CN116840724A - Lithium ion battery health condition estimation and residual service life prediction method - Google Patents
Lithium ion battery health condition estimation and residual service life prediction method Download PDFInfo
- Publication number
- CN116840724A CN116840724A CN202310888955.4A CN202310888955A CN116840724A CN 116840724 A CN116840724 A CN 116840724A CN 202310888955 A CN202310888955 A CN 202310888955A CN 116840724 A CN116840724 A CN 116840724A
- Authority
- CN
- China
- Prior art keywords
- health condition
- health
- lithium ion
- ion battery
- model
- 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
- 230000036541 health Effects 0.000 title claims abstract description 129
- 238000000034 method Methods 0.000 title claims abstract description 36
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 title claims abstract description 32
- 229910001416 lithium ion Inorganic materials 0.000 title claims abstract description 32
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 238000007637 random forest analysis Methods 0.000 claims abstract description 22
- 125000004122 cyclic group Chemical group 0.000 claims abstract description 14
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 11
- 238000000556 factor analysis Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 230000006870 function Effects 0.000 abstract description 35
- 238000013135 deep learning Methods 0.000 abstract description 3
- 238000010801 machine learning Methods 0.000 abstract description 3
- 238000011156 evaluation Methods 0.000 description 14
- 230000003862 health status Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 3
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 3
- 238000007500 overflow downdraw method Methods 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
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]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- 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
- 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]
- G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a lithium ion battery health condition estimation and residual service life prediction method, which comprises the following steps: extracting health indexes to perform classical modal decomposition on the signals with known health conditions; obtaining the correlation between the intrinsic mode function component set, the residual signal, the health index data and the health condition by using the Pearson coefficient; obtaining a variance contribution rate through a factor analysis method; respectively inputting the health index data into a GRU cyclic neural network model and an RF random forest model for processing to obtain a final health condition estimated value; and respectively carrying out parameter updating on the two models according to the final health condition estimated value, and inputting the health index data of the lithium ion battery to be tested into the two models with the updated parameters to complete prediction. According to the invention, the health condition signals are decomposed through classical mode decomposition, so that the data prediction accuracy is improved; by introducing variance contribution ratio and establishing a prediction model by utilizing deep learning and machine learning, the calculated amount is reduced and the prediction accuracy is improved.
Description
Technical Field
The invention relates to the technical field of lithium ion batteries, in particular to a lithium ion battery health condition estimation and residual service life prediction method.
Background
The lithium ion battery has the advantages of high energy density, high voltage, small self-discharge, high environmental protection and the like, and is widely applied to the fields of new energy automobiles, power grid energy storage and the like. However, with the deep application of the lithium ion battery, potential safety hazards such as spontaneous combustion and explosion of the lithium ion battery are increasingly displayed. The accurate estimation of the health state of the lithium ion battery is an important measure for guaranteeing the safe and stable operation of the battery, and is also a key place for the health management and intelligent operation and maintenance of the lithium ion battery.
Common methods for estimating SOH of lithium ion batteries include model-based methods, data-driven methods, and fusion methods. The model-based method has complex parameters, is difficult to identify the parameters and takes time for measurement; the method based on data driving utilizes a single model, has certain limitation and has low prediction precision; although the fusion method has good prediction effect, the fusion method has complex parameters, large calculated amount and high time cost.
Disclosure of Invention
Aiming at the defects in the prior art, the method for estimating the health condition and predicting the residual service life of the lithium ion battery solves the problems of low prediction precision, complex calculation and high time cost in the prior art.
In order to achieve the aim of the invention, the invention adopts the following technical scheme:
the method for estimating the health condition and predicting the residual service life of the lithium ion battery comprises the following steps:
s1, taking the charge equal-pressure difference time and the discharge voltage of the battery as the lowest point time as health indexes; performing classical modal decomposition on the known health condition signals of the lithium ion battery to obtain an initial intrinsic modal function component set and an initial residual signal;
s2, respectively analyzing the health condition signal and the health index, the health condition signal and the initial intrinsic mode function set and the health condition signal and the initial residual error signal through the Pearson coefficient to respectively obtain corresponding correlations; the initial natural mode function component, the related data of the health index and the initial residual signal with the correlation not lower than the threshold value are reserved, and a final natural mode function component set, health index data and residual signals are obtained;
s3, obtaining a corresponding variance contribution rate through a factor analysis method according to the correlation between the health condition signal and the initial natural mode function component set;
s4, building a GRU cyclic neural network model and an RF random forest model; inputting the health index data into a GRU (generalized particle swarm optimization) cyclic neural network model to obtain a first health condition estimation set; inputting the health index data into the RF random forest model to obtain a second health condition estimated value;
s5, multiplying the first health condition estimation set and the variance contribution rate to obtain a third health condition estimation set; adding all the health condition estimated values of the second health condition estimated value and the third health condition estimated set to obtain a final health condition estimated value;
s6, carrying out parameter updating on the GRU circulating neural network model by a gradient descent method according to the final health condition estimated value to obtain an updated GRU circulating neural network model;
s7, updating parameters of the RF random forest model through a grid search method according to the final health condition estimated value to obtain an updated RF random forest model;
and S8, inputting the health index data of the lithium ion battery to be tested into the GRU cyclic neural network model and the RF random forest model after parameter updating, and completing the estimation of the health condition and the prediction of the residual service life of the lithium ion battery.
Further, the specific method of classical modal decomposition in step S1 is:
s1-1, acquiring all limit value points of a health condition signal; connecting the maximum value point and the minimum value point into an upper envelope line and an upper envelope line respectively by using a cubic spline functionThe lower envelope curve is used for solving the average value of the upper envelope curve and the lower envelope curve; the average value of the upper envelope line and the lower envelope line is differenced with the health condition signal to obtain a new health condition signal h 1(t) ;
S1-2, repeating the step S1-1 until the current health condition signal meets the condition that the number of extreme points and the number of zero crossing points are equal or differ by 1, or the average value of two envelope lines of a local maximum point and a local minimum point is 0, and taking the health condition signal h n(t) A first component that is a function of the natural mode; the health signal and the first component h n(t) Making a difference to obtain a new health condition signal;
s1-3, repeating the step S1-2 until the intrinsic mode function component is small enough or is a monotonic function, and ending the operation to obtain an intrinsic mode function component set and a residual signal.
Further, the correlation threshold of the initial natural mode function component in step S2 is 0.9; the absolute value of the correlation threshold of the correlation data of the health index is 0.05; the correlation threshold of the initial residual signal is 0.99.
Further, the GRU cyclic neural network model in the step S6 adopts a mean square error as a loss function; the RF random forest model in step S7 uses the mean square error as the loss function.
Further, the GRU recurrent neural network model introduces an update gate and a reset gate, the formulas of which are as follows:
z t =σ(W z x t +U z h t-1 +b z )
r t =σ(W r x t +U r h t-1 +b r )
wherein x is t Is the data input at time t, namely health index and inherentA set of modal function components; z t 、r t Representing an update gate and a reset gate, respectively;representing candidate states at time t; h is a t The output of the hidden layer at the moment t is represented; h is a t-1 The output of the hidden layer at the moment t-1 is represented; w (W) z 、W r And W is h Representing the weight parameters; b r 、b h And b z Is a bias parameter; u (U) z 、U r And U h Representing the weight parameters; sigma (·) is a sigmoid activation function; tan h (·) is the hyperbolic tangent activation function.
Further, the specific method of the RF random forest model is as follows:
generating a random vector sequence according to the input data; randomly sampling input data by using a Bootstrap sampling method to obtain k sub-sample sets; respectively establishing a regression model for each sub-sample set; performing k-round training on the regression model to obtain a regression tree model sequence;
for any new sample, the predicted result is the average summary of the results of k rounds of training, and the formula is as follows:
wherein x represents an input sample, f (x) represents a second health condition estimate, h i (x) Representing the results of the ith regression tree model.
The beneficial effects of the invention are as follows: according to the prediction method, the health condition signals are decomposed through classical mode decomposition, so that non-stationary data are stabilized, and the data prediction precision is improved; the inherent mode function components are measured by introducing the variance contribution ratio, and the prediction model is built by utilizing deep learning and machine learning, so that the calculated amount is reduced, and the prediction accuracy is improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a block diagram of a GRU recurrent neural network model of the invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
As shown in fig. 1, a method for estimating health status and predicting remaining service life of a lithium ion battery includes the steps of:
s1, taking the charge equal-pressure difference time and the discharge voltage of the battery as the lowest point time as health indexes; performing classical modal decomposition on the known health condition signals of the lithium ion battery to obtain an initial intrinsic modal function component set and an initial residual signal;
s2, respectively analyzing the health condition signal and the health index, the health condition signal and the initial intrinsic mode function set and the health condition signal and the initial residual error signal through the Pearson coefficient to respectively obtain corresponding correlations; the initial natural mode function component, the related data of the health index and the initial residual signal with the correlation not lower than the threshold value are reserved, and a final natural mode function component set, health index data and residual signals are obtained;
s3, obtaining a corresponding variance contribution rate through a factor analysis method according to the correlation between the health condition signal and the initial natural mode function component set;
s4, building a GRU cyclic neural network model and an RF random forest model; inputting the health index data into a GRU (generalized particle swarm optimization) cyclic neural network model to obtain a first health condition estimation set; inputting the health index data into the RF random forest model to obtain a second health condition estimated value;
s5, multiplying the first health condition estimation set and the variance contribution rate to obtain a third health condition estimation set; adding all the health condition estimated values of the second health condition estimated value and the third health condition estimated set to obtain a final health condition estimated value;
s6, carrying out parameter updating on the GRU circulating neural network model by a gradient descent method according to the final health condition estimated value to obtain an updated GRU circulating neural network model;
s7, updating parameters of the RF random forest model through a grid search method according to the final health condition estimated value to obtain an updated RF random forest model;
and S8, inputting the health index data of the lithium ion battery to be tested into the GRU cyclic neural network model and the RF random forest model after parameter updating, and completing the estimation of the health condition and the prediction of the residual service life of the lithium ion battery.
The specific method of classical modal decomposition in step S1 is:
s1-1, acquiring all limit value points of a health condition signal; connecting the maximum value point and the minimum value point into an upper envelope curve and a lower envelope curve by using a cubic spline function respectively, and solving the average value of the upper envelope curve and the lower envelope curve; the average value of the upper envelope line and the lower envelope line is differenced with the health condition signal to obtain a new health condition signal h 1(t) ;
S1-2, repeating the step S1-1 until the current health condition signal meets the condition that the number of extreme points and the number of zero crossing points are equal or differ by 1, or the average value of two envelope lines of a local maximum point and a local minimum point is 0, and taking the health condition signal h n(t) A first component that is a function of the natural mode; the health signal and the first component h n(t) Making a difference to obtain a new health condition signal;
s1-3, repeating the step S1-2 until the intrinsic mode function component is small enough or is a monotonic function, and ending the operation to obtain an intrinsic mode function component set and a residual signal.
The correlation threshold of the initial natural mode function component in the step S2 is 0.9; the absolute value of the correlation threshold of the correlation data of the health index is 0.05; the correlation threshold of the initial residual signal is 0.99.
The GRU cyclic neural network model in the step S6 adopts the mean square error as a loss function; the RF random forest model in step S7 uses the mean square error as the loss function.
As shown in fig. 2, the GRU recurrent neural network model introduces an update gate and a reset gate, and its formula is as follows:
z t =σ(W z x t +U z h t-1 +b z )
r t =σ(W r x t +U r h t-1 +b r )
wherein x is t The data input at the moment t, namely health index and intrinsic mode function component set; z t 、r t Representing an update gate and a reset gate, respectively;representing candidate states at time t; h is a t The output of the hidden layer at the moment t is represented; h is a t-1 The output of the hidden layer at the moment t-1 is represented; w (W) z 、W r And W is h Representing the weight parameters; b r 、b h And b z Is a bias parameter; u (U) z 、U r And U h Representing the weight parameters; sigma (·) is a sigmoid activation function; tan h (·) is the hyperbolic tangent activation function.
The specific method of the RF random forest model is as follows:
generating a random vector sequence according to the input data; randomly sampling input data by using a Bootstrap sampling method to obtain k sub-sample sets; respectively establishing a regression model for each sub-sample set; performing k-round training on the regression model to obtain a regression tree model sequence;
for any new sample, the predicted result is the average summary of the results of k rounds of training, and the formula is as follows:
wherein x represents an input sample, f (x) represents a second health condition estimate, h i (x) Representing the results of the ith regression tree model.
In one embodiment of the invention, the pearson correlation coefficient is shown as follows:
where Pearson represents the correlation, X, Y represents the correlation variable, and E (-) represents the desire. The related variables include a health status signal and a health index, a health status signal and an initial set of natural mode functions, and a health status signal and an initial residual signal.
The invention adopts average absolute error MAE, root mean square error RMSE, average absolute percentage error MAPE and determination coefficient R 2 As an evaluation index of the prediction result. The smaller the average absolute error MAE, the root mean square error RMSE and the average absolute percentage error MAPE, the better the fitting effect of the prediction model is, and the better the accuracy is; determining the coefficient R 2 The closer to 1 the value of (c) is, the better the predictive model fitting effect is.
The data of the step S2 and the step S3 are divided into a training set and a testing set according to the proportion of 8:2 in the experiment; the residual signal ratio is set to 100%; lithium ion batteries numbered B0005, B0006, B0007 and B0018 are adopted to enter different prediction models for prediction. The prediction model comprises a prediction model, a GRU prediction model, an RF prediction model, an EMD-RF prediction model and an EMD-GRU prediction model of the invention; the prediction model of the invention is an EMD-VCR-GRU-RF prediction model, EMD is classical modal decomposition, and VCR is variance contribution rate.
Table 1 shows evaluation indexes of health condition estimation values of the RF prediction model and the GRU prediction model.
Table 2 shows evaluation indexes of health status evaluation values of the EMD-RF prediction model and the EMD-GRU prediction model.
Table 3 shows the evaluation index of the health condition estimation value of the prediction model of the present invention.
Table 1 evaluation index of health status evaluation values of RF prediction model and GRU prediction model
TABLE 2 evaluation index of health evaluation values of EMD-RF prediction model and EMD-GRU prediction model
TABLE 3 evaluation index of health evaluation value of predictive model of the invention
As can be seen from tables 1, 2 and 3, the average absolute error MAE, the root mean square error RMSE and the average absolute percentage error MAPE of the prediction model of the present invention are the smallest, and the coefficient R is determined by comparing the GRU prediction model, the RF prediction model, the EMD-RF prediction model and the EMD-GRU prediction model 2 The value of (2) is closest to 1, which indicates that the fitting effect of the prediction model of the present invention is better than the other four prediction models.
Regarding the residual service life of the lithium ion battery, a new evaluation index, namely an integral prediction error AE, is added in the experiment; its prediction starting points are set to 60 and 80, respectively. Table 4 shows the evaluation index of the remaining life of the prediction model of the present invention.
TABLE 4 evaluation index of residual service Life of the prediction model of the invention
As can be seen from table 4, the prediction error of 80 is smaller than the prediction error of 60. This means that the training data set is increased, contributing to an improved prediction accuracy. The AE overall prediction error is smaller, which indicates that the robustness and adaptability of the proposed model are better.
In conclusion, the health condition signals are decomposed through classical mode decomposition, so that non-stationary data are stabilized, and the data prediction precision is improved; the inherent mode function components are measured by introducing the variance contribution ratio, and the prediction model is built by utilizing deep learning and machine learning, so that the calculated amount is reduced, and the prediction accuracy is improved.
Claims (6)
1. A lithium ion battery health condition estimation and residual service life prediction method is characterized in that: the method comprises the following steps:
s1, taking the charge equal-pressure difference time and the discharge voltage of the battery as the lowest point time as health indexes; performing classical modal decomposition on the known health condition signals of the lithium ion battery to obtain an initial intrinsic modal function component set and an initial residual signal;
s2, respectively analyzing the health condition signal and the health index, the health condition signal and the initial intrinsic mode function set and the health condition signal and the initial residual error signal through the Pearson coefficient to respectively obtain corresponding correlations; the initial natural mode function component, the related data of the health index and the initial residual signal with the correlation not lower than the threshold value are reserved, and a final natural mode function component set, health index data and residual signals are obtained;
s3, obtaining a corresponding variance contribution rate through a factor analysis method according to the correlation between the health condition signal and the initial natural mode function component set;
s4, building a GRU cyclic neural network model and an RF random forest model; inputting the health index data into a GRU (generalized particle swarm optimization) cyclic neural network model to obtain a first health condition estimation set; inputting the health index data into the RF random forest model to obtain a second health condition estimated value;
s5, multiplying the first health condition estimation set and the variance contribution rate to obtain a third health condition estimation set; adding all the health condition estimated values of the second health condition estimated value and the third health condition estimated set to obtain a final health condition estimated value;
s6, carrying out parameter updating on the GRU circulating neural network model by a gradient descent method according to the final health condition estimated value to obtain an updated GRU circulating neural network model;
s7, updating parameters of the RF random forest model through a grid search method according to the final health condition estimated value to obtain an updated RF random forest model;
and S8, inputting the health index data of the lithium ion battery to be tested into the GRU cyclic neural network model and the RF random forest model after parameter updating, and completing the estimation of the health condition and the prediction of the residual service life of the lithium ion battery.
2. The method for estimating health and predicting remaining life of a lithium ion battery according to claim 1, wherein: the specific method of classical modal decomposition in the step S1 is as follows:
s1-1, acquiring all limit value points of a health condition signal; connecting the maximum value point and the minimum value point into an upper envelope curve and a lower envelope curve by using a cubic spline function respectively, and solving the average value of the upper envelope curve and the lower envelope curve; the average value of the upper envelope line and the lower envelope line is differenced with the health condition signal to obtain a new health condition signal h 1(t) ;
S1-2, repeating the step S1-1 until the current health condition signal meets the condition that the number of extreme points and the number of zero crossing points are equal or differ by 1, or the average value of two envelope lines of a local maximum point and a local minimum point is 0, and taking the health condition signal h n(t) A first component that is a function of the natural mode; the health signal and the first component h n(t) Making a difference to obtain a new health condition signal;
s1-3, repeating the step S1-2 until the intrinsic mode function component is small enough or is a monotonic function, and ending the operation to obtain an intrinsic mode function component set and a residual signal.
3. The method for estimating health and predicting remaining life of a lithium ion battery according to claim 1, wherein: the correlation threshold of the initial natural mode function component in the step S2 is 0.9; the absolute value of the correlation threshold of the correlation data of the health index is 0.05; the correlation threshold of the initial residual signal is 0.99.
4. The method for estimating health and predicting remaining life of a lithium ion battery according to claim 1, wherein: the GRU cyclic neural network model in the step S6 adopts a mean square error as a loss function; the RF random forest model in step S7 uses the mean square error as the loss function.
5. The method for estimating health and predicting remaining life of a lithium ion battery according to claim 1, wherein: the GRU cyclic neural network model introduces an update gate and a reset gate, and the formula is as follows:
z t =σ(W z x t +U z h t-1 +b z )
r t =σ(W r x t +U r h t-1 +b r )
wherein x is t The data input at the moment t, namely health index and intrinsic mode function component set; z t 、r t Representing an update gate and a reset gate, respectively;representing candidate states at time t; h is a t The output of the hidden layer at the moment t is represented; h is a t-1 The output of the hidden layer at the moment t-1 is represented; w (W) z 、W r And W is h Representing the weight parameters; b r 、b h And b z Is a bias parameter; u (U) z 、U r And U h Representing the weight parameters; sigma (·) is a sigmoid activation function; tan h (·) is the hyperbolic tangent activation function.
6. The method for estimating health and predicting remaining life of a lithium ion battery according to claim 1, wherein: the specific method of the RF random forest model is as follows:
generating a random vector sequence according to the input data; randomly sampling input data by using a Bootstrap sampling method to obtain k sub-sample sets; respectively establishing a regression model for each sub-sample set; performing k-round training on the regression model to obtain a regression tree model sequence;
for any new sample, the predicted result is the average summary of the results of k rounds of training, and the formula is as follows:
wherein x represents an input sample, f (x) represents a second health condition estimate, h i (x) Representing the results of the ith regression tree model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310888955.4A CN116840724A (en) | 2023-07-18 | 2023-07-18 | Lithium ion battery health condition estimation and residual service life prediction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310888955.4A CN116840724A (en) | 2023-07-18 | 2023-07-18 | Lithium ion battery health condition estimation and residual service life prediction method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116840724A true CN116840724A (en) | 2023-10-03 |
Family
ID=88168797
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310888955.4A Pending CN116840724A (en) | 2023-07-18 | 2023-07-18 | Lithium ion battery health condition estimation and residual service life prediction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116840724A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117805662A (en) * | 2024-02-26 | 2024-04-02 | 深圳市拓湃新能源科技有限公司 | Battery parameter processing method, device, equipment and storage medium |
-
2023
- 2023-07-18 CN CN202310888955.4A patent/CN116840724A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117805662A (en) * | 2024-02-26 | 2024-04-02 | 深圳市拓湃新能源科技有限公司 | Battery parameter processing method, device, equipment and storage medium |
CN117805662B (en) * | 2024-02-26 | 2024-06-11 | 深圳市拓湃新能源科技有限公司 | Battery parameter processing method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yang et al. | Lifespan prediction of lithium-ion batteries based on various extracted features and gradient boosting regression tree model | |
Lin et al. | State of health estimation of lithium-ion battery based on an adaptive tunable hybrid radial basis function network | |
CN110398697B (en) | Lithium ion health state estimation method based on charging process | |
CN113805064B (en) | Lithium ion battery pack health state prediction method based on deep learning | |
CN110888059B (en) | Charge state estimation algorithm based on improved random forest combined volume Kalman | |
CN113702843B (en) | Lithium battery parameter identification and SOC estimation method based on suburb optimization algorithm | |
CN111090047A (en) | Lithium battery health state estimation method based on multi-model fusion | |
CN111856287B (en) | Lithium battery health state detection method based on stacked residual causal convolutional neural network | |
Li et al. | On-line estimation method of lithium-ion battery health status based on PSO-SVM | |
Mamo et al. | Long short-term memory with attention mechanism for state of charge estimation of lithium-ion batteries | |
CN109932656A (en) | A kind of service life of lithium battery estimation method based on IMM-UPF | |
CN114397577A (en) | New energy automobile lithium battery health state assessment method based on ASTUKF-GRA-LSTM model | |
CN115201686B (en) | Lithium ion battery health state assessment method under incomplete charge and discharge data | |
CN113917334A (en) | Battery health state estimation method based on evolution LSTM self-encoder | |
Zhao et al. | State-of-health estimation with anomalous aging indicator detection of lithium-ion batteries using regression generative adversarial network | |
CN116840724A (en) | Lithium ion battery health condition estimation and residual service life prediction method | |
CN114966436A (en) | Lithium battery state of charge prediction method, device, equipment and readable storage medium | |
CN115792680A (en) | Retired battery health state assessment method based on Bayes deep learning | |
CN114839538A (en) | Method for extracting degradation characteristics of lithium ion battery for estimating residual life | |
Bak et al. | Accurate estimation of battery SOH and RUL based on a progressive lstm with a time compensated entropy index | |
CN115308608A (en) | All-vanadium redox flow battery voltage prediction method, device and medium | |
CN112327169B (en) | Lithium battery residual life prediction method | |
Hatherall et al. | Remaining discharge energy estimation for lithium-ion batteries using pattern recognition and power prediction | |
Lyu et al. | State-of-charge estimation of lithium-ion batteries based on deep neural network | |
CN116908727A (en) | Multi-feature combination-based energy storage power station lithium battery capacity estimation method |
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 |