CN114895206A - Lithium ion battery SOH estimation method based on RBF neural network of improved wolf optimization algorithm - Google Patents
Lithium ion battery SOH estimation method based on RBF neural network of improved wolf optimization algorithm Download PDFInfo
- Publication number
- CN114895206A CN114895206A CN202210448979.3A CN202210448979A CN114895206A CN 114895206 A CN114895206 A CN 114895206A CN 202210448979 A CN202210448979 A CN 202210448979A CN 114895206 A CN114895206 A CN 114895206A
- Authority
- CN
- China
- Prior art keywords
- neural network
- wolf
- rbf neural
- improved
- soh
- 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
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]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
- G01R31/388—Determining ampere-hour charge capacity or SoC involving voltage measurements
-
- 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]
- G01R31/389—Measuring internal impedance, internal conductance or related variables
-
- 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]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/70—Energy storage systems for electromobility, e.g. batteries
Abstract
The invention discloses a lithium ion battery SOH estimation method of an RBF neural network based on an improved wolf optimization algorithm, which comprises the following steps: 1. acquiring historical data of a battery to be tested, 2, extracting characteristics based on a charging voltage curve and an increment capacity curve, 3, constructing an improved RBF neural network, optimizing parameters in the network by adopting an improved wolf optimization algorithm based on the training set, and 4, estimating SOH based on a trained model. The method can solve the problems that the traditional RBF neural network is difficult to track the global linear change of the battery SOH and is easy to fall into the local optimum, thereby improving the estimation precision of the model to the SOH of the lithium ion battery.
Description
Technical Field
The invention belongs to the technical field of power battery management, and particularly relates to a lithium ion battery SOH estimation method based on an improved Grey wolf optimization algorithm RBF neural network.
Background
The lithium ion battery has been developed into the most important driving energy of the electric vehicle at present by virtue of the advantages of high energy density, low self-discharge rate and the like. However, as the battery continues to operate, its performance changes and may also present safety issues, which directly limit the widespread use of electric vehicles. As one of the most important parameters of a power battery management system, the state of health (SOH) of a battery can be accurately estimated, which not only can help to solve the above problems, but also can improve the estimation accuracy of other parameters, such as: state of charge, energy state.
Current SOH estimation methods are broadly divided into model-based methods and data-driven based methods. Although the model-based method has high interpretability, the effect of the model-based method is greatly different along with the selection of the model, and the estimation precision is low. Due to the development of big data technology and man-machine artificial intelligence technology, the SOH estimation method based on data driving has good application prospect. Among them, neural network algorithms are widely used in SOH estimation. However, the SOH estimation method based on neural network still has some limitations. First, from the principle of battery aging, it is clear that the capacity decreases with increasing charge-discharge cycles, indicating that the battery state of health degradation data can be viewed as a long-term time series. However, since the hidden layer function of the neural network nonlinearly transforms the input characteristic, the direct linear correlation between the characteristic and the SOH is reduced. Therefore, the estimated SOH gradually deviates from the true value in the late stage, and the long-term linear variation process cannot be accurately followed. Secondly, the traditional neural network has the problems of low convergence speed, easy falling into local optimum, overfitting and the like in parameter selection. In recent years, the application of meta-heuristics in neural network training has attracted much attention of scientists, and the performance of neural networks has been greatly improved. There are many disadvantages to be studied, such as slow convergence rate, poor intelligent search capability, etc. Therefore, how to optimize the parameters of the neural network in the SOH estimation becomes another problem.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and provides a lithium ion battery SOH estimation method based on an RBF neural network with an improved Grey wolf optimization algorithm, so that the problems that the traditional RBF neural network is difficult to track the global linear change of the battery SOH and is easy to fall into local optimization can be solved, and the estimation precision of a model on the SOH of the lithium ion battery can be improved.
The invention discloses a lithium ion battery SOH estimation method based on an improved Grey wolf optimization algorithm RBF neural network, which is characterized by comprising the following steps of:
step S1, data acquisition:
charging and discharging any lithium battery for n times, and recording the current, voltage, temperature and time data of the lithium battery in the charging and discharging process and the capacity of complete discharge each time;
step S2, charging at constant current for k-th charging process 1 k Constant voltage charging time
Peak of incremental capacity curve
And peak value
Corresponding position
As the k-th input feature vector is
The real state of health (SOH) value of the kth lithium battery is taken as an output value
From the input feature matrix F ═ F 1 ；F 2 ；…；F k ；…；F n ]And the output vector
Forming a training set;
s3, constructing an improved RBF neural network, and optimizing parameters in the network by adopting an improved wolf optimization algorithm based on the training set;
step 3.1, constructing an improved RBF neural network by using the formula (1):
in the formula (1), w i Representing the connection weight of the ith hidden layer node and the output node in the RBF neural network, and recording the vector formed by the connection weights of all hidden layer nodes and the output nodes in the RBF neural network as w, c i Representing the central point of the activation function at the ith hidden layer node in the RBF neural network, recording a matrix formed by the central points of the activation functions at all the hidden layer nodes in the RBF neural network as c, phi is a coefficient of a linear polynomial, and h represents the number of the hidden layer nodes; g (F-c) i ) Is the activation function of the hidden layer and is obtained by equation (2):
in the formula (2), σ i Is a variance parameter of an ith hidden layer node in the RBF neural network; then the vector formed by the variance parameters of all hidden layer nodes in the RBF neural network is marked as sigma;
step 3.2, optimizing five parameters [ sigma, c, w, phi, h ] of the improved RBF neural network by improving a wolf optimization algorithm:
step 3.2.1, defining the current iteration number as t, and initializing t to be 1;
defining and initializing various parameters in the improved graying algorithm, including: the size M of the wolf group, the maximum iteration number T, the search space dimension D and the T-th generation coefficient vector a t ；
Step 3.2.2, randomly initiating the twolf generation population { X j (t) | j ═ 1,2, …, M }, where X j (t) represents the direction vectors of jth individual wolf in the tth generation, and the direction vectors of each individual wolf are five parameters in the improved RBF neural network;
step 3.2.3, calculating fitness function value find of j th wolf individual in the t-th generation wolf population by using the formula (3) j (t), thereby obtaining the fitness function value of each wolf individual in the tth generation wolf population;
in the formula (3), the reaction mixture is,
and Y k The real value and the fitting value of the SOH are respectively, and lambda is a parameter;
step 3.2.4, the fitness function values of all the wolf individuals in the tth generation wolf population are sorted in an ascending order, the wolf individuals with the fitness function ranking the first three are found out and are respectively marked as alpha (t), beta (t) and gamma (t), and the direction vector of the tth generation wolf individual alpha (t) is marked as X α (t) and the direction vector of the Tth generation wolf individual beta (t) is marked as X β (t) and the direction vector of the tth grey wolf individual gamma (t) is marked as X γ (t)；
Step 3.2.5, determining the direction vector X (t +1) of the optimal wolf individual in the t +1 th generation by using the formula (4):
in formula (4): gauchy (0,1) represents a random number with a mean value of 0 and a variance of 1;
step 3.2.6, updating the t-th generation coefficient vector a by using the formula (5) t ：
In formula (5): rand represents a random number between (0, 1);
step 3.2.7, after T +1 is assigned to T, whether T is greater than T is judged, if yes, the direction vector of the Tth generation of the optimal wolf individual is output and is used as the optimal parameter [ sigma ] of the optimized improved RBF neural network * ,c * ,w * ,Φ * ,h * ]Otherwise, turning to the step 3.2.3 to execute in sequence;
step S4, saving the optimal parameter [ sigma ] * ,c * ,w * ,Φ * ,h * ]The corresponding RBF neural network is used as a SOH estimation model; for state of health (SOH) estimation of any of the input feature matrices.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention provides a method for estimating the SOH of a battery based on an improved gray wolf optimization algorithm and an RBF neural network. The method optimizes the parameters of the improved RBF neural network through the improved wolf optimization algorithm, and effectively solves the problems that the estimated value of the traditional RBF neural network is easy to deviate from the true value and fall into local optimum at the later stage of SOH estimation, thereby effectively improving the accuracy and robustness of SOH estimation.
2. The invention adds a linear polynomial directly related to characteristic variables into the traditional RBF neural network function to track the linear trend of the battery in the long-term aging process. Compared with the traditional RBF neural network, the method can effectively track the local nonlinear change generated by capacity regeneration and simulate the capacity attenuation trend in the long-term battery aging process, thereby being beneficial to improving the SOH estimation precision.
3. The invention introduces a nonlinear self-adaptive convergence strategy and Cauchy variation in the traditional Greek wolf optimization algorithm to improve the diversity of population and the global exploration and development capability, avoid falling into local optimum and improve the convergence precision. The nonlinear self-adaptive convergence strategy converts the linear change of the convergence factor into the nonlinear change to increase the diversity of the algorithm, and the method is more suitable for the application scene in the real world. In addition, the gray wolf position information of the global search is mutated by introducing a Cauchy distribution function method, so that the global search capability of the gray wolf is improved, the algorithm can better achieve global optimization, and the estimation accuracy of the SOH of the lithium ion battery is greatly improved.
Drawings
FIG. 1 is a charging voltage curve diagram of an experimental lithium ion battery at different cycle times;
FIG. 2 is a graph of incremental capacity of an experimental lithium ion battery at different cycle times;
FIG. 3 is a block diagram of an improved RBF neural network of the present invention;
FIG. 4 is a flow chart of a lithium ion battery SOH estimation method based on an improved wolf optimization algorithm and an RBF neural network according to the present invention;
FIG. 5 is a SOH estimation result and error map based on different algorithms.
Detailed Description
In this embodiment, a lithium ion battery SOH estimation method based on an improved grayish optimization algorithm RBF neural network includes the following steps, as shown in fig. 4:
step S1, data acquisition:
and collecting historical operation data of the battery to be tested, wherein the historical operation data comprises the charge and discharge times, the voltage, the current, the operation time, the complete discharge capacity and the like of each charge and discharge cycle. The battery was subjected to 166 charge-discharge cycles in total. Then, the battery SOH is defined from the viewpoint of capacity as shown in equation (1):
in formula (1): SOH k Is the SOH value of the current cycle, C k Is the battery capacity of the current cycle, C 0 The rated capacity of the battery in a brand new state.
Step S2, fromFeatures relating to battery degradation are extracted from the battery charging voltage curve. Fig. 1 shows the charging voltage curve at different cycle numbers. As the number of cycles increases, the constant-current charging duration decreases, and the constant-voltage charging time increases. Features related to battery degradation are extracted from the incremental capacity curve and are obtained from equation (2). Fig. 2 shows incremental capacity curves at different cycle numbers. As the number of cycles increases, the peak of the incremental capacity curve decreases and the peak-to-peak position shifts to the right. The feature extraction is as follows: with constant current charging time f of the kth charging process 1 k Constant voltage charging time
Peak of incremental capacity curve
And peak value
Corresponding position
As the k-th input feature vector is
The real state of health (SOH) value of the kth lithium battery is taken as an output value
From said input feature matrix F ═ F 1 ；F 2 ；…；F n ]And the output vector
Forming a training set, wherein the characteristics and SOH data of the previous 116 times are the training set, and the rest is a testing set;
s3, constructing an improved RBF neural network, and optimizing parameters in the network by adopting an improved wolf optimization algorithm based on a training set;
and 3.1, constructing an improved RBF neural network by using the formula (3), wherein the network structure is shown in figure 3. The RBF neural network comprises three layers, namely an input layer, a hidden layer and an output layer. Different from the traditional RBF neural network, the output layer of the improved RBF neural network is added with a linear polynomial directly related to the characteristics on the original basis so as to improve the estimation accuracy of the SOH. The reason for this is that it is clear from the principle of battery aging that the capacity decreases with the increase in the number of charge-discharge cycles, which indicates that the battery state of health degradation data can be regarded as a long-term time series. However, since the function of the neural network hidden layer nonlinearly transforms the input features, the direct linear dependence of the features on the SOH is reduced. Therefore, the estimated SOH gradually deviates from the true value in the late stage and cannot accurately follow the long-term linear variation process. Therefore, the improved RBF neural network can not only solve the problem that the SOH of the battery has local nonlinear change due to various uncertain factors, but also can accurately track the long-term linear change of the SOH of the battery;
in the formula (3), w i Representing the connection weight of the ith hidden layer node and the output node in the RBF neural network, and recording the vector formed by the connection weights of all hidden layer nodes and the output nodes in the RBF neural network as w, c i Representing the central point of the activation function at the ith hidden layer node in the RBF neural network, recording a matrix formed by the central points of the activation functions at all the hidden layer nodes in the RBF neural network as c, phi is a coefficient of a linear polynomial, and h represents the number of the hidden layer nodes; g (F-c) i ) Is the activation function of the hidden layer and is obtained by equation (4):
in the formula (4), σ i Is a variance parameter of an ith hidden layer node in the RBF neural network; then the vector formed by the variance parameters of all hidden layer nodes in the RBF neural network is marked as sigma;
and 3.2, optimizing five parameters [ sigma, c, w, phi and h ] of the improved RBF neural network by improving a wolf optimization algorithm. The invention mainly adopts a nonlinear adaptive strategy and Cauchy variation to improve the traditional Greensbane optimization algorithm and improve the problem that the traditional Greensbane optimization algorithm is easy to fall into local optimum, so that the improved Greensbane optimization algorithm can obtain the optimal RBF neural network parameters:
step 3.2.1, defining the current iteration number as t, and initializing t to be 1;
defining and initializing various parameters in the improved graying algorithm, including: the size M of the wolf group, the maximum iteration number T, the search space dimension D and the T-th generation coefficient vector a t ；
Step 3.2.2, randomly initiating the twolf generation population { X j (t) | j ═ 1,2, …, M }, where X j (t) represents the direction vectors of jth individual wolf in the tth generation, and the direction vectors of each individual wolf are five parameters in the improved RBF neural network;
step 3.2.3, calculating fitness function value find of j th wolf individual in the t-th generation wolf population by using the formula (5) j (t), thereby obtaining the fitness function value of each wolf individual in the tth generation wolf population;
in the formula (5), the reaction mixture is,
and Y k The true and fitted values of SOH, respectively, λ is a small number that needs to be kept an order of magnitude from the previous one.
Step 3.2.4, the fitness function value of each wolf individual in the t generation wolf population is subjected to ascending orderSorting, finding out the gray wolf individuals with the fitness function ranking three above as alpha (t), beta (t) and gamma (t), and recording the direction vector of the t-th gray wolf individual alpha (t) as X α (t) and the direction vector of the Tth generation wolf individual beta (t) is marked as X β (t) and the direction vector of the tth grey wolf individual gamma (t) is marked as X γ (t)；
And 3.2.5, determining the direction vector X (t +1) of the optimal wolf individual in the t +1 th generation by using the formula (6), wherein aiming at the characteristic that the wolf optimization algorithm is easy to fall into local optimal, the variety of the population is increased by using Cauchy variation, the global search capability of the algorithm is improved, and the search space is increased. The peak value of the Cauchy distribution function at the origin is small, but the distribution at two ends is long, and the Cauchy distribution function can generate larger disturbance near the current variant Huilus wolf individual by utilizing Cauchy variation, so that the range of the Cauchy distribution function is wider, and the local optimal value can be more easily jumped out by adopting the distribution at two ends of the Cauchy variation;
in formula (6): gauchy (0,1) represents a random number with a mean value of 0 and a variance of 1;
step 3.2.6, updating the t-th generation coefficient vector a by using the formula (7) t Convergence factor a in the traditional gray wolf optimization algorithm t The linear decrease is along with the increase of the iteration number, however, when the problem of a plurality of unknown parameters is solved, the mechanism cannot reflect the convergence problem in the actual optimization process, and therefore, a nonlinear adaptive strategy is introduced to update the convergence factor a t ；
In formula (12): rand represents a random number between (0, 1);
step 3.2.7, after T +1 is assigned to T, whether T is greater than T is judged, if yes, the direction vector of the Tth generation of the optimal wolf individual is output and is used as the optimal parameter [ sigma ] of the optimized improved RBF neural network * ,c * ,w * ,Φ * ,h * ]Otherwise, turning to the step 3.2.3 to execute in sequence;
step S4, saving the optimal parameter [ sigma ] * ,c * ,w * ,Φ * ,h * ]The corresponding RBF neural network is used as a SOH estimation model; for state of health (SOH) estimation of any of the input feature matrices.
In order to verify the superiority of the present invention, the predicted performances of the algorithm of the present invention and the three algorithms of SVR and GPR were compared. Fig. 5 shows SOH estimation results and errors of the three algorithms on a test set, and it can be seen from part (a) in fig. 5 that the SOH estimation value of the lithium battery is very close to the true value by using the method of the present invention, which indicates that the present invention has the characteristic of accurate estimation results. Through the error analysis of part (b) in fig. 5, it can be seen that the algorithm provided by the present invention can control the estimation error within 2% and has higher estimation accuracy compared with the other three algorithms.
Claims (1)
1. A lithium ion battery SOH estimation method based on an improved Grey wolf optimization algorithm RBF neural network is characterized by comprising the following steps:
step S1, data acquisition:
charging and discharging any lithium battery for n times, and recording the current, voltage, temperature and time data of the lithium battery in the charging and discharging process and the capacity of complete discharge each time;
step S2, charging at constant current for k-th charging process 1 k Constant voltage charging time
Peak value f of incremental capacity curve 3 k And peak value f 3 k Corresponding position
As the k-th input feature vector is
The real state of health (SOH) value of the kth lithium battery is taken as an output value
From the input feature matrix F ═ F 1 ；F 2 ；…；F k ；…；F n ]And the output vector
Forming a training set;
s3, constructing an improved RBF neural network, and optimizing parameters in the network by adopting an improved wolf optimization algorithm based on the training set;
step 3.1, constructing an improved RBF neural network by using the formula (1):
in the formula (1), w i Representing the connection weight of the ith hidden layer node and the output node in the RBF neural network, and recording the vector formed by the connection weights of all hidden layer nodes and the output nodes in the RBF neural network as w, c i Representing the central point of the activation function at the ith hidden layer node in the RBF neural network, recording a matrix formed by the central points of the activation functions at all the hidden layer nodes in the RBF neural network as c, phi is a coefficient of a linear polynomial, and h represents the number of the hidden layer nodes; g (F-c) i ) Is the activation function of the hidden layer and is obtained by equation (2):
in the formula (2), σ i Is a variance parameter of an ith hidden layer node in the RBF neural network; then the vector formed by the variance parameters of all hidden layer nodes in the RBF neural network is marked as sigma;
step 3.2, optimizing five parameters [ sigma, c, w, phi, h ] of the improved RBF neural network by improving a wolf optimization algorithm:
step 3.2.1, defining the current iteration number as t, and initializing t to be 1;
defining and initializing various parameters in the improved graying algorithm, including: the size M of the wolf group, the maximum iteration number T, the search space dimension D and the T-th generation coefficient vector a t ；
Step 3.2.2, randomly initiating the twolf generation population { X j (t) | j ═ 1,2, …, M }, where X j (t) represents the direction vectors of jth individual wolf in the tth generation, and the direction vectors of each individual wolf are five parameters in the improved RBF neural network;
step 3.2.3, calculating fitness function value find of j th wolf individual in the t-th generation wolf population by using the formula (3) j (t), thereby obtaining the fitness function value of each wolf individual in the tth generation wolf population;
in the formula (3), the reaction mixture is,
and Y k The real value and the fitting value of the SOH are respectively, and lambda is a parameter;
step 3.2.4, the fitness function values of all the wolf individuals in the tth generation wolf population are sorted in an ascending order, the wolf individuals with the fitness function ranking the first three are found out and are respectively marked as alpha (t), beta (t) and gamma (t), and the direction vector of the tth generation wolf individual alpha (t) is marked as X α (t) and the direction vector of the Tth generation wolf individual beta (t) is marked as X β (t) and the direction vector of the tth grey wolf individual gamma (t) is marked as X γ (t)；
Step 3.2.5, determining the direction vector X (t +1) of the optimal wolf individual in the t +1 th generation by using the formula (4):
in formula (4): gauchy (0,1) represents a random number with a mean value of 0 and a variance of 1;
step 3.2.6, updating the t-th generation coefficient vector a by using the formula (5) t ：
In formula (5): rand represents a random number between (0, 1);
step 3.2.7, after T +1 is assigned to T, whether T is greater than T is judged, if yes, the direction vector of the Tth generation of the optimal wolf individual is output and is used as the optimal parameter [ sigma ] of the optimized improved RBF neural network * ,c * ,w * ,Φ * ,h * ]Otherwise, turning to the step 3.2.3 to execute in sequence;
step S4, saving the optimal parameter [ sigma ] * ,c * ,w * ,Φ * ,h * ]The corresponding RBF neural network is used as a SOH estimation model; for state of health (SOH) estimation of any of the input feature matrices.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210448979.3A CN114895206B (en) | 2022-04-26 | 2022-04-26 | Lithium ion battery SOH estimation method based on RBF neural network of improved gray wolf optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210448979.3A CN114895206B (en) | 2022-04-26 | 2022-04-26 | Lithium ion battery SOH estimation method based on RBF neural network of improved gray wolf optimization algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114895206A true CN114895206A (en) | 2022-08-12 |
| CN114895206B CN114895206B (en) | 2023-04-28 |
Family
ID=82719150
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210448979.3A Active CN114895206B (en) | 2022-04-26 | 2022-04-26 | Lithium ion battery SOH estimation method based on RBF neural network of improved gray wolf optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114895206B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116609672A (en) * | 2023-05-16 | 2023-08-18 | 国网江苏省电力有限公司淮安供电分公司 | Energy storage battery SOC estimation method based on improved BWOA-FNN algorithm |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007056870A1 (en) * | 2005-11-21 | 2007-05-24 | The Governors Of The University Of Alberta | Methods and compositions for pharmacologically controlled targeted immunotherapy |
| EP2808829A1 (en) * | 2013-05-28 | 2014-12-03 | Vrije Universiteit Brussel | Method and apparatus for predicting the perceptual quality of a signal |
| US20170164603A1 (en) * | 2015-12-11 | 2017-06-15 | Joseph E. Kovarik | System and method to drive away geese |
| DE102017103377A1 (en) * | 2016-02-26 | 2017-08-31 | Ford Global Technologies, Llc | SYSTEMS AND METHOD FOR MONITORING THE HEALTH CONDITION OF A BATTERY |
| CN107741568A (en) * | 2017-11-08 | 2018-02-27 | 中南大学 | A kind of lithium battery SOC estimation method that optimization RBF neural is shifted based on state |
| CN108805217A (en) * | 2018-06-20 | 2018-11-13 | 山东大学 | A kind of health state of lithium ion battery method of estimation and system based on support vector machines |
| CN110309920A (en) * | 2019-06-24 | 2019-10-08 | 北京计算机技术及应用研究所 | The network RBF evaluation method of RBF neural based on Hybrid Hierarchy Genetic Algorithm |
| CN110377957A (en) * | 2019-06-20 | 2019-10-25 | 浙江大学 | A kind of overhead crane neural network modeling approach of whale search strategy grey wolf algorithm |
| CN111443293A (en) * | 2020-03-30 | 2020-07-24 | 青岛大学 | Lithium battery state of health (SOH) estimation method based on data driving |
| CN111709524A (en) * | 2020-07-03 | 2020-09-25 | 江苏科技大学 | RBF neural network optimization method based on improved GWO algorithm |
| CN112348080A (en) * | 2020-11-06 | 2021-02-09 | 北京石油化工学院 | RBF improvement method, device and equipment based on industrial control abnormity detection |
| CN112581263A (en) * | 2020-12-23 | 2021-03-30 | 百维金科(上海)信息科技有限公司 | Credit evaluation method for optimizing generalized regression neural network based on wolf algorithm |
| CN113406521A (en) * | 2021-06-11 | 2021-09-17 | 上海电力大学 | Lithium battery health state online estimation method based on feature analysis |
| WO2021185308A1 (en) * | 2020-03-18 | 2021-09-23 | 北京理工大学 | Online determination method and system for state of health of power battery pack of electric vehicle |
| CN113449464A (en) * | 2021-06-11 | 2021-09-28 | 淮阴工学院 | Wind power prediction method based on improved depth extreme learning machine |
-
2022
- 2022-04-26 CN CN202210448979.3A patent/CN114895206B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007056870A1 (en) * | 2005-11-21 | 2007-05-24 | The Governors Of The University Of Alberta | Methods and compositions for pharmacologically controlled targeted immunotherapy |
| EP2808829A1 (en) * | 2013-05-28 | 2014-12-03 | Vrije Universiteit Brussel | Method and apparatus for predicting the perceptual quality of a signal |
| US20170164603A1 (en) * | 2015-12-11 | 2017-06-15 | Joseph E. Kovarik | System and method to drive away geese |
| DE102017103377A1 (en) * | 2016-02-26 | 2017-08-31 | Ford Global Technologies, Llc | SYSTEMS AND METHOD FOR MONITORING THE HEALTH CONDITION OF A BATTERY |
| CN107741568A (en) * | 2017-11-08 | 2018-02-27 | 中南大学 | A kind of lithium battery SOC estimation method that optimization RBF neural is shifted based on state |
| CN108805217A (en) * | 2018-06-20 | 2018-11-13 | 山东大学 | A kind of health state of lithium ion battery method of estimation and system based on support vector machines |
| CN110377957A (en) * | 2019-06-20 | 2019-10-25 | 浙江大学 | A kind of overhead crane neural network modeling approach of whale search strategy grey wolf algorithm |
| CN110309920A (en) * | 2019-06-24 | 2019-10-08 | 北京计算机技术及应用研究所 | The network RBF evaluation method of RBF neural based on Hybrid Hierarchy Genetic Algorithm |
| WO2021185308A1 (en) * | 2020-03-18 | 2021-09-23 | 北京理工大学 | Online determination method and system for state of health of power battery pack of electric vehicle |
| CN111443293A (en) * | 2020-03-30 | 2020-07-24 | 青岛大学 | Lithium battery state of health (SOH) estimation method based on data driving |
| CN111709524A (en) * | 2020-07-03 | 2020-09-25 | 江苏科技大学 | RBF neural network optimization method based on improved GWO algorithm |
| CN112348080A (en) * | 2020-11-06 | 2021-02-09 | 北京石油化工学院 | RBF improvement method, device and equipment based on industrial control abnormity detection |
| CN112581263A (en) * | 2020-12-23 | 2021-03-30 | 百维金科(上海)信息科技有限公司 | Credit evaluation method for optimizing generalized regression neural network based on wolf algorithm |
| CN113406521A (en) * | 2021-06-11 | 2021-09-17 | 上海电力大学 | Lithium battery health state online estimation method based on feature analysis |
| CN113449464A (en) * | 2021-06-11 | 2021-09-28 | 淮阴工学院 | Wind power prediction method based on improved depth extreme learning machine |
Non-Patent Citations (2)
| Title |
|---|
| MINGQIANG LIN 等: "State of health estimation of lithium-ion battery based on an adaptive tunable hybrid radial basis function network" * |
| 李龙刚 等: "基于改进灰狼优化与支持向量回归的锂电池健康状态预测" * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116609672A (en) * | 2023-05-16 | 2023-08-18 | 国网江苏省电力有限公司淮安供电分公司 | Energy storage battery SOC estimation method based on improved BWOA-FNN algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114895206B (en) | 2023-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110888059B (en) | Charge state estimation algorithm based on improved random forest combined volume Kalman | |
| CN113917337A (en) | Battery health state estimation method based on charging data and LSTM neural network | |
| Jiang et al. | Li-ion battery temperature estimation based on recurrent neural networks | |
| CN110806541B (en) | AD-BAS-based lithium battery model parameter identification method | |
| CN113064093A (en) | Energy storage battery state of charge and state of health joint estimation method and system | |
| CN114861527A (en) | Lithium battery life prediction method based on time series characteristics | |
| CN113917334A (en) | Battery health state estimation method based on evolution LSTM self-encoder | |
| CN112734002A (en) | Service life prediction method based on data layer and model layer joint transfer learning | |
| CN113687242A (en) | Lithium ion battery SOH estimation method for optimizing and improving GRU neural network based on GA algorithm | |
| CN114895206B (en) | Lithium ion battery SOH estimation method based on RBF neural network of improved gray wolf optimization algorithm | |
| CN116298936A (en) | Intelligent lithium ion battery health state prediction method in incomplete voltage range | |
| CN112731183B (en) | Improved ELM-based lithium ion battery life prediction method | |
| CN116774086B (en) | Lithium battery health state estimation method based on multi-sensor data fusion | |
| CN116774088A (en) | Lithium ion battery health state estimation method based on multi-objective optimization | |
| CN117007974A (en) | Solid-state battery SOC estimation method based on model fusion | |
| CN116540134A (en) | GWO-LSTM-based lithium ion battery health state estimation method | |
| CN116736133A (en) | Early prediction method for capacity degradation track of lithium ion battery in full life cycle | |
| CN116609678A (en) | Lithium ion battery residual life prediction method based on improved mixed model | |
| CN113884936A (en) | Lithium ion battery health state prediction method based on ISSA coupling DELM | |
| CN114545279A (en) | Lithium battery health state estimation method based on neural network ordinary differential equation | |
| Zhao et al. | Lithium-ion battery state-of-health estimation method using isobaric energy analysis and pso-lstm | |
| CN114487890A (en) | Lithium battery health state estimation method for improving long-term and short-term memory neural network | |
| CN116362418B (en) | Online prediction method for application-level manufacturing capacity of intelligent factory of high-end battery | |
| CN115963407A (en) | ICGWO (intensive care unit) optimization ELM (element-based robust model) based lithium battery SOC estimation method | |
| CN117233627A (en) | Energy storage battery capacity estimation method based on wavelet packet transformation-deep convolution gate control neural network |
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 |