CN116774088A - Lithium ion battery health state estimation method based on multi-objective optimization - Google Patents
Lithium ion battery health state estimation method based on multi-objective optimization Download PDFInfo
- Publication number
- CN116774088A CN116774088A CN202310747259.1A CN202310747259A CN116774088A CN 116774088 A CN116774088 A CN 116774088A CN 202310747259 A CN202310747259 A CN 202310747259A CN 116774088 A CN116774088 A CN 116774088A
- Authority
- CN
- China
- Prior art keywords
- health state
- health
- lithium ion
- model
- state estimation
- 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 96
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000005457 optimization Methods 0.000 title claims abstract description 32
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 title claims abstract description 29
- 229910001416 lithium ion Inorganic materials 0.000 title claims abstract description 29
- 238000010277 constant-current charging Methods 0.000 claims abstract description 19
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 7
- 238000013528 artificial neural network Methods 0.000 claims abstract description 5
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 241000283153 Cetacea Species 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 5
- 230000008569 process Effects 0.000 abstract description 14
- 230000032683 aging Effects 0.000 abstract description 10
- 230000015556 catabolic process Effects 0.000 abstract description 6
- 238000006731 degradation reaction Methods 0.000 abstract description 6
- 238000012163 sequencing technique Methods 0.000 abstract 1
- 238000007600 charging Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 7
- 238000013178 mathematical model Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 229910052744 lithium Inorganic materials 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 229910012851 LiCoO 2 Inorganic materials 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000007774 positive electrode material Substances 0.000 description 1
- 238000005295 random walk Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000005070 sampling 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The application relates to a lithium ion battery health state estimation method based on multi-objective optimization, which comprises the steps of firstly collecting charge and discharge cycle data of a lithium ion battery and preprocessing; then, extracting a plurality of health features including constant-current charging time and polynomial coefficients fitting a constant-current charging voltage-time curve by using a Logit model from the charge-discharge cycle data; and finally, establishing a health state estimation model based on the neural network of the extreme learning machine, training the health state estimation model by using lithium ion battery charge-discharge cycle data, optimizing the health state estimation model by using a multi-objective optimization algorithm based on non-dominant sequencing and crowding distances in the training process, and using the optimized health state estimation model for estimating the health state of the lithium ion battery. The polynomial coefficient extracted by the application reflects the nonlinear aging rule in the constant current charging process, indirectly represents the battery degradation, directly represents the battery degradation by the constant current charging time, and combines the prediction precision and stability of the model by model optimization.
Description
Technical Field
The application belongs to the technical field of health state estimation of battery management systems, and particularly relates to a lithium ion battery health state estimation method based on multi-objective optimization.
Background
The lithium ion battery is widely applied in the field of electric automobiles because of the advantages of high energy density, large output capacity, high cost performance and the like. In the process of recycling the battery, the internal structure of the battery can generate phenomena of recyclable lithium loss, collector corrosion and the like, and the battery performance can be continuously degraded. In order to ensure safe and stable operation of the electric automobile, the power battery is managed and controlled on line through a battery management system. The State of health (SOH) of the battery reflects the capacity of the current battery relative to the new battery, and the capacity level of the battery can be known in advance by accurately predicting the State of health of the battery, which has important significance for optimizing the control strategy of the electric automobile, improving the energy utilization rate, reducing the energy consumption cost and the like.
At present, the commonly used lithium ion battery health state assessment method is divided into a model-based method and a data driving-based method, and the physical process of the lithium ion battery is described by establishing a mathematical model based on the model method, but the method has the defects of high model complexity, difficult parameter determination, poor applicability and the like. The method is based on a data driving method, a model is built by means of historical data of the battery, machine learning and the like, and then the health state of the battery is predicted, and compared with the method based on the model, the method has better prediction accuracy and robustness, but the method has the following problems that firstly, in the aspect of health feature extraction, most of current researches are conducted on feature expansion such as battery capacity increment, equal pressure rise time, equal pressure fall time and average voltage, but the characteristics cannot reflect nonlinear aging rules of the battery in the constant current charging process, and the degradation characteristics of the health state of the lithium ion battery cannot be accurately represented. Secondly, in the aspect of a data driving model, most of the application of the traditional single-target optimization model in the aspect of health state evaluation only focuses on improving the prediction precision, but capacity rebound phenomenon can occur in the process of recycling batteries, fluctuation generated by capacity rebound cannot be accurately tracked, and further the prediction stability of the model is poor, so that the influence of capacity rebound on the prediction stability is ignored by the traditional single-target optimization model.
Therefore, the application provides a lithium ion battery health state estimation method based on multi-objective optimization, which is characterized in that on one hand, health features reflecting nonlinear aging in the constant current charging process are extracted, and on the other hand, the model is enabled to pay attention to the accuracy and stability of a prediction result at the same time.
Disclosure of Invention
Aiming at the defects existing in the prior art, the application aims to provide a lithium ion battery health state estimation method based on multi-objective optimization.
The application solves the technical problems by adopting the following technical scheme:
the lithium ion battery health state estimation method based on multi-objective optimization is characterized by comprising the following steps of:
step S1: collecting charge-discharge cycle data of a lithium ion battery;
step S2: preprocessing charge-discharge cycle data, extracting a plurality of health features from the charge-discharge cycle data, wherein one health feature is constant-current charging time, and the other health features are polynomial coefficients for fitting a constant-current charging voltage-time curve by using a Logit model;
step S3: establishing a health state estimation model based on the extreme learning machine neural network;
step S4: training a health state estimation model by using collected lithium ion battery charge-discharge cycle data, wherein an input feature matrix of the model is F= [ F ] 1 ;F 2 ;…;F n ;…;F N ]The output vector isInput feature vector representing the nth cycle, < ->Representing healthy features extracted from the nth cycle, < + >>Representing the nth cycleThe true health state of the ring, N represents the number of charge and discharge cycles of the full life cycle;
taking two parameters of input weight from an input layer to a hidden layer and bias of the hidden layer of the health state estimation model as parameters to be optimized, and optimizing the health state estimation model by using a multi-objective optimization algorithm based on non-dominant sorting and crowding distance; the fitness function of the multi-objective optimization algorithm comprises a mean square error and a stability index, and the formula is as follows:
wherein MSE is mean square error, SDEX is stability index, K is sample number, y i 、The real health state and the predicted health state of the ith sample are respectively, and std (·) represents the standard deviation;
step S5: and using the optimized health state estimation model for estimating the health state of the lithium ion battery.
Compared with the prior art, the application has the beneficial effects that:
1. according to the application, a voltage-time curve in the constant current charging process is fitted by using the Logit model, polynomial coefficients of the Logit model are extracted as health features, and compared with traditional health features such as capacity increment, equal pressure rise time, equal pressure fall time and average voltage, the polynomial coefficients can fully reflect nonlinear aging rules in the constant current charging process of the battery, can indirectly represent the degradation of the lithium ion battery, and meanwhile, the constant current charging time directly representing the degradation of the lithium ion battery is also taken as health features, so that SOH estimation accuracy is improved.
2. The health state estimation model is optimized by adopting a multi-objective whale optimization algorithm based on non-dominant sorting and crowding distance calculation, and the mean square error and the stability index (SDEX) are used as objective functions of multi-objective optimization, so that the prediction precision and stability of the model are considered, the problem that the conventional single-objective optimization model only considers the prediction precision and ignores the prediction stability caused by battery capacity rising is effectively solved, and relatively high prediction precision and stability can be achieved in the health state estimation of the whole life cycle of the lithium ion battery.
Drawings
FIG. 1 is an overall flow chart of the present application;
FIG. 2 is a graph of SOH decay for three cells in a Maryland dataset;
FIG. 3 is a charge voltage versus time graph;
FIG. 4 is a graph of the third order fitting accuracy variation of the CS35 battery full life cycle voltage curve;
FIG. 5 is a graph of normalized four health features versus cycle number;
FIG. 6 is a block diagram of an ELM neural network;
FIG. 7 (a) is a graph of state of health estimation for a CS37 battery for different models;
FIG. 7 (b) is a graph of relative error versus cycle number for different models;
FIG. 8 (a) is a relative error histogram of the PSO-ELM model;
FIG. 8 (b) is a relative error histogram of the WOMOA-ELM model.
Detailed Description
The following specific embodiments are given by way of illustration only and not by way of limitation of the scope of the application.
The application discloses a lithium ion battery health state estimation method based on multi-objective optimization, which comprises the following steps:
step S1: collecting charge and discharge cycle data in the aging test process of the lithium ion battery, and constructing a data set;
taking the maryland university aging data set as an example, the data set comprises aging test data of three batteries, wherein the battery types are CS 2-35 (CS 35), CS 2-36 (CS 36) and CS 2-37 (CS 37), the positive electrode materials of the batteries are lithium cobaltate (LiCoO 2), and the nominal capacities of the batteries are all 1.1Ah; the charge and discharge standards of the battery aging test are as follows: under the condition of 24 ℃ at room temperature, taking 3.8V as constant current charging initial voltage, and carrying out constant current charging at 0.5C until reaching charging cut-off voltage of 4.2V; subsequently, charging was continued in a constant voltage mode until the current dropped to 50mA; finally, discharging to a discharge cut-off voltage of 2.7V by using a constant current of 1C;
step S2: data preprocessing and health feature extraction;
s2.1: removing noise to finish data preprocessing;
s2.2: SOH is defined as the full discharge capacity as shown in formula (1); each battery retains charge-discharge cycle data from 100% soh down to 70% soh interval, three batteries retain 641, 635 and 708 charge-discharge cycle data respectively, and the health state decay curves of the whole life cycle of the three batteries are shown in fig. 2;
in SOH n SOH value for the nth cycle, C n For the capacity value of the nth cycle, C 0 Is the rated capacity of the battery;
fig. 3 is a charge voltage-time curve under constant current charge conditions, and the constant current charge time gradually shortens with increasing cycle times, so that the constant current charge time reflects battery degradation, and the constant current charge time is taken as a health feature and is denoted as HF1, and the expression is:
ΔT=T (4.2V) -T (3.8V) (2)
wherein T is (4.2V) 、T (3.8V) Respectively representing the charging time when the battery is charged to the constant-current charging initial voltage of 3.8V and the charging cut-off voltage of 4.2V;
s2.3: as can be seen from fig. 3, the charging voltage-time curve under the constant-current charging condition is nonlinear, and gradually shifts leftwards as the cycle number increases, and the charging voltage-time curve under the constant-current charging condition contains health characteristics representing aging of the battery; performing curve fitting on voltage and charging time in the constant current charging process by using a Logit model of a formula (3), wherein polynomial coefficients of the fitted curve can reflect nonlinear aging rules of the battery, so that the polynomial coefficients of the fitted curve are regarded as health features and are used for estimating the health state of the battery;
in the formula, U represents a voltage,representing polynomial coefficients, T representing charging time, c representing intercept, n representing number of cycles, and m representing fitting order;
selecting a decision coefficient (R 2 ) And the Root Mean Square Error (RMSE) are used for evaluating the precision of different fitting orders of the Logit model, and the calculation formulas are shown in formulas (4) and (5);
wherein U is j 、Represents the j-th real voltage and the fitting voltage, respectively, < >>The average value of the real voltage is represented, and S is the number of sampling points;
the result shows that when the fitting order is more than or equal to 3, R 2 The RMSE is more than or equal to 0.996 and less than or equal to 0.0063, the accuracy can meet the requirement, so that the fitting order is taken as 3 order, the coefficients of the primary term, the secondary term and the tertiary term of the Logit model are used as health features HF 2-HF 4, and the fitting order is taken as 3 orderEstimating a state of health of the battery; the third-order fitting precision change of the CS35 battery full life cycle voltage curve is shown in fig. 4;
s2.4: the correlation between the health features HF1-HF4 and the SOH of the battery is verified by using a formula (6), the pearson correlation coefficients between the four health features and the SOH of the battery are shown in a table 1, and the absolute values of the pearson correlation coefficients are all above 0.97, so that the four selected health features are reasonable and can reflect the health state of the battery; the relationship between the normalized four health features and the cycle number is shown in fig. 5;
wherein Pearson is a Pearson correlation coefficient, E (·) is an expected value, α represents a health feature vector, and θ represents a battery SOH vector;
TABLE 1 Pearson correlation coefficient between health characteristics and SOH
Step S3: establishing a health state estimation model based on an extreme learning machine (Extreme Learning Machine, ELM) neural network; as shown in fig. 6, the health state estimation model includes an input layer, a hidden layer, and an output layer, wherein the number of nodes of the input layer is 4, the number of nodes of the hidden layer is 9, and the number of nodes of the output layer is 1; suppose there are K sample dataThe state of health estimation model may be expressed as:
wherein L is the number of hidden layer nodes, h (·) is a Sigmoid activation function, ω r =[ω r1 ,ω r2 ,…,ω rK ] T For the input weights of the input layer to the hidden layer,b r =[b r1 ,b r2 ,…,b rK ] T to conceal layer bias, beta r =[β r1 ,β r2 ,…,β rK ]The output layer weight;
the output result of the health state estimation model after training can be infinitely close to the true value, and the health state estimation model can be equivalent to H beta=y and output weight beta=h + Y,H + The inverse matrix of the output matrix H of the health state estimation model is specifically developed as:
step S4: training the health state estimation model, and optimizing the health state estimation model by utilizing a multi-objective optimization algorithm to obtain an optimized health state estimation model;
s4.1: taking CS35 and CS36 battery data as training sets and CS37 battery data as test sets; the input characteristic matrix of the model is F= [ F ] 1 ;F 2 ;…;F n ;…;F N ]The output vector isWherein (1)>Input feature vector representing the nth cycle, < ->Four health features representing the n-th cycle extraction, < > and->The real health state of the nth cycle is represented, and N represents the number of charge and discharge cycles in the full life cycle;
s4.2: parameter omega of health state estimation model r And b r As parameters to be optimized of the multi-objective optimization algorithm, initializing parameters; multiple onesThe population number Agents_no=20, the maximum iteration number Max_iter=10, the current iteration number is defined as GEN, and the GEN=1 is initialized;
the Mean Square Error (MSE) and stability index (SDEX) are taken as two fitness functions, and the calculation formula is as follows:
wherein y is i 、The real health state and the predicted health state of the ith sample are respectively, and std (·) represents the standard deviation;
s4.3: taking a Multi-Objective Whale optimization algorithm (Multi-Objective Whale OptimizationAlgorithm, MOWOA) as an example, performing Multi-Objective balance optimization on two parameters to be optimized of the health state estimation model by using the Multi-Objective Whale optimization algorithm; the multi-objective whale algorithm is a multi-objective optimization algorithm based on non-dominant ranking and crowding distance calculation, specifically, after the parent and the offspring of whale population are combined, all individuals are divided into a plurality of non-dominant fronts by using the non-dominant ranking; secondly, sorting the solutions of each non-dominant front according to the congestion distance from the maximum to the minimum; finally, selecting individuals as the next population according to the non-dominant front level and the crowding distance, wherein the specific optimizing process comprises the following steps:
s4.3.1: calculating and initializing the fitness value of the whale individual in the target space; non-dominant ranking of all whale individuals according to fitness value, taking two whale individuals as known two decision variables e= (e) 1 ,e 2 ,···,e d ) Sum o= (o) 1 ,o 2 ,···,o d ) If and only if it meetsThen ePareto dominates o, noted e > o; if a decision variable e= (e) 1 ,e 2 ,···,e d ) Satisfy->Then we call e= (e) 1 ,e 2 ,···,e d ) For Pareto optimal solutions, a set of all Pareto optimal solutions is called an optimal solution set;
the main purpose of non-dominant ordering is to find a set of solutions so that, among all possible solutions, the set of solutions is not dominated by other solutions; dividing the population into M layers based on the dominant relationship, wherein the first layer is a non-dominant individual set of whale population, the second layer is a non-dominant individual set of the residual population obtained by removing the first layer of individuals, and the M layer is a non-dominant individual set of the residual population obtained by removing the previous M-1 layer of individuals;
s4.3.2: dividing all solutions into layers such that the solutions within each layer are non-dominant to each other; the crowding distance of each solution is calculated, a group of excellent non-dominant solutions is selected in an auxiliary mode, so that the solutions on the Pareto front have certain distribution, and a specific calculation formula is as follows:
wherein: v (V) u (k-1) and V u (k+1) is the value of the u-th function of the solutions before and after the kth solution,and->The maximum and minimum values of the u-th function respectively;
s4.3.3: updating the position of each whale according to predefined exercise rules, and evaluating new solutions, wherein the exercise process comprises three parts of surrounding predation, spiral updating and hunting;
surrounding predation: each whale searches the surrounding environment according to the current position and speed in the searching process, and the direction most likely to contain the hunting object is found, and the mathematical model is shown as follows:
D=|C·X P (t)-X(t)| (12)
X(t+1)=X P (t)-A·D (13)
wherein X is P (t) is the position of the prey, X (t) represents the current position of the whale, D represents the distance between the whale and the prey, t represents the current iteration number, A and C represent vector coefficients, and the definition is as follows:
where a is a convergence factor, linearly decreasing from 2 to 0; r is a random number, and the value range is 0-1;
spiral updating: when a whale individual approaches the optimal solution but cannot reach it will increase its exploration through a spiral path movement, the mathematical model is as follows:
X t-1 =D 3 ·e bl ·cos(2πl)+X * (t) (15)
wherein b is a constant, and the value affects the spiral shape of the spiral; l is a random number, and the value range is [ -1,1]; furthermore, when the probability is less than 0.5, the wrapping shrink phase is in; when the number of the spiral updating phases is larger than 0.5, the spiral updating phases are in the spiral updating phase;
searching for hunting: when |A| >1, whale cannot obtain effective information, and enters a searching stage in a random walk mode, and a mathematical model is as follows:
D=|C×X rand (t)-X(t)| (16)
X(t+1)=X rand (t)-A·D (17)
wherein X is rand (t) a whale location randomly selected from the current population;
s4.3.4: non-dominant sorting is carried out on the new solutions, the crowding distance is calculated, and whether the new solutions are accepted or not is determined according to the number of layers and the crowding distance to which the new solutions belong;
s4.3.5: repeating steps S4.3.3 and S4.3.4 until the iteration is terminated to obtain a set of optimal solutions, wherein the solutions are uniformly distributed on the Pareto front and have certain diversity and feasibility; the health state estimation model corresponding to the optimal solution is the optimized health state estimation model.
Step S5: the optimized health state estimation model is used for estimating the health state of the lithium ion battery, namely estimating the health state of the CS37 battery in the whole life cycle;
evaluating model estimation accuracy by using a Relative Error (RE), an average absolute error (MAE) and a Root Mean Square Error (RMSE), wherein the calculation formulas are shown as formulas (18) to (20); the model estimated stability was evaluated using the stability index (SDEX), see formula (10).
Fig. 7 (a) and (b) are graphs of health estimation curves and relative errors of three models of LSTM, PSO-ELM and WOMOA-ELM, respectively, with respect to a CS37 battery, and fig. 8 (a) and (b) are histograms of relative errors of the PSO-ELM and WOMOA-ELM models, respectively, which are health estimation models of the present application. As can be seen from the figure, the relative error of the WOOA-ELM model is obviously smaller than PSO-ELM, and the model has higher stability, because the parameters of the health state estimation model are subjected to multi-objective balance optimization by utilizing a multi-objective whale optimization algorithm based on non-dominant sorting and crowding distance calculation, the MOWOA-ELM model shows higher prediction precision and stability in the whole prediction interval, and the PSO-ELM model has higher overall estimation precision, but has poorer estimation precision in certain local intervals and unstable estimation result.
The application is applicable to the prior art where it is not described.
Claims (3)
1. The lithium ion battery health state estimation method based on multi-objective optimization is characterized by comprising the following steps of:
step S1: collecting charge-discharge cycle data of a lithium ion battery;
step S2: preprocessing charge-discharge cycle data, extracting a plurality of health features from the charge-discharge cycle data, wherein one health feature is constant-current charging time, and the other health features are polynomial coefficients for fitting a constant-current charging voltage-time curve by using a Logit model;
step S3: establishing a health state estimation model based on the extreme learning machine neural network;
step S4: training a health state estimation model by using collected lithium ion battery charge-discharge cycle data, wherein an input feature matrix of the model is F= [ F ] 1 ;F 2 ;…;F n ;…;F N ]The output vector is Input feature vector representing the nth cycle, < ->Representing healthy features extracted from the nth cycle, < + >>The real health state of the nth cycle is represented, and N represents the number of charge and discharge cycles in the full life cycle;
taking two parameters of input weight from an input layer to a hidden layer and bias of the hidden layer of the health state estimation model as parameters to be optimized, and optimizing the health state estimation model by using a multi-objective optimization algorithm based on non-dominant sorting and crowding distance; the fitness function of the multi-objective optimization algorithm comprises a mean square error and a stability index, and the formula is as follows:
wherein MSE is mean square error, SDEX is stability index, K is sample number, y i 、The real health state and the predicted health state of the ith sample are respectively, and std (·) represents the standard deviation;
step S5: and using the optimized health state estimation model for estimating the health state of the lithium ion battery.
2. The method for estimating the health state of a lithium ion battery based on multi-objective optimization according to claim 1, wherein in step S2, the accuracy of different fitting orders of the Logit model is estimated by determining coefficients and root mean square errors, and when the fitting order is greater than or equal to 3, the accuracy can meet the requirement, so that the coefficients of the first term, the second term and the third term of the Logit model are selected as health features.
3. The multi-objective optimization-based lithium ion battery state of health estimation method according to claim 1 or 2, wherein the multi-objective optimization algorithm is a multi-objective whale optimization algorithm.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310747259.1A CN116774088A (en) | 2023-06-25 | 2023-06-25 | Lithium ion battery health state estimation method based on multi-objective optimization |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310747259.1A CN116774088A (en) | 2023-06-25 | 2023-06-25 | Lithium ion battery health state estimation method based on multi-objective optimization |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116774088A true CN116774088A (en) | 2023-09-19 |
Family
ID=87994250
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310747259.1A Pending CN116774088A (en) | 2023-06-25 | 2023-06-25 | Lithium ion battery health state estimation method based on multi-objective optimization |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116774088A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117233630A (en) * | 2023-11-16 | 2023-12-15 | 深圳屹艮科技有限公司 | Method and device for predicting service life of lithium ion battery and computer equipment |
CN117849628A (en) * | 2024-03-08 | 2024-04-09 | 河南科技学院 | Lithium ion battery health state estimation method based on time sequence transformation memory network |
-
2023
- 2023-06-25 CN CN202310747259.1A patent/CN116774088A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117233630A (en) * | 2023-11-16 | 2023-12-15 | 深圳屹艮科技有限公司 | Method and device for predicting service life of lithium ion battery and computer equipment |
CN117233630B (en) * | 2023-11-16 | 2024-03-15 | 深圳屹艮科技有限公司 | Method and device for predicting service life of lithium ion battery and computer equipment |
CN117849628A (en) * | 2024-03-08 | 2024-04-09 | 河南科技学院 | Lithium ion battery health state estimation method based on time sequence transformation memory network |
CN117849628B (en) * | 2024-03-08 | 2024-05-10 | 河南科技学院 | Lithium ion battery health state estimation method based on time sequence transformation memory network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110568359B (en) | Lithium battery residual life prediction method | |
CN110888059B (en) | Charge state estimation algorithm based on improved random forest combined volume Kalman | |
CN116774088A (en) | Lithium ion battery health state estimation method based on multi-objective optimization | |
CN113253116A (en) | Lithium ion battery state of charge estimation method and storage medium | |
CN113917337A (en) | Battery health state estimation method based on charging data and LSTM neural network | |
CN111366848A (en) | Battery health state prediction method based on PSO-ELM algorithm | |
CN111398832A (en) | Bus battery SOC prediction method based on ANFIS model | |
CN112883632B (en) | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm | |
CN116449218B (en) | Lithium battery health state estimation method | |
CN115856678A (en) | Lithium ion battery health state estimation method | |
CN115308606B (en) | Lithium ion battery health state estimation method based on adjacent characteristics | |
CN114384435A (en) | WSA-LSTM algorithm-based self-adaptive prediction method for residual service life of new energy automobile power battery | |
CN114217234B (en) | IDE-ASRCKF-based lithium ion battery parameter identification and SOC estimation method | |
CN115963407A (en) | ICGWO (intensive care unit) optimization ELM (element-based robust model) based lithium battery SOC estimation method | |
CN117031310A (en) | Method for predicting residual service life of power battery of electric automobile | |
CN111260015A (en) | Lithium ion battery model parameter identification method based on chaotic cat swarm algorithm | |
CN110232432B (en) | Lithium battery pack SOC prediction method based on artificial life model | |
CN117007974A (en) | Solid-state battery SOC estimation method based on model fusion | |
CN117169743A (en) | Battery health state estimation method and device based on partial data and model fusion | |
CN116258059B (en) | Lithium battery temperature real-time prediction model based on multi-time space width learning model and construction method | |
CN116774045A (en) | Lithium battery health state prediction method based on HHO-SVR | |
CN113884936B (en) | ISSA coupling DELM-based lithium ion battery health state prediction method | |
CN114895206B (en) | Lithium ion battery SOH estimation method based on RBF neural network of improved gray wolf optimization algorithm | |
CN111337833A (en) | Lithium battery capacity integrated prediction method based on dynamic time-varying weight | |
CN115542168A (en) | Lithium battery residual service life prediction method based on fusion data driving 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 |