CN115078999A - Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network - Google Patents
Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network Download PDFInfo
- Publication number
- CN115078999A CN115078999A CN202210554879.9A CN202210554879A CN115078999A CN 115078999 A CN115078999 A CN 115078999A CN 202210554879 A CN202210554879 A CN 202210554879A CN 115078999 A CN115078999 A CN 115078999A
- Authority
- CN
- China
- Prior art keywords
- training
- lithium battery
- hidden layer
- neural network
- output
- 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.)
- Granted
Links
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 107
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 107
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 72
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000012549 training Methods 0.000 claims abstract description 105
- 238000005070 sampling Methods 0.000 claims description 49
- 230000036541 health Effects 0.000 claims description 33
- 238000012360 testing method Methods 0.000 claims description 32
- 230000003044 adaptive effect Effects 0.000 claims description 16
- 238000010606 normalization Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 9
- 230000004913 activation Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 230000001351 cycling effect Effects 0.000 claims description 3
- 238000007599 discharging Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 241000544061 Cuculus canorus Species 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 101100493820 Caenorhabditis elegans best-1 gene Proteins 0.000 description 1
- 101100004280 Caenorhabditis elegans best-2 gene Proteins 0.000 description 1
- 208000032953 Device battery issue Diseases 0.000 description 1
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000005431 greenhouse gas Substances 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 229910001416 lithium ion Inorganic materials 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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]
- G01R31/3644—Constructional arrangements
- G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention provides a lithium battery condition prediction method based on a self-adaptive hidden layer BP neural network, and belongs to the technical field of lithium batteries. The method introduces the charging time into the training of the BP neural network, and improves the SOH estimation precision of the lithium battery; the method has the advantages of wide application range, high prediction precision, strong tracking performance and the like, reduces the dependence of the traditional BP neural network on the initial structure setting of the network, and effectively improves the accuracy of the model.
Description
Technical Field
The invention relates to the technical field of lithium batteries, in particular to a lithium battery condition prediction method based on a self-adaptive hidden layer BP neural network.
Background
In recent years, the rapid development of lithium ion battery technology has driven the rapid development of vehicle electrification, and the popularity of electric vehicles is rapidly increasing. At present, in order to reduce the emission of greenhouse gases and save non-renewable energy, many countries have come out of related policies to promote the development of electric vehicles.
A Battery Management System (BMS) is an important system for securing safety and reliability of a lithium battery, of which the most important are battery state of charge (SOC) estimation and state of health (SOH) estimation. Accurate SOH estimation is critical to ensure the healthy operation of the battery and to prevent sudden battery failure from causing a series of disasters that could be avoided. The change in SOH of the battery may be reflected in a change in parameters inside the battery, such as a decrease in capacity and an increase in ohmic internal resistance.
The SOH estimation method for lithium batteries at present can be roughly divided into the following 2 types:
1. the model-based method comprises two methods based on an electrochemical model and an equivalent circuit model. Although the method based on the electrochemical model has high prediction accuracy, the model structure is very complex and difficult to build because the complicated aging mechanism of the lithium battery needs to be considered, and the method has large calculation amount and usually needs to be combined with other artificial intelligence algorithms. The equivalent model based method is simple in structure and easy to build, but has the cost of large error, low precision, poor robustness and the like.
2. The data driving method comprises machine learning, artificial intelligence algorithm and the like, and the method regards the lithium battery as a black box model and excavates the law of battery performance evolution from a large amount of data. However, the method needs a large amount of data training, the accuracy degree of the data and the structure of the algorithm determine the accuracy of health condition estimation, and the key problem of estimating the SOH of the lithium battery by a data-driven method is to find a proper algorithm and correctly process the original data.
Chinese patent application publication (CN112881914A) discloses a method for predicting the health status of a lithium battery. The method mainly comprises the following steps: a BP neural network improved by a cuckoo algorithm is adopted to represent the structural characteristics of the lithium battery of the electric automobile, a battery health condition prediction model is established aiming at the data driving method, the capacity parameter of the battery pack is tracked and predicted to obtain a capacity estimation value of the lithium battery at a certain moment, and the capacity estimation value is compared with an actual value to realize the prediction of the health condition of the lithium battery of the electric automobile. Although the method for searching the optimal initial value of the network by the cuckoo can effectively avoid the problem that the model is trapped in local optimization, the hidden layer is set to be a fixed value, the accuracy of the hidden layer is influenced to a certain extent when the number of input training sets is reduced, and the method has the characteristic of poor robustness.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a BP neural network with a self-adaptive hidden layer for predicting the health condition of a lithium battery aiming at the problem that the accuracy cannot reach the optimum due to the fact that the original structure setting of a neural network structure is kept under the condition that the number of training sets is reduced or the type of a battery is changed in the prior art, and the BP neural network has the advantages of wide application range, high prediction accuracy, strong tracking performance and the like, meanwhile, the dependence of the traditional BP neural network on the initial structure setting of the network is reduced, and the accuracy of a model is effectively improved.
In order to solve the technical problem of the invention, the adopted technical scheme is as follows: the lithium battery condition prediction method based on the self-adaptive hidden layer BP neural network comprises the following steps of:
step 1, cycling the charge and discharge of a lithium battery for n times, sampling the lithium battery in the charge and discharge cycles for n times, and forming a lithium battery charge and discharge data set by using sampling data;
extracting the following lithium battery sampling parameters related to the health condition of the lithium battery from the charging and discharging data set and recording the parameters as sampling health parameters: obtaining five groups of sampling health parameters by using the sampling end voltage in n times of charge and discharge cycles, the sampling end current in n times of charge and discharge cycles, the sampling lithium battery temperature in n times of charge and discharge cycles, the sampling lithium battery charging time in n times of charge and discharge cycles and the sampling lithium battery capacity in n times of charge and discharge cycles, and marking any one group in the five groups as a sampling health parameter group j, wherein j is 1,2,3,4, 5; any one of n charge-discharge cycles is recorded as a cycle i, i is 1, 2.. n; taking the sampling health parameter group j as a column vector, constructing a health parameter data matrix, and recording as a health matrix H n×5 The expression is as follows:
wherein, V i The sampled terminal voltage measured in cycle i; i is i The current of the sampling end measured in the cycle i is used as the current of the sampling end; t is i Sampling the temperature of the lithium battery measured in the cycle i; t is t i Sampling the lithium battery charging time measured in the cycle i; c i The lithium battery charge capacity measured in cycle i is sampled.
Step 2, carrying out normalization processing on the 5 groups of column vectors to obtain the following data after normalization processing:
the terminal voltages of the n lithium batteries after normalization processing are recorded as terminal voltages
n end currents after normalization processing, and recording the end currents in the circulation i as end currents
The temperature of the n normalized lithium batteries is recorded as the temperature of the lithium batteries in the cycle i
The charging time of the n normalized lithium batteries is recorded as the charging time of the lithium batteries in the cycle i
The charging capacities of the n lithium batteries after normalization processing are recorded as the charging capacity of the lithium battery in the cycle i
N terminal voltages
n end currents
Temperature of n lithium batteries
And n lithium battery charging times
Composing an input dataset X n×4 N lithium battery charging capacity
Form an output data set Y n×1 。
Step 3, inputting a data set X n×4 Divided into two groups, which are input training set X in turn train And input test set X test ；
Will output data set Y n×1 Divided into two groups, which are output training set Y in turn train And output test set Y test ；
Wherein, the proportion of the two training sets in the original data set is a given preset value gamma which is 50-90%;
set output training set Y train The total number of the m data is included, and any one of the m data is recorded as an output training value y f ，f＝1,2,...,m，m＜n。
And step 4, establishing a three-layer BP neural network A and training.
Step 4.1, the structure of the three-layer BP neural network A comprises an input layer, a hidden layer and an output layer; the input layer comprises 4 input layer cells and the output layer comprises only 1 output layer cell; the input layer and the hidden layer are in full connection, and the hidden layer and the output layer are in full connection;
the number of hidden layer units is determined by an empirical formula and an adaptive adjustment formula, wherein the empirical formula is used for estimating the maximum value of the number of hidden layer units, and the empirical formula is as follows:
h is the number of hidden layer units, h is an integer, r is the number of input layer units, s is the number of output layer units, a is a constant less than or equal to 10, namely the number of hidden layer units h is less than or equal to 12;
step 4.2, setting a total of 12 training hidden layer unit numbers h 'in a mode of increasing 1 unit each time, wherein the 1 st training hidden layer unit number h' is 1; successively substituting the number h 'of the 12 training hidden layer units into the BP neural network A established in the step 4.1 to obtain 12 training hidden layers, and then training until the number h' of the training hidden layer units is 12; the process of one training is as follows:
will input into training set X train Inputting a BP neural network A comprising a training hidden layer for training to obtain a set of output predicted values
Wherein
To output a predicted value, a set of predicted values is output
And output training set Y train Keeping the dimensionality consistent and outputting a predicted value
And output training value y f Substituting the average absolute value error formula into a training error MAE, wherein the average absolute value error formula has the expression:
after 12 times of training is finished, recording the number h' of the training hidden layer units of the 12 th training as the number h of new hidden layer units new And new hidingNumber of layer units h new The corresponding training error MAE is recorded as the new error MAE new The minimum value of the first 11 training errors MAE is recorded as the minimum error MAE pre And will be equal to the minimum error MAE pre The corresponding unit number h' of the training hidden layer is recorded as the unit number h of the minimum error hidden layer pre ；
Determining the optimal hidden layer unit number h according to the adaptive regulation formula best The expression of the adaptive adjustment formula is as follows:
the optimum number h of hidden layer units best And (4) substituting the number of the hidden layer units into the BP neural network A obtained in the step 4.1, and finishing the construction of the three-layer BP neural network A.
Step 5, inputting the training set X obtained in the step 3 train And outputting a training set Y train And (4) adding the BP neural network A established in the step (4), training the weight and the threshold of the BP neural network A, finishing the final training of the BP neural network A, and recording as a BP neural network B.
Step 6, inputting the test set X obtained in the step 3 test And output test set Y test And (5) testing the lithium battery by adding the lithium battery into the BP neural network B trained in the step (5), and predicting the health state of the lithium battery.
Preferably, in the step 4, a three-layer BP neural network a is established and trained, and the target for training the BP neural network is 10 -6 The hidden layer activation function is poslin, the output layer activation function is purelin, the training function is trainlm, and the learning rate is 0.0001.
The lithium battery SOH prediction method based on the self-adaptive hidden layer BP neural network quickly and accurately realizes lithium battery SOH prediction under the condition of data set change, and has the advantages that:
1. the type of the battery and the number of data sets of the training network are not fixed, and the structure of the network can be changed in real time to adapt to the change of the training set.
2. Charging time is added as an input parameter, so that the accuracy of lithium battery SOH prediction is effectively improved.
3. The method has the advantages of reducing the calculation complexity, shortening the calculation time, improving the application range, and having high precision and strong tracking performance.
Drawings
FIG. 1 is a BP neural network structure diagram constructed in the lithium battery condition prediction method based on the adaptive hidden layer BP neural network of the present invention;
FIG. 2 is a flow chart of a lithium battery condition prediction method based on an adaptive hidden layer BP neural network according to the present invention;
fig. 3 is a graph comparing the prediction results of the lithium battery condition of the BP neural network without adaptive hidden layer setting under the same data set training according to the embodiment of the present invention.
Fig. 4 is a comparison graph of the prediction results of the lithium battery condition of the BP neural network without adaptive hidden layer setting under different battery data set training according to the embodiment of the present invention.
Detailed Description
The technical solution of the present invention is further described in detail below with reference to the accompanying drawings.
The invention relates to a lithium battery condition prediction method based on a self-adaptive hidden layer BP neural network, wherein the lithium battery condition is the lithium battery health condition, the flow of the lithium battery condition is shown in a figure 2, and the prediction method comprises the following steps:
step 1, cycling the charge and discharge of a lithium battery for n times, sampling the lithium battery in the charge and discharge cycles for n times, and forming a lithium battery charge and discharge data set by using sampling data;
extracting the following lithium battery sampling parameters related to the health condition of the lithium battery from the charging and discharging data set and recording the parameters as sampling health parameters: sampling end voltage in n times of charge and discharge cycles, sampling end current in n times of charge and discharge cycles, sampling lithium battery temperature in n times of charge and discharge cycles, sampling lithium battery charging time in n times of charge and discharge cycles and sampling lithium battery capacity in n times of charge and discharge cycles to obtain five groups of sampling health parameters, and marking any one group in the five groups as a sampling health parameter group j, wherein j is a sampling health parameter group j1,2,3,4, 5; any one of n charge-discharge cycles is recorded as a cycle i, i is 1, 2.. n; taking the sampling health parameter group j as a column vector, constructing a health parameter data matrix, and recording as a health matrix H n×5 The expression is as follows:
wherein, V i The sampled terminal voltage measured in cycle i; i is i The current of the sampling end measured in the cycle i is used as the current of the sampling end; t is i Sampling the temperature of the lithium battery measured in the cycle i; t is t i Sampling the lithium battery charging time measured in the cycle i; c i The lithium battery charge capacity measured in cycle i is sampled.
Step 2, carrying out normalization processing on the 5 groups of column vectors to obtain the following data after normalization processing:
the terminal voltages of the n lithium batteries after normalization processing are recorded as terminal voltages
n end currents after normalization processing, and recording the end currents in the circulation i as end currents
The temperature of the n normalized lithium batteries is recorded as the temperature of the lithium batteries in the cycle i
The charging time of the n normalized lithium batteries is recorded as the charging time of the lithium batteries in the cycle i
The charging capacities of the n lithium batteries after normalization processing are recorded as the charging capacity of the lithium battery in the cycle i
N terminal voltages
n end currents
Temperature of n lithium batteries
And n lithium battery charging times
Composing an input dataset X n×4 N lithium battery charging capacity
Form an output data set Y n×1 。
Step 3, inputting a data set X n×4 Divided into two groups, which are input training set X in turn train And input test set X test ；
Will output data set Y n×1 Divided into two groups, which are output training set Y in turn train And output test set Y test ；
Wherein, the proportion of the two training sets in the original data set is a given preset value gamma which is 50-90%;
set output training set Y train The total number of the m data is included, and any one of the m data is recorded as an output training value y f ，f＝1,2,...,m，m＜n。
And step 4, establishing a three-layer BP neural network A and training.
Step 4.1, the structure of the three-layer BP neural network A comprises an input layer, a hidden layer and an output layer; the input layer comprises 4 input layer cells and the output layer comprises only 1 output layer cell; the input layer and the hidden layer are in full connection, and the hidden layer and the output layer are in full connection;
the number of hidden layer units is determined by an empirical formula and an adaptive adjustment formula, wherein the empirical formula is used for estimating the maximum value of the number of hidden layer units, and the empirical formula is as follows:
h is the number of hidden layer units, h is an integer, r is the number of input layer units, s is the number of output layer units, a is a constant less than or equal to 10, namely the number of hidden layer units h is less than or equal to 12;
step 4.2, setting 12 training hidden layer units h 'in a mode of increasing 1 unit each time, wherein the 1 st training hidden layer unit h' is 1; successively substituting the number h 'of the 12 training hidden layer units into the BP neural network A established in the step 4.1 to obtain 12 training hidden layers, and then training until the number h' of the training hidden layer units is 12; the process of one training is as follows:
will input into training set X train Inputting a BP neural network A comprising a training hidden layer for training to obtain a set of output predicted values
Wherein
To output a predicted value, a set of predicted values is output
And output training set Y train Keeping the dimensionality consistent and outputting a predicted value
And output training value y f Substituting the average absolute value error formula into the average absolute value error formula to calculate to obtain a training error MAE, wherein the average absolute value error formula has the expression:
after 12 times of training is finished, recording the number h' of the training hidden layer units of the 12 th training as the number h of new hidden layer units new And the number of new hidden layer units h new The corresponding training error MAE is recorded as the new error MAE new The minimum value of the first 11 training errors MAE is recorded as the minimum error MAE pre And will be equal to the minimum error MAE pre The corresponding unit number h' of the training hidden layer is recorded as the unit number h of the minimum error hidden layer pre ；
Determining the optimal hidden layer unit number h according to the adaptive regulation formula best The expression of the adaptive adjustment formula is as follows:
the optimum number h of hidden layer units best And (4) substituting the number of the hidden layer units into the BP neural network A obtained in the step 4.1, and finishing the construction of the three-layer BP neural network A.
Step 5, inputting the training set X obtained in the step 3 train And outputting a training set Y train And (4) adding the BP neural network A established in the step (4), training the weight and the threshold of the BP neural network A, finishing the final training of the BP neural network A, and recording as a BP neural network B.
Step 6, inputting the test set X obtained in the step 3 test And output test set Y test And (5) testing the lithium battery by adding the lithium battery into the BP neural network B trained in the step (5), and predicting the health state of the lithium battery.
In this embodiment, in the step 4, a three-layer BP neural network a is established and trained, and the target for training the BP neural network is 10 -6 The hidden layer activation function is poslin, the output layer activation function is purelin, the training function is trainlm, and the learning rate is 0.0001.
In order to verify the effect of the present invention, simulation was performed.
Fig. 1 is a block diagram of a BP neural network constructed in accordance with the present invention.
In this embodiment, a three-layer BP neural network a composed of an input layer, a hidden layer, and an output layer is first constructed, where the input layer and the hidden layer are fully connected, and the hidden layer and the output layer are fully connected. The input layer comprises 4 input layer units, the output layer only comprises 1 output layer unit, and the hidden layer unit is obtained through training and testing.
In this embodiment, two different sets of battery data, denoted as data set a and data set B, are used for simulation verification.
FIG. 3 is a comparison of simulation results of the embodiment of the present invention in FIG. 1. In this simulation, the data is taken from dataset A, preset value γ 1 In this embodiment, the optimum number of hidden layer units h is 50% best1 7. The abscissa of the graph is the charge-discharge cycle of data set a, and the ordinate is the lithium battery capacity in the charge-discharge cycle corresponding to data set a. The actual value curve is a curve drawn according to original data of the data set A, the adaptive BP neural network curve is a curve drawn according to training and testing results of the prediction method, the neural network curve is a curve drawn according to the training and testing results of the traditional BP neural network, and the traditional BP neural network is a BP neural network which is not set by the adaptive hidden layer. As can be seen from fig. 3, the health condition of the lithium battery obtained by the prediction method provided by the present invention is closer to the actual value, i.e., the accuracy is higher.
FIG. 4 is a comparison of simulation results of the present invention embodiment with FIG. 2. In this simulation, the data is taken from data set B, preset value γ 2 In this embodiment, the optimum number of hidden layer units h is 50% best2 9. The abscissa of the graph is the charge-discharge cycle of the data set B, and the ordinate is the lithium battery capacity in the charge-discharge cycle corresponding to the data set B. The actual value curve is a curve drawn according to original data of a data set B, the self-adaptive BP neural network curve is a curve drawn according to the training and testing results of the prediction method, the BP neural network curve is a curve drawn according to the training and testing results of the traditional BP neural network, and the traditional BP neural network is set for a hidden layer which is not subjected to self-adaptationA BP neural network is arranged. As can be seen from fig. 4, the health condition of the lithium battery obtained by the prediction method provided by the present invention is closer to the actual value, i.e., the accuracy is higher.
Claims (2)
1. A lithium battery condition prediction method based on a self-adaptive hidden layer BP neural network is characterized by comprising the following steps of:
step 1, cycling the charge and discharge of a lithium battery for n times, sampling the lithium battery in the charge and discharge cycles for n times, and forming a lithium battery charge and discharge data set by using sampling data;
extracting the following lithium battery sampling parameters related to the health condition of the lithium battery from the charging and discharging data set and recording the parameters as sampling health parameters: sampling end voltage in n times of charge and discharge cycles, sampling end current in n times of charge and discharge cycles, sampling lithium battery temperature in n times of charge and discharge cycles, sampling lithium battery charging time in n times of charge and discharge cycles and sampling lithium battery capacity in n times of charge and discharge cycles to obtain five groups of sampling health parameters, and marking any one group in the five groups as a sampling health parameter group j, wherein j is 1,2,3,4, 5; any one of n charge-discharge cycles is recorded as a cycle i, i is 1, 2.. n; taking the sampling health parameter group j as a column vector, constructing a health parameter data matrix, and recording as a health matrix H n×5 The expression is as follows:
wherein, V i The sampled terminal voltage measured in cycle i; i is i The current of the sampling end measured in the cycle i is used as the current of the sampling end; t is i Sampling the temperature of the lithium battery measured in the cycle i; t is t i Sampling the lithium battery charging time measured in the cycle i; c i The lithium battery charge capacity measured in cycle i is sampled.
Step 2, carrying out normalization processing on the 5 groups of column vectors to obtain the following data after normalization processing:
n normalized lithiumAnd (3) battery terminal voltage, and recording the lithium battery terminal voltage in the cycle i as the terminal voltage
n end currents after normalization processing, and recording the end currents in the circulation i as end currents
The temperature of the n normalized lithium batteries is recorded as the temperature of the lithium batteries in the cycle i
The charging time of the n normalized lithium batteries is recorded as the charging time of the lithium batteries in the cycle i
The charging capacities of the n lithium batteries after normalization processing are recorded as the charging capacity of the lithium battery in the cycle i
N terminal voltages
n end currents
Temperature of n lithium batteries
And n lithium battery charging times
Composing an input dataset X n×4 N lithium battery charging capacity
Form an output data set Y n×1 。
Step 3, inputting a data set X n×4 Divided into two groups, which are input training set X in turn train And input test set X test ；
Will output data set Y n×1 Divided into two groups, which are output training set Y in turn train And output test set Y test ；
Wherein, the proportion of the two training sets in the original data set is a given preset value gamma which is 50-90%;
set output training set Y train The total number of the m data is included, and any one of the m data is recorded as an output training value y f ，f＝1,2,...,m，m＜n。
And step 4, establishing a three-layer BP neural network A and training.
Step 4.1, the structure of the three-layer BP neural network A comprises an input layer, a hidden layer and an output layer; the input layer comprises 4 input layer cells and the output layer comprises only 1 output layer cell; the input layer and the hidden layer are in full connection, and the hidden layer and the output layer are in full connection;
the number of hidden layer units is determined by an empirical formula and an adaptive adjustment formula, wherein the empirical formula is used for estimating the maximum value of the number of hidden layer units, and the empirical formula is as follows:
h is the number of hidden layer units, h is an integer, r is the number of input layer units, s is the number of output layer units, a is a constant less than or equal to 10, namely the number of hidden layer units h is less than or equal to 12;
step 4.2, setting a total of 12 training hidden layer unit numbers h 'in a mode of increasing 1 unit each time, wherein the 1 st training hidden layer unit number h' is 1; successively substituting the number h 'of the 12 training hidden layer units into the BP neural network A established in the step 4.1 to obtain 12 training hidden layers, and then training until the number h' of the training hidden layer units is 12; the process of one training is as follows:
will input into training set X train Inputting a BP neural network A comprising a training hidden layer for training to obtain a set of output predicted values
Wherein
To output a predicted value, a set of predicted values is output
And output training set Y train Keeping the dimensionality consistent and outputting a predicted value
And output training value y f Substituting the average absolute value error formula into a training error MAE, wherein the average absolute value error formula has the expression:
after 12 times of training is finished, recording the number h' of the training hidden layer units of the 12 th training as the number h of new hidden layer units new And the number h of new hidden layer units new The corresponding training error MAE is recorded as the new error MAE new Recording the minimum value of the first 11 training errors MAE as the minimum error MAE pre And will be equal to the minimum error MAE pre The corresponding unit number h' of the training hidden layer is recorded as the unit number h of the minimum error hidden layer pre ；
Determining the optimal hidden layer unit number h according to the adaptive regulation formula best Expression of said adaptive regulation formulaThe formula is as follows:
the optimum number h of hidden layer units best And (4) substituting the number of the hidden layer units into the BP neural network A obtained in the step 4.1, and finishing the construction of the three-layer BP neural network A.
Step 5, inputting the training set X obtained in the step 3 train And outputting a training set Y train And (4) adding the BP neural network A established in the step (4), training the weight and the threshold of the BP neural network A, finishing the final training of the BP neural network A, and recording as a BP neural network B.
Step 6, inputting the test set X obtained in the step 3 test And output test set Y test And (5) testing the lithium battery by adding the lithium battery into the BP neural network B trained in the step (5), and predicting the health state of the lithium battery.
2. The method for predicting the condition of the lithium battery based on the adaptive hidden layer BP neural network as claimed in claim 1, wherein the step 4 is to establish the three-layer BP neural network A and train the BP neural network with a target of 10 -6 The hidden layer activation function is poslin, the output layer activation function is purelin, the training function is trainlm, and the learning rate is 0.0001.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554879.9A CN115078999B (en) | 2022-05-20 | 2022-05-20 | Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554879.9A CN115078999B (en) | 2022-05-20 | 2022-05-20 | Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115078999A true CN115078999A (en) | 2022-09-20 |
| CN115078999B CN115078999B (en) | 2024-04-26 |
Family
ID=83249920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210554879.9A Active CN115078999B (en) | 2022-05-20 | 2022-05-20 | Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115078999B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112881914A (en) * | 2021-01-12 | 2021-06-01 | 常州大学 | Lithium battery health state prediction method |
| JPWO2022080377A1 (en) * | 2020-10-15 | 2022-04-21 |
-
2022
- 2022-05-20 CN CN202210554879.9A patent/CN115078999B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPWO2022080377A1 (en) * | 2020-10-15 | 2022-04-21 | ||
| CN112881914A (en) * | 2021-01-12 | 2021-06-01 | 常州大学 | Lithium battery health state prediction method |
Non-Patent Citations (2)
| Title |
|---|
| 封进;: "BP神经网络预估锂离子电池SOC训练数据选择", 电源技术, no. 02, 20 February 2016 (2016-02-20) * |
| 徐元中;曹翰林;吴铁洲;: "基于SA-BP神经网络算法的电池SOH预测", 电源技术, no. 03, 20 March 2020 (2020-03-20) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115078999B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tian et al. | Consistency evaluation and cluster analysis for lithium-ion battery pack in electric vehicles | |
| CN110045298B (en) | Method for diagnosing parameter inconsistency of power battery pack | |
| Yu et al. | OCV-SOC-temperature relationship construction and state of charge estimation for a series–parallel lithium-ion battery pack | |
| CN107741568B (en) | Lithium battery SOC estimation method based on state transition optimization RBF neural network | |
| CN110703101B (en) | Lithium ion battery inter-partition cycle capacity decline prediction method | |
| CN111007399A (en) | Lithium battery state of charge prediction method based on improved generation countermeasure network | |
| CN112630662B (en) | Power battery SOH estimation method based on data driving and multi-parameter fusion | |
| CN111458646A (en) | Lithium battery SOC estimation method based on PSO-RBF neural network | |
| CN112163372B (en) | SOC estimation method of power battery | |
| CN113128672B (en) | Lithium ion battery pack SOH estimation method based on transfer learning algorithm | |
| CN114839536B (en) | Lithium ion battery health state estimation method based on multiple health factors | |
| CN115639481B (en) | Battery data preprocessing system and method based on big data prediction SOC | |
| CN112881914B (en) | Lithium battery health state prediction method | |
| CN113807039A (en) | Power state prediction method of series battery system | |
| CN115308606B (en) | Lithium ion battery health state estimation method based on adjacent characteristics | |
| US20230349977A1 (en) | Method and apparatus for estimating state of health of battery | |
| CN115389946A (en) | Lithium battery health state estimation method based on isobaric rise energy and improved GRU | |
| CN116298936A (en) | Intelligent lithium ion battery health state prediction method in incomplete voltage range | |
| CN114545279A (en) | Lithium battery health state estimation method based on neural network ordinary differential equation | |
| CN114545275A (en) | Indirect prediction method for remaining service life of lithium ion battery | |
| CN113791351A (en) | Lithium battery life prediction method based on transfer learning and difference probability distribution | |
| CN115078999B (en) | Lithium battery condition prediction method based on self-adaptive hidden layer BP neural network | |
| CN116774045A (en) | Lithium battery health state prediction method based on HHO-SVR | |
| CN114062948B (en) | Lithium ion battery SOC estimation method based on 3DCNN | |
| CN114355218A (en) | Lithium ion battery charge state prediction method based on multi-feature quantity screening |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |