true) | 228 | 11 |
Lithium precipitation (true) | 21 | 426 |
Table 3 training set evaluation results
Training set | Accuracy rate of | Recall rate of recall | f1_score | Number of samples |
Lithium is not separated out | 0.92 | 0.95 | 0.93 | 239 |
Lithium precipitation | 0.97 | 0.95 | 0.96 | 447 |
Step five: substituting the test set sample, and checking the fitting condition of the model.
Substituting the test set data into the XGB model to obtain result output, wherein the evaluation indexes comprise accuracy, recall rate, precision and f1_fraction.
The test set detection results are shown in table 5, and the test set accuracy is 0.88, namely 294 battery samples correctly predict 258 labels of the battery samples through an XGB model; the accuracy of the lithium-ion battery is 180/199=0.90; recall 180/197 = 0.91; f1_score=2×0.90×0.91)/(0.90+0.91) =0.90, and the model detection effect is good, and no obvious overfitting condition is found.
Table 4 test set data types
Test set | Non-precipitation lithium (prediction) | Lithium precipitation (prediction) |
Non-precipitation lithium (true) | 78 | 19 |
Lithium precipitation (true) | 17 | 180 |
Table 5 test set test results
As shown in fig. 4, the embodiment of the application further provides a battery lithium-precipitation early warning system, which includes: the device comprises a sampling module, a preprocessing module, an extracting module, a training module, an optimizing module, an analyzing module and an early warning module.
The sampling module is used for acquiring first measured data in the charge-discharge cycle process of the experimental battery; the acquisition method comprises the following steps: charging is started after the charge of the experimental battery is exhausted, and every t 1 Second acquisitionTaking first measured data once until the electric quantity of the experimental battery is full, and standing for more than 30 seconds; discharging the test cell from full charge to depleted charge, and every t 2 Acquiring first measured data for one second, and standing for more than 30 seconds; the first measured data comprises one or more of a step name, a step number, a cycle number, a charge and discharge current, a charge and discharge voltage, a battery capacity and a charge and discharge time.
The preprocessing module is used for preprocessing the first measured data to construct a data sample set; the method of preprocessing includes preprocessing the first measured data including one or more of data deduplication, decommissioning, or consistent surveys.
The extraction module is used for extracting lithium analysis characteristics of the lithium analysis battery according to the data sample set; the specific extraction method comprises the following steps: disassembling the experimental battery, and judging according to the internal condition of the experimental battery; forming a data sample set from first measured data of an experimental battery with lithium precipitation condition; then drawing a change curve of charge-discharge voltage, charge-discharge current, charge-discharge time or battery capacity of a single lithium-ion battery; comparing the change curves of a plurality of lithium-separating batteries to obtain lithium-separating characteristics; the lithium separation characteristics comprise general characteristics and important characteristics, the number of the general characteristics is at least 30, and the number of the important characteristics is at least 10.
The training module is used for inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training; the specific training method comprises the following steps: establishing an XGB model, dividing a data sample set into a training set and a testing set by adopting hierarchical sampling, and inputting data and lithium analysis characteristics in the training set into the XGB model for training; during training, matching the data in the training set with lithium analysis features; and adopting a cross verification iteration XGB model, and simultaneously using a particle swarm algorithm to identify key parameters of the XGB model and distinguish final parameters of the XGB model.
The optimizing module is used for obtaining the lithium separation threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method comprises the steps of carrying out a first treatment on the surface of the The specific optimizing method comprises the following steps: obtaining a lithium separation threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method of (1) comprises: initializing mParticles, each of which is randomly distributed in an n-dimensional space, wherein the position of the particle in each dimension is within 0-1;
establishing a fitness function: cross entropy loss;
comparing the current fitness of each particle with the historical fitness;
if the current fitness is greater than the historical fitness, updating the current fitness of the particle to the historical fitness which is smaller;
comparing the minimum value of each particle in the particle swarm with the group history minimum value;
if the current fitness is lower, updating the group history best to the current group best fitness;
updating the speed and position of each particle: speed of particle id at k+1st generation:
wherein,representing an inertia part, which consists of inertia weight and particle self-velocity, and represents the trust of the particle on the previous self-motion state; ω represents inertial weight; />The representative cognitive part represents the thinking of the particle, namely the experience part of the particle, and is the distance and direction between the current position of the particle and the self history optimal position;representing a social part, representing information sharing and cooperation among particles, and being the distance and direction between the current position of the particles and the optimal position of the group history; in the formula, k represents the number of iterations; id represents a particle; v represents the particle velocity itself; c 1 Representing individual learning factors; c 2 Representing a population learning factor; r is (r) 1 And r 2 All of which represent a random number and,interval [0,1 ]]The search randomness is increased; />Representing the historical optimal position of the particle i in the d dimension in the kth iteration, namely, searching the optimal solution obtained by the ith particle after the kth iteration; />After the kth iteration, searching the optimal solution obtained by the ith particle in the representative group; />Representing the position vector of particle i in the d-th dimension in the k-th iteration; />A speed variable representing the d-th dimension of particle i in the kth iteration;
in the embodiment of the application, the particle speed and the position are updated, and if the particle position exceeds the range of 0-1, the particle position is set as 0 or 1 of the boundary; checking whether the optimal parameter meets the minimum threshold standard at the moment, and if so, taking the parameter; if not, continuing to update the track until the track is satisfied; determination of the lithium separation threshold S by multiple cycles 1 And a lithium non-precipitation threshold S 2 。
The analysis module is used for inputting second measured data of the battery to be tested into the XGB model for analysis to obtain lithium precipitation probability of the battery to be tested; specifically, first, second measured data of a battery to be measured needs to be obtained; and drawing a change curve according to the second measured data, and inputting the change curve or the second measured data into the XGB model for analysis and judgment to obtain the lithium precipitation probability.
The early warning module is used for judging that the lithium precipitation probability is larger than the lithium precipitation threshold S 1 The battery to be measured is subjected to early warning. Specifically, the obtained lithium precipitation probability and the lithium precipitation threshold S 1 Comparing, if the lithium precipitation probability is greater than the lithium precipitation threshold S 1 Indicating that the lithium precipitation condition of the battery to be tested exists; when the lithium separation probability is at the lithium separation threshold S 1 And a lithium non-precipitation threshold S 2 Between which are locatedWhen the lithium ion battery to be detected is in the lithium ion battery state, the lithium ion battery to be detected is not in the lithium ion battery state, but the lithium ion battery is in the lithium ion battery state; when the lithium precipitation probability is lower than the lithium non-precipitation threshold S 2 And when the lithium ion battery to be detected is not in the lithium ion condition, the lithium ion risk is avoided.
The embodiment of the application also provides electronic equipment, which comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the battery lithium precipitation early warning method provided by any embodiment when executing the computer program.
The foregoing describes in detail a method, a system and an electronic device for early warning of battery lithium precipitation provided by the embodiments of the present application, and specific examples are applied to illustrate the principles and embodiments of the present application, where the descriptions of the foregoing examples are only used to help understand the technical solution and core ideas of the present application; those of ordinary skill in the art will appreciate that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application.
Claims (11)
1. The battery lithium precipitation early warning method is characterized by comprising the following steps of:
acquiring first measured data in the charge-discharge cycle process of an experimental battery;
preprocessing the first measured data to construct a data sample set;
extracting lithium analysis characteristics of the lithium analysis battery according to the data sample set;
inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training, and obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 ;
Inputting second measured data of the battery to be tested into the XGB model for analysis to obtain lithium precipitation probability of the battery to be tested;
for the lithium separation probability being greater than the lithium separation threshold S 1 And (3) carrying out early warning on the battery to be detected.
2. The battery lithium-ion early warning method according to claim 1, wherein the method for extracting the lithium-ion characteristic of the lithium-ion battery according to the data sample set comprises the following steps:
drawing a change curve of charge-discharge voltage, charge-discharge current, charge-discharge time or battery capacity of the single lithium-ion battery;
comparing the change curves of a plurality of lithium-ion batteries to obtain lithium-ion characteristics;
the lithium separation characteristic comprises general characteristics and important characteristics, the number of the general characteristics is at least 30, and the number of the important characteristics is at least 10.
3. The battery lithium-ion early warning method of claim 1, further comprising discriminating a lithium-ion condition of the experimental battery before extracting the lithium-ion characteristic of the lithium-ion battery, the discriminating method comprising:
disassembling the experimental battery, and judging according to the internal condition of the experimental battery;
the first measured data of the experimental battery with lithium precipitation condition is formed into a data sample set.
4. The battery lithium-ion early warning method of claim 1, wherein the method for inputting the extracted lithium-ion characteristics into a pre-constructed XGB model for training comprises the following steps:
establishing an XGB model, and dividing a data sample set into a training set and a testing set by adopting hierarchical sampling;
inputting the data in the training set and the lithium analysis characteristics into an XGB model for training;
iterating the XGB model by adopting cross verification, and identifying key parameters of the XGB model by using a particle swarm algorithm at the same time to identify final parameters of the XGB model;
the parameters include one or more of a depth of the tree, a minimum number of samples of leaf nodes, a learning rate, an L1 regularization term, and an L2 regularization term.
5. The battery lithium-ion early warning method according to claim 1, characterized in that the lithium-ion threshold S is obtained by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 The method of (1) comprises:
initializing m particles, wherein each particle is randomly distributed in an n-dimensional space, and the position of the particle in each dimension is within 0-1;
establishing a fitness function: cross entropy loss;
comparing the current fitness of each particle with the historical fitness;
if the current fitness is greater than the historical fitness, updating the current fitness of the particle to the historical fitness which is smaller;
comparing the minimum value of each particle in the particle swarm with the group history minimum value;
if the current fitness is lower, updating the group history best to the current group best fitness;
updating the speed and position of each particle: speed of particle id at k+1st generation:
wherein,representing an inertia part, which consists of inertia weight and particle self-velocity, and represents the trust of the particle on the previous self-motion state; ω represents inertial weight; />The representative cognitive part represents the thinking of the particle, namely the experience part of the particle, and is the distance and direction between the current position of the particle and the self history optimal position;representative societyThe meeting part represents information sharing and cooperation among particles, and is the distance and direction between the current position of the particles and the optimal position of the group history; in the formula, k represents the number of iterations; id represents a particle; v represents the particle velocity itself; c 1 Representing individual learning factors; c 2 Representing a population learning factor; r is (r) 1 And r 2 All represent random numbers, interval 0,1]The search randomness is increased; />Representing the historical optimal position of the particle i in the d dimension in the kth iteration, namely, searching the optimal solution obtained by the ith particle after the kth iteration; />After the kth iteration, searching the optimal solution obtained by the ith particle in the representative group; />Representing the position vector of particle i in the d-th dimension in the k-th iteration; />A speed variable representing the d-th dimension of particle i in the kth iteration;
updating the particle speed and the position, and setting the particle position as 0 or 1 of the boundary if the particle position exceeds the range of 0-1;
checking whether the optimal parameter meets the minimum threshold standard at the moment, and if so, taking the parameter;
if not, continuing to update the track until the track is satisfied;
determination of the lithium separation threshold S by multiple cycles 1 And a lithium non-precipitation threshold S 2 。
6. The battery lithium-ion early warning method of claim 1, wherein the method for preprocessing the first measured data comprises:
the method of preprocessing the first measured data includes one or more of data deduplication, decommissioning, or consistent investigation.
7. The battery lithium-ion early warning method according to claim 1, wherein the method for acquiring the first measured data in the charge-discharge cycle of the experimental battery comprises the following steps:
charging is started after the charge of the experimental battery is exhausted, and every t 1 Acquiring first measured data for one time until the electric quantity of the experimental battery is full, and standing for more than 30 seconds;
discharging the test cell from full charge to depleted charge and every t 2 Acquiring the first measured data once in a second, and standing for more than 30 seconds;
the charging mode comprises one or more of constant-current charging, constant-current constant-voltage charging and constant-power charging;
the discharging mode comprises one or two of constant-current discharging or constant-power discharging.
8. The battery lithium-ion early warning method of claim 1, wherein the first measured data includes one or more of a step name, a step number, a number of cycles, a charge-discharge current, a charge-discharge voltage, a battery capacity, and a charge-discharge time.
9. The battery lithium-ion early warning method according to claim 1, wherein the experimental battery and the battery to be tested are lithium batteries, and the lithium batteries comprise aluminum shells or soft packages taking graphite as a negative electrode;
the charging multiplying power of the lithium battery is 0.01-5 ℃.
10. A battery lithium-ion early warning system, the system comprising:
the sampling module is used for acquiring first measured data in the charge-discharge cycle process of the experimental battery;
the preprocessing module is used for preprocessing the first measured data and then constructing a data sample set;
the extraction module is used for extracting lithium-precipitation characteristics of the lithium-precipitation battery according to the data sample set;
the training module is used for inputting the extracted lithium analysis characteristics into a pre-constructed XGB model for training;
the optimizing module is used for obtaining a lithium analysis threshold S by adopting an optimizing algorithm 1 And a lithium non-precipitation threshold S 2 ;
The analysis module is used for inputting second measured data of the battery to be tested into the XGB model for analysis, and obtaining lithium precipitation probability of the battery to be tested;
the early warning module is used for carrying out early warning on the fact that the lithium separation probability is larger than the lithium separation threshold S 1 And (3) carrying out early warning on the battery to be detected.
11. An electronic device comprising a memory and a processor, the memory storing a computer program, wherein the processor, when executing the computer program, implements the battery lithium-out pre-warning method of any one of claims 1 to 9.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311096169.7A CN117110892A (en) | 2023-08-28 | 2023-08-28 | Early warning method and system for lithium precipitation of battery and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311096169.7A CN117110892A (en) | 2023-08-28 | 2023-08-28 | Early warning method and system for lithium precipitation of battery and electronic equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117110892A true CN117110892A (en) | 2023-11-24 |
Family
ID=88794295
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311096169.7A Pending CN117110892A (en) | 2023-08-28 | 2023-08-28 | Early warning method and system for lithium precipitation of battery and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117110892A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117521857A (en) * | 2024-01-05 | 2024-02-06 | 宁德时代新能源科技股份有限公司 | Battery cell lithium analysis method and device, readable storage medium and electronic equipment |
-
2023
- 2023-08-28 CN CN202311096169.7A patent/CN117110892A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117521857A (en) * | 2024-01-05 | 2024-02-06 | 宁德时代新能源科技股份有限公司 | Battery cell lithium analysis method and device, readable storage medium and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111707951B (en) | Battery pack consistency evaluation method and system | |
Yang et al. | A fractional-order model-based battery external short circuit fault diagnosis approach for all-climate electric vehicles application | |
CN112092675B (en) | Battery thermal runaway early warning method, system and server | |
CN109655754B (en) | Battery performance evaluation method based on multi-dimensional grading of charging process | |
CN116502112B (en) | New energy power supply test data management method and system | |
CN103792495B (en) | Method for evaluating battery performance based on Delphi method and grey relation theory | |
WO2019162749A1 (en) | Method and system for online assessing state of health of a battery | |
CN113391213B (en) | Battery micro-fault diagnosis method based on battery pack consistency evolution | |
CN117110892A (en) | Early warning method and system for lithium precipitation of battery and electronic equipment | |
CN114818831B (en) | Bidirectional lithium ion battery fault detection method and system based on multi-source perception | |
CN115587527A (en) | Battery life prediction method, system, terminal device and computer readable medium | |
Liu et al. | Comparative study of incremental capacity curve determination methods for lithium-ion batteries considering the real-world situation | |
CN114280479A (en) | Electrochemical impedance spectrum-based rapid sorting method for retired batteries | |
CN109613446A (en) | A kind of aging of lithium battery detection method based on time series analysis | |
CN112816881A (en) | Battery differential pressure abnormality detection method, battery differential pressure abnormality detection device and computer storage medium | |
CN115994441A (en) | Big data cloud platform online battery life prediction method based on mechanism information | |
Chen et al. | A novel battery health indicator and PSO-LSSVR for LiFePO4 battery SOH estimation during constant current charging | |
Zhao et al. | Data-driven lithium-ion battery degradation evaluation under overcharge cycling conditions | |
Zhao et al. | Battery degradation stage detection and life prediction without accessing historical operating data | |
CN117173100B (en) | Polymer lithium ion battery production control system and method thereof | |
CN117169743A (en) | Battery health state estimation method and device based on partial data and model fusion | |
CN116859259A (en) | Battery health state estimation method based on integrated learning and real vehicle big data | |
US20230349977A1 (en) | Method and apparatus for estimating state of health of battery | |
CN110826645A (en) | Adaboost algorithm-based lithium battery retirement detection method and system | |
CN116047339A (en) | Lithium ion battery pack SOC estimation method and device based on thermoelectric coupling model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |