CN113281671A - Lithium ion battery remaining service life prediction method and system based on IGS-SVM - Google Patents

Lithium ion battery remaining service life prediction method and system based on IGS-SVM Download PDF

Info

Publication number
CN113281671A
CN113281671A CN202110722824.XA CN202110722824A CN113281671A CN 113281671 A CN113281671 A CN 113281671A CN 202110722824 A CN202110722824 A CN 202110722824A CN 113281671 A CN113281671 A CN 113281671A
Authority
CN
China
Prior art keywords
lithium battery
data
capacity
parameters
igs
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
Application number
CN202110722824.XA
Other languages
Chinese (zh)
Inventor
李�杰
张志新
李润然
贾渊杰
孟凡熙
张子辰
闫柯朴
赵世明
牛惠萌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changan University
Original Assignee
Changan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Changan University filed Critical Changan University
Priority to CN202110722824.XA priority Critical patent/CN113281671A/en
Publication of CN113281671A publication Critical patent/CN113281671A/en
Pending legal-status Critical Current

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]
    • G01R31/392—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
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Computing Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)
  • Secondary Cells (AREA)

Abstract

The invention discloses a lithium ion battery remaining service life prediction method and system based on IGS-SVM, wherein a health factor HI with the same degradation capability as the NASA lithium battery capacity is constructed by using the voltage of a discharge end; establishing a data set for predicting the RUL of the lithium battery according to the health factor HI and the extracted historical data, and dividing the data set into a training set and a testing set; optimizing the parameters of the support vector machine on the training set by using an improved grid search method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model; and putting the test set into an IGS-SVM model to obtain the average absolute error value, the root mean square error value and the value of the fitting degree coefficient of the test set in model training, measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time. The method is suitable for the on-line RUL prediction of the lithium battery and has good practicability.

Description

Lithium ion battery remaining service life prediction method and system based on IGS-SVM
Technical Field
The invention belongs to the technical field of battery detection, and particularly relates to a lithium ion battery remaining service life prediction method and system based on an IGS-SVM.
Background
With the development of science and technology, lithium ion batteries gradually become important energy storage and supply carriers in many industries due to the comprehensive advantages of small size, high energy density, high working voltage, long service life and the like. Prediction of Remaining service Life (RUL) of a lithium ion battery as a leading-edge technology of lithium ion battery fault diagnosis and Health Management (PHM) is paid more and more attention by researchers, and is gradually becoming a research hotspot of electronic system Health Management and fault diagnosis. The conventional lithium battery RUL prediction method can be divided into a failure physical model and a data driving model. The method aims to establish a mathematical model for describing the degradation behavior of the battery, predict the RUL of the lithium battery by establishing the relation between the degradation sequence of the lithium battery and time, deeply understand the internal chemical mechanism of the lithium battery, analyze the influence factors such as electrolyte, anode and cathode materials, internal impedance and the like in the lithium battery, and establish a corresponding analytical model to predict the residual life of the lithium battery. The data driving method directly extracts characteristic parameters related to the service life of the lithium battery from performance degradation data of the battery, outputs a prediction result through model training, and provides decision information for system monitoring and maintenance.
Characteristic parameter selection and Health Indicator (HI) construction are two key factors influencing the prediction effect of the data-driven-based method. The main parameters of lithium batteries include cycle number, current, voltage, temperature, impedance, capacity, etc. In the existing studies, the capacity and internal resistance are often selected as HI, because these HI directly reflect the physical property degradation of the lithium battery. When the capacity and impedance of the battery are obtained, the lithium battery is required to complete a complete charging and discharging process, which causes inconvenience to online RUL prediction of the lithium battery. Tong et al found that open circuit voltage is another suitable HI for battery State Of Health (SOH). However, the measurement of the open circuit voltage is very time consuming, since the battery requires a long rest time to reach a steady state. Tseng et al observed that replacing the open circuit voltage with a voltage 60 seconds after full discharge was more feasible and accurate in SOH modeling, and that the open circuit voltage could be replaced, but it still has the problem of harsh monitoring conditions. Widodo et al propose the use of sample entropy of discharge voltage in the framework of prediction of battery health assessment. The method provides a useful calculation tool for evaluating the predictability of the time series, and can also quantify the regularity of the data series. However, this approach is time consuming and requires a capacity parameter to evaluate the sample entropy index. Han et al use differential voltages to estimate SOH, Zhou et al propose average voltage drop construction of HI, Liu et al use the time interval of equal discharge voltage difference during each cycle as HI for measuring capacity degradation in SOH modeling, and train with the time interval difference as HI. The research beneficially explores the HI construction, obtains a plurality of achievements with innovation, and still needs to improve the HI construction precision and robustness. The HI of the lithium battery is constructed by adopting an equal-time voltage difference method, and the HI sequences constructed by the voltage differences of different time intervals are respectively compared and researched, so that the capacity degradation is represented by directly using the terminal voltage change. Therefore, in the prediction of the RUL of the lithium battery, the data representing the capacity degradation can be directly obtained on line, and the reliability and the engineering practicability of the RUL prediction are improved.
Algorithms which are applied to the prediction of the RUL of the lithium battery include naive Bayes, Gaussian Process Regression (GPR), Auto Regression (AR) models, Relevance Vector Machines (RVM), Support Vector Machines (SVM), Artificial Neural Networks (ANN) and the like. Chenlin et al presented a naive Bayes model and studied battery degradation models at different ambient temperatures and under different usage conditions. Zhou Jianbao et al used the RVM algorithm to directly estimate the prediction of the remaining life of the lithium ion battery. Sbarufatti and the like adopt an Autoregressive (AR) model to track the degradation trend of the battery capacity, and then determine the order of the AR model by using a particle swarm optimization algorithm, so that the online application of the RUL prediction of the lithium ion battery is realized. And the Liuyue peak and the like fuse the RVM, the PF and the autoregressive AR model to improve the prediction capability of the model. And Surendar and the like optimize SVM parameters by using a GS method to estimate the charge state of the lead storage battery. And (3) optimizing SVM (support vector machine) parameters by using a Particle Swarm Optimization (PSO) to perform charging and RUL prediction of the lithium battery, so as to improve the prediction precision. Li et al used Genetic Algorithm (GA) to perform iterative optimization to find the best hyper-parameter of SVM to predict the RUL of lithium battery. The research achieves better results, but how to select optimal parameters based on the SVR method and ensure high accuracy of RUL prediction still remains a problem.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method and a system for predicting the remaining service life of a lithium ion battery based on an IGS-SVM (integrated gate-support vector machine), aiming at the defects in the prior art, an HI (high-voltage differential) with the same degradation process as a capacity sequence of the lithium battery is constructed by adopting an equal-time voltage difference method, the capacity degradation of the lithium battery is represented by using the change of terminal voltage, the method and the system are suitable for on-line RUL prediction of the lithium battery, have good practicability, and provide a new method for constructing the HI in the RUL prediction of the lithium battery.
The invention adopts the following technical scheme:
a lithium ion battery residual service life prediction method based on an IGS-SVM comprises the following steps:
s1, constructing a health factor HI with the same degradation capacity as the lithium battery capacity by using the voltage of the discharge end;
s2, establishing a data set for predicting the RUL of the lithium battery according to the health factor HI established in the step S1 and the extracted historical data, and dividing the data set into a training set and a testing set;
s3, optimizing the parameters of the support vector machine on the training set obtained in the step S2 by using an improved grid search method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model;
s4, putting the test set obtained in the step S2 into the IGS-SVM model trained in the step S3, and obtaining the average absolute error value, the root mean square error value and the fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
Specifically, in step S1, the history data of the lithium battery includes the capacity, the charge and discharge voltage, the charge and discharge current, and the charge and discharge temperature data of the lithium battery, the discharge voltage data of the lithium battery is extracted by using an equal time voltage difference method, and the discharge voltage is used to construct the health factor HI having the same degradation capability as the capacity of the lithium battery.
Specifically, in step S2, the extracted history data specifically includes:
Figure BDA0003137069370000041
wherein i represents the number of discharge cycles of the lithium battery, VCRepresents the voltage sequence of 50s charging of the lithium battery, ICRepresents the current sequence T of the lithium battery charging for 3500sCSequence representing the highest temperature composition of a lithium battery during charging, TDIIndicating the end of the discharge of the lithium battery, TDThe sequence representing the highest temperature composition during discharge of the lithium battery.
Specifically, in step S2, the preprocessing is performed on the data set, specifically: deleting abnormal points of the data set, processing noise of the data set, supplementing missing values of the data set, performing Person correlation analysis, and mapping the constructed health factor HI and the capacity of the lithium battery.
Further, the correlation coefficient r of the Person correlation analysis is:
Figure BDA0003137069370000042
wherein n is the number of sequences, Xi、YiIs the ith sequence observation corresponding to the variable,
Figure BDA0003137069370000043
are the average of the variables x, y, respectively.
Specifically, step S3 specifically includes:
at C is [10 ]-4,102]G is [10 ]-4,104]Performing rough search by adopting a large step pitch within the range of the step pitch of 20, evaluating a group (C, g) with the highest classification accuracy by using cross validation, and selecting a group with the smallest parameter C if a plurality of groups (C, g) correspond to the highest validation classification accuracy in the parameter selection process; if the corresponding minimum C has a plurality of groups of g, selecting the searched first group (C, g) as the optimal parameter as the local optimal parameter obtained by large-step-distance searching; after finding the locally optimal parameter, one is chosen [10 ] to the left and right of the set of parameters-2,10]And (4) carrying out secondary fine search by adopting a small step pitch 1 to find out the final global optimal parameters (C, g).
Specifically, in step S4, the average absolute error MAE:
Figure BDA0003137069370000051
mean square error RMSE:
Figure BDA0003137069370000052
capacity prediction Error:
Figure BDA0003137069370000053
wherein n is the sequence length,
Figure BDA0003137069370000054
to predict capacity, y is the actual capacity.
Specifically, in step S4, the fitting degree decision coefficient R2Comprises the following steps:
Figure BDA0003137069370000055
wherein y is the observed value of the capacity of the lithium battery,
Figure BDA0003137069370000056
is the average value of the observed values of the capacity of the lithium battery,
Figure BDA0003137069370000057
the method comprises the steps of obtaining a lithium battery capacity estimated value, y is a lithium battery capacity value of the i cycle at the first time, SST is a total square sum, SSR is a regression square sum, and SSE is a residual square sum.
Specifically, in step S4, the charge and discharge voltage, the charge and discharge current, and the charge and discharge temperature of the lithium battery are collected; processing the acquired data, and storing the original data into historical data for real-time updating; putting the processed data into a computing unit to predict the RUL of the lithium battery; and visualizing the RUL prediction result of the lithium battery.
Another technical solution of the present invention is a lithium ion battery remaining service life prediction system based on IGS-SVM, comprising:
the extraction module is used for constructing a health factor HI with the same degradation capacity as the capacity of the lithium battery by using the voltage of the discharge end;
the data module is used for establishing a data set for predicting the RUL of the lithium battery according to the health factor HI established by the extraction module and the extracted historical data, and dividing the data set into a training set and a test set;
the updating module is used for optimizing the parameters of the support vector machine on the training set obtained by the data module by using an improved grid searching method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model;
the prediction module is used for putting the test set of the data module into the IGS-SVM model trained by the updating module to obtain the average absolute error value, the root mean square error value and the fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
Compared with the prior art, the invention has at least the following beneficial effects:
the invention relates to a lithium ion battery remaining service life prediction method based on an IGS-SVM, which adopts an IGS-SVM fusion algorithm, because the capacity degradation sequence of a lithium battery is highly nonlinear, a support vector machine can process a complex nonlinear classification problem through mapping to a high-dimensional space, the quality of the performance of the support vector machine mainly depends on the selection of a kernel function, the nonlinear expression capability of an RBF radial basis kernel function is stronger, the local nonlinear change trend in the battery degradation process can be captured, and compared with other kernel functions, the method has the advantage of small variable parameters. The GS algorithm is a global search algorithm, the evaluation indexes of all parameter values can be obtained by traversing all parameters, and a global optimal solution is selected according to the change condition of the evaluation indexes. In order to ensure that the optimal parameters can be found globally with efficiency and accuracy, the IGS algorithm is used for carrying out global optimization on the parameters of the SVM, and the efficiency and the accuracy of finding the optimal parameters are improved by adding a Cross Validation (CV) and large and small step pitch combination method.
Further, the method of equal time voltage difference is adopted to construct the health factor having the same degradation process with the capacity sequence of the lithium battery, because the lithium battery is required to complete a complete charging and discharging process when the capacity and impedance of the battery are obtained, which causes inconvenience to the online RUL prediction of the lithium battery. The HI is indirectly constructed from the voltage of the discharge end of the lithium battery to replace capacity and impedance parameters capable of directly representing the degradation of the lithium battery, so that the problems that the capacity and the impedance of the battery belong to internal parameters of the battery, the battery is difficult to obtain on line, and the RUL on-line prediction of the lithium battery cannot be realized are effectively solved.
Furthermore, due to the complex electrochemical working environment of the lithium battery, the monitored performance parameters have strong randomness, and the parameters have complex interrelations and obvious nonlinear characteristics, so that the change trend of the performance parameters is difficult to describe by using a clear model. Based on data driving, the fusion model constructed by the deep learning method can be directly learned from a large amount of real experimental data without considering the mathematical relationship between input and output, so that the problem of high nonlinearity can be solved, the modeling process is simpler, and the prediction effect is more accurate. The NASA lithium battery historical data comprises lithium battery capacity, charging and discharging voltage, charging and discharging current and charging and discharging temperature data, the data are parameter characteristics of the lithium battery in the working process, the parameters continuously occur along with the circulating charging and discharging process of the lithium battery, certain relation exists between the parameters and the degradation process of the lithium battery, and the parameters can be used as training characteristics of a lithium battery model training process.
Furthermore, preprocessing the constructed data set is favorable for eliminating abnormal points and noise in the data and filling missing values, and because the acquired data points deviate from the real data too much or are lost due to factors such as sensitivity of a sensor, data point loss, interference, stability of an equipment system and the like in the data acquisition process, the quality of the data set can be improved through preprocessing the data set, and the stability of a training model and the overall prediction effect of the model are improved. The division of the training set and the test set is beneficial to evaluating the quality of model training, optimizing and improving the model in time and training the model with the best effect.
Furthermore, a Pearson (Person) correlation coefficient can reflect the linear correlation degree between two variables, and through Person correlation analysis, the correlation between HI constructed by an equal time voltage difference method and the capacity of the lithium battery can be judged, wherein the stronger the correlation is, the closer the correlation coefficient r is to 1, which represents that the constructed HI can represent the degradation process of the lithium battery; through correlation analysis and analysis, the correlation between each extracted characteristic parameter of the lithium battery and the capacity of the lithium battery can reflect the correlation between the characteristic parameter and the degradation process of the lithium battery, and indirectly reflect the contribution of the characteristic parameter to the model.
Further, the parameters of the SVM are optimized on the training set obtained in the step S2 by using an IGS method to obtain optimal parameters (C, g), the parameters of the support vector machine model are updated by using the optimal parameters (C, g) to obtain an IGS-SVM model, so that the optimal model parameters can be obtained in the global range, the obtained model is guaranteed to be globally optimal, and the accuracy and the reliability of the RUL prediction of the lithium battery can be guaranteed.
Further, step S4 needs to further evaluate and optimize the model performance on the basis of determining the prediction model framework based on the RUL of the IGS-SVM lithium battery. Through training and testing the model and recording the change of the algorithm performance set in the training process, the prediction capability of the model on the lithium battery RUL is reflected. The method adopts average absolute error (MAE) and Root Mean Square Error (RMSE) and error (Er) as algorithm performance indexes, measures the error between a predicted value and a true value, and evaluates the RUL of the lithium battery based on the IGS-SVM fusion method to predict the model. The MAE is also defined as the Loss function (Loss).
Further, the evaluation index uses a fitness function R2(also called coefficient of determination) measures goodness of fit, R, of actual and predicted values in a regression training process2The closer to 1 the value of (A) the better the fitting.
Further, training data is constructed by using the training model through the data processed in the steps S1 to S2, the IGS optimization method provided in the steps S3 and S4 is used for optimizing parameters (C, g) of the SVM on the training data through the IGS method to obtain optimal (C, g) parameters, and the parameters of the SVM model are updated to obtain the IGS-SVM training model.
In conclusion, the invention provides an equal-time voltage difference method for constructing HI with the same degradation process as the capacity sequence of the lithium battery, and can represent the capacity degradation of the lithium battery by using the change of the terminal voltage, because the terminal voltage parameter is easy to directly measure, the method is suitable for the on-line RUL prediction of the lithium battery, has good practicability, and provides a new method for constructing HI in the RUL prediction of the lithium battery; compared with the approximate algorithm or the heuristic intelligent optimization algorithm, the Grid Search (GS) algorithm can traverse the parameters of the corresponding values through global optimization, and particularly can greatly improve the possibility of obtaining a global optimal solution on the premise of not excessively increasing the calculation amount aiming at the optimization problem with lower dimensionality. Therefore, aiming at the constructed HI of the lithium battery, the method improves the GS, performs global optimization on the parameters of the SVM through the IGS algorithm, and verifies that the method provided by the invention has higher accuracy and robustness.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
FIG. 1 is a flow chart of IGS-SVM model prediction;
FIG. 2 shows a schematic diagram of a RUL prediction system for a lithium battery;
FIG. 3 is a block diagram of a data acquisition system;
FIG. 4 is a comparison diagram of the cycle charge and discharge process of a lithium battery;
FIG. 5 is a graph comparing the degradation process of the capacity of a lithium battery;
FIG. 6 is a comparison graph of the voltage variation process of the discharge end of the discharge cycle of the lithium battery;
FIG. 7 is a schematic diagram of HI extraction at the discharge end of a lithium battery;
FIG. 8 is a diagram of lithium battery HI and capacity mapping results;
FIG. 9 is a diagram of an IGS-based SVM parameter optimization process;
FIG. 10 is a diagram showing the RUL prediction result of a lithium battery.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
Various structural schematics according to the disclosed embodiments of the invention are shown in the drawings. The figures are not drawn to scale, wherein certain details are exaggerated and possibly omitted for clarity of presentation. The shapes of various regions, layers and their relative sizes and positional relationships shown in the drawings are merely exemplary, and deviations may occur in practice due to manufacturing tolerances or technical limitations, and a person skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions, according to actual needs.
The invention provides a lithium ion battery remaining service life prediction method and system based on an IGS-SVM (integrated gate-support vector machine). A Health factor (HI) with the same degradation process as a capacity sequence of a lithium battery is constructed by adopting an equal time voltage difference method, and terminal voltage information which is easy to acquire on line is used for representing the capacity degradation of the lithium battery. On the basis, an Improved Grid Search (IGS) algorithm is provided to perform global optimization on parameters of a Support Vector Machine (SVM), and the IGS-SVM fusion algorithm is used for predicting the RUL of the lithium battery, so that the real-time online prediction capability of the lithium battery is realized.
Compared with an approximation method or a heuristic intelligent optimization method, the Grid Search (GS) method can traverse parameters of corresponding values through global optimization, and particularly can greatly improve the possibility of obtaining a global optimal solution on the premise of not excessively increasing the calculated amount aiming at the optimization problem of lower dimensionality. Therefore, aiming at the constructed HI of the lithium battery, the GS is Improved, an IGS-SVM method is designed, global optimization and optimization are carried out on parameters of the SVM through an IGS (Improved Grid Search) algorithm, and the method provided by the invention is verified to have higher accuracy and robustness.
The invention predicts the RUL of the lithium battery, and performs regression analysis and prediction on data acquired by a sensor carried by a lithium battery health management system, wherein the RUL prediction flow chart of the lithium battery is shown in figure 1, and the RUL prediction system design prediction flow chart of the lithium battery is shown in figure 2 and can be divided into the following three parts.
Referring to fig. 1, the method for predicting the remaining service life of a lithium ion battery based on an IGS-SVM of the present invention includes the following steps:
s1, constructing a health factor HI with the same degradation capacity as the lithium battery capacity by using the voltage of the discharge end;
firstly, extracting characteristics of lithium battery capacity, charge-discharge voltage, charge-discharge current and charge-discharge temperature data, extracting discharge voltage data of the lithium battery by adopting an equal time voltage difference method, analyzing the correlation of the factors and the lithium battery capacity, and constructing a health factor HI with the same degradation capacity as the lithium battery capacity by using the discharge terminal voltage; the comparative graph of the capacity degradation process of the lithium battery is shown in fig. 5, and the comparative graph of the discharge end voltage change process of the discharge cycle of the lithium battery is shown in fig. 6.
S2, establishing a lithium battery RUL predicted data set according to the HI constructed in the step S1 and the extracted features, and dividing the data set into a 50% training set and a 50% testing set;
the characteristic parameters of the lithium battery are extracted as follows:
Figure BDA0003137069370000121
wherein i represents the number of discharge cycles of the lithium battery, VCRepresents the voltage sequence of 50s charging of the lithium battery, ICRepresents the current sequence T of the lithium battery charging for 3500sCSequence representing the highest temperature composition of a lithium battery during charging, TDIIndicating the end of the discharge of the lithium battery, TDThe sequence representing the highest temperature composition during discharge of the lithium battery.
The HI of the lithium battery charging voltage end at different time intervals, such as a delta V sequence of time intervals of 500s, 1500s, 2300s and the like, is extracted, the extraction principle is shown in fig. 7, and the mapping result of the HI and the capacity of the lithium battery is shown in fig. 8. The corresponding sequences are respectively delta V500、ΔV1500And Δ V2300As the health factor HI, there are:
Figure BDA0003137069370000122
in the prediction of the RUL of the lithium battery, the input factor is the charging voltage VCCharging current ICTemperature T of chargingCD discharge current TDITemperature T of dischargeDMeanwhile, with the HI shown in equation 1 as output, the following data set matrix is designed:
Figure BDA0003137069370000123
Figure BDA0003137069370000131
Figure BDA0003137069370000132
wherein i represents the number of discharge cycles of the lithium battery, D500Is expressed as Δ V500Data set of HI, D1500Is expressed as Δ V1500Data set of HI, D2300Is expressed as Δ V2300Data set of HI.
S3, an IGS method;
the modified grid is empirically set to have a range for C and g, respectively, with the range for C set to [10 ]-5,105]G is set to [10 ]-4,104]Based on the training set, the local optimum range (C, g) is found by using the large step pitch 20, after the local optimum parameter is found, a small interval is selected near the group of parameters, and the final optimum parameter (C, g) is found by adopting the small step pitch 1 to carry out secondary fine search.
And (3) constructing training data by using the data processed in the steps S1 to S2, optimizing parameters (C, g) of the SVM on the training data by using the IGS optimization method provided in the steps S3 and S4 to obtain optimal (C, g) parameters, and updating the parameters of the SVM model to obtain an IGS-SVM training model.
The results of the optimization of the IGS-SVM parameters are shown in Table 1
TABLE 1 IGS optimization SVM parameter pre-post comparison
Figure BDA0003137069370000133
Figure BDA0003137069370000141
And combining the constructed training data, and performing data training by using an IGS-SVM model, wherein the optimization process of the parameters is shown in FIG. 9.
S4, testing the model;
putting the data of the test set into the model trained in step S3 to obtain Mean Absolute Error (MAE) value, Root Mean Squared Error (RMSE) value and fitting degree decision coefficient R of the test set in the model training2And (4) verifying the quality of the model.
And S5, encapsulating the model with high accuracy into a computing unit, facilitating real-time prediction of new data collected later, and finally visualizing the prediction result.
Online prediction of remaining life of lithium battery
S501, collecting sensor data on the lithium battery through a data collector;
referring to fig. 3 and 4, after the construction of the RUL prediction fusion model based on the IGS-SVM lithium battery is completed, in order to predict an updated data set of the lithium battery in real time, the real-time state parameters (charge-discharge voltage, charge-discharge current, and charge-discharge temperature) of the lithium battery need to be collected, detected and collected by a data collection system. The data acquisition system generally comprises a sensor module, a signal conditioning circuit module, a data acquisition unit module, a computer and an application software part, and the specific flow is represented as follows:
(1) a sensor on the lithium battery detects a corresponding state signal and converts the state signal into a corresponding output signal;
(2) the signal conditioning circuit carries out filtering, conversion, amplification and other processing on the analog signal output by the sensor;
(3) the data acquisition unit discretizes the continuous analog signals after the signals are conditioned, converts the continuous analog signals into discrete digital signals through sampling with certain frequency, and transmits the acquired digital signals to a computer processing system.
(4) The computer processing system stores the conversion data acquired by the data acquisition unit and transmits the data to the data processing module; in addition, the collected signals are visualized by computer application software for monitoring the change of the sensor in real time.
The improper use and working conditions of the lithium battery are the main reasons for the rapid reduction of the cycle life of the lithium battery and even the occurrence of safety accidents such as combustion, explosion and the like. In order to ensure the precision and stability of the acquired data, the acquisition system meets the requirements of working under the following conditions:
(1) the lithium battery needs to work at the ambient temperature of-20 to 66 ℃, which is a temperature threshold space acceptable for the normal work of the lithium battery;
(2) the overcharge and overdischarge have great influence on the service life of the lithium battery, the voltage range of a common lithium manganate battery is 2.8-4.2V, the voltage range of a lithium iron phosphate battery is 2.0-3.6V, and the normal working voltage range of the lithium battery is wide;
(3) the charge and discharge multiplying power represents the large-current working capacity of the lithium battery, the maximum charge multiplying power and the maximum discharge multiplying power respectively refer to the maximum current allowed by the lithium battery during charge and discharge, and the multiplying powers of the lithium batteries of different materials and different purposes are different and work according to the multiplying power requirement of the lithium battery in actual use;
(4) a wider range of acquisition rates is required. The collection rate is generally required to be from 1 to 200 kHz;
(5) the measurement accuracy (or precision) and resolution requirements are high. For example, when some parameters of the lithium battery are subjected to AD conversion, an AD converter with 32-bit resolution is required to be used, so that the system accuracy is up to +/-0.01% FS to +/-0.03% FS;
(6) the sensor can timely reflect the change of temperature, namely the time constant of the sensor is small (less than 0.1 s);
(7) the non-linearity error of the amplifier is less than 0.25%.
S502, processing the acquired data, and storing the original data into historical data for real-time updating;
in order to enable the collected data to be normally and rapidly identified by a lithium battery RUL prediction module in a core computing unit, a data processing process in a data processing module mainly comprises the following 3 aspects:
(1) updating the data set: arranging the data according to the collected time sequence and adding labels to form a data form consistent with the historical data of the lithium battery so as to facilitate the next processing;
(2) data preprocessing: carrying out data preprocessing on the constructed data, and inputting the data into a core computing unit to carry out lithium battery RUL prediction;
(3) and (3) data set storage: and sequencing the original data and the preprocessed data and storing the data to realize real-time updating.
S503, the processed data are put into a computing unit to predict the RUL of the lithium battery;
s504, visualizing the RUL prediction result of the lithium battery.
The real-time results of the sensor information converted by the data collector and the RUL prediction of the lithium battery predicted by the core computing unit are shown in fig. 10.
In another embodiment of the present invention, a lithium ion battery remaining service life prediction system based on an IGS-SVM is provided, which can be used to implement the above method for predicting the remaining service life of a lithium ion battery based on an IGS-SVM.
The extraction module is used for constructing a health factor HI with the same degradation capability in the capacity of the lithium battery by using the voltage of the discharge end;
the data module is used for establishing a data set for predicting the RUL of the lithium battery according to the health factor HI established by the extraction module and the extracted historical data, and dividing the data set into a training set and a test set;
the updating module is used for optimizing the parameters of the support vector machine on the training set obtained by the data module by using an improved grid searching method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model;
the prediction module is used for putting the test set of the data module into the IGS-SVM model trained by the updating module to obtain the average absolute error value, the root mean square error value and the fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
In yet another embodiment of the present invention, a terminal device is provided that includes a processor and a memory for storing a computer program comprising program instructions, the processor being configured to execute the program instructions stored by the computer storage medium. The Processor may be a Central Processing Unit (CPU), or may be other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable gate array (FPGA) or other Programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc., which is a computing core and a control core of the terminal, and is adapted to implement one or more instructions, and is specifically adapted to load and execute one or more instructions to implement a corresponding method flow or a corresponding function; the processor provided by the embodiment of the invention can be used for the operation of the lithium ion battery residual service life prediction method based on the IGS-SVM, and comprises the following steps:
constructing a health factor HI with the same degradation capacity as the capacity of the lithium battery by using the voltage of the discharge end; establishing a data set for predicting the RUL of the lithium battery according to the health factor HI and the extracted historical data, and dividing the data set into a training set and a testing set; optimizing the parameters of the support vector machine on the training set by using an improved grid search method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model; putting the test set into an IGS-SVM model to obtain an average absolute error value, a root mean square error value and a fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
In still another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a Memory device in a terminal device and is used for storing programs and data. It is understood that the computer readable storage medium herein may include a built-in storage medium in the terminal device, and may also include an extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer-readable storage medium may be a high-speed RAM memory, or may be a non-volatile memory (non-volatile memory), such as at least one disk memory.
The computer-readable storage medium can be loaded with one or more instructions and executed by a processor to implement the corresponding steps of the method for predicting the remaining service life of the lithium ion battery based on the IGS-SVM in the above embodiments; one or more instructions in the computer-readable storage medium are loaded by the processor and perform the steps of:
constructing a health factor HI with the same degradation capacity as the capacity of the lithium battery by using the voltage of the discharge end; establishing a data set for predicting the RUL of the lithium battery according to the health factor HI and the extracted historical data, and dividing the data set into a training set and a testing set; optimizing the parameters of the support vector machine on the training set by using an improved grid search method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model; putting the test set into an IGS-SVM model to obtain an average absolute error value, a root mean square error value and a fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
In summary, the method for predicting the remaining service life of the lithium ion battery based on the IGS-SVM can realize the on-line prediction of the RUL of the lithium ion battery by constructing the HI and IGS-SVM fusion algorithm. The advantages are as follows:
firstly, the method comprises the following steps: due to the complex electrochemical working environment of the lithium battery, the monitored performance parameters have strong randomness, and the parameters have complex interrelations and obvious nonlinear characteristics, so that the change trend of the performance parameters is difficult to describe by using a clear model. Based on data driving, the fusion model constructed by the deep learning method can be directly learned from a large amount of real experimental data without considering the mathematical relationship between input and output, so that the problem of high nonlinearity can be solved, the modeling process is simpler, and the prediction effect is more accurate.
Secondly, the method comprises the following steps: the health factor with the same degradation process as the capacity sequence of the lithium battery is constructed by adopting an equal time voltage difference method, and the lithium battery is required to complete a complete charging and discharging process when the capacity and the impedance of the battery are obtained, so that the online RUL prediction of the lithium battery is inconvenient. The HI is indirectly constructed from the voltage of the discharge end of the lithium battery to replace capacity and impedance parameters capable of directly representing the degradation of the lithium battery, so that the problems that the capacity and the impedance of the battery belong to internal parameters of the battery, the battery is difficult to obtain on line, and the RUL on-line prediction of the lithium battery cannot be realized are effectively solved.
Thirdly, the method comprises the following steps: by adopting the IGS-SVM fusion algorithm, as the capacity degradation sequence of the lithium battery is highly nonlinear, the performance of the support vector machine is mainly determined by the selection of the kernel function, the nonlinear expression capability of the RBF radial basis kernel function is stronger, the local nonlinear change trend in the battery degradation process can be captured, compared with other kernel functions, the method has the advantage of small variable parameters, a Gaussian kernel (RBF radial basis kernel) function is selected as the kernel function for model training, and the RBF radial basis kernel function SVM model has two important parameters, namely C and g. The GS algorithm is a global search algorithm, the evaluation indexes of all parameter values can be obtained by traversing all parameters, and a global optimal solution is selected according to the change condition of the evaluation indexes. In order to ensure that the optimal parameters can be found globally with efficiency and accuracy, an Improved Grid Search (IGS) algorithm is used for carrying out global optimization on the parameters of a Support Vector Machine (SVM), and the efficiency and accuracy of finding the optimal parameters are Improved by adding a Cross Validation (CV) and large and small step pitch combination method.
Fourthly: the data of the lithium battery are collected in real time by using a data collector, and then the collected data are preprocessed and input into a model trained by a neural network, so that the on-line prediction of the RUL of the lithium battery can be realized, and the result is visualized. Meanwhile, collected data are monitored, correspondingly processed and added into historical data for retraining, parameters in the model are optimized, and a real-time prediction model for predicting the RUL of the lithium battery is established, so that the purpose of real-time updating is achieved, and prediction accuracy is improved.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above-mentioned contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modification made on the basis of the technical idea of the present invention falls within the protection scope of the claims of the present invention.

Claims (10)

1. A lithium ion battery residual service life prediction method based on an IGS-SVM is characterized by comprising the following steps:
s1, constructing a health factor HI with the same degradation capacity as the lithium battery capacity by using the voltage of the discharge end;
s2, establishing a data set for predicting the RUL of the lithium battery according to the health factor HI established in the step S1 and the extracted historical data, and dividing the data set into a training set and a testing set;
s3, optimizing the parameters of the support vector machine on the training set obtained in the step S2 by using an improved grid search method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model;
s4, putting the test set obtained in the step S2 into the IGS-SVM model trained in the step S3, and obtaining the average absolute error value, the root mean square error value and the fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
2. The method according to claim 1, wherein in step S1, the history data of the lithium battery includes the capacity, the charging and discharging voltage, the charging and discharging current and the charging and discharging temperature data of the lithium battery, the discharging voltage data of the lithium battery is extracted by using an equal time voltage difference method, and the discharging voltage is used to construct the health factor HI having the same degradation capability as the capacity of the lithium battery.
3. The method according to claim 1, wherein in step S2, the extracted history data is specifically:
Figure FDA0003137069360000011
wherein i represents the number of discharge cycles of the lithium battery, VCRepresents the voltage sequence of 50s charging of the lithium battery, ICRepresents the current sequence T of the lithium battery charging for 3500sCSequence representing the highest temperature composition of a lithium battery during charging, TDIIndicating the end of the discharge of the lithium battery, TDThe sequence representing the highest temperature composition during discharge of the lithium battery.
4. The method according to claim 1, wherein in step S2, the data set is preprocessed, specifically: deleting abnormal points of the data set, processing noise of the data set, supplementing missing values of the data set, performing Person correlation analysis, and mapping the constructed health factor HI and the capacity of the lithium battery.
5. The method of claim 4, wherein the correlation coefficient r of the Person correlation analysis is:
Figure FDA0003137069360000021
wherein n is the number of sequences, Xi、YiIs the ith sequence observation corresponding to the variable,
Figure FDA0003137069360000022
are the average of the variables x, y, respectively.
6. The method according to claim 1, wherein step S3 is specifically:
at C is [10 ]-4,102]G is [10 ]-4,104]And roughly searching by adopting a large step pitch within the range of step pitch of 20, evaluating a group (C, g) with the highest classification accuracy by using cross validation, and if multiple groups (C, g) correspond to the highest validation classification accuracy in the parameter selection process, determining that the parameters are classified according to the highest validation classification accuracySelecting a group with the minimum parameter C; if the corresponding minimum C has a plurality of groups of g, selecting the searched first group (C, g) as the optimal parameter as the local optimal parameter obtained by large-step-distance searching; after finding the locally optimal parameter, one is chosen [10 ] to the left and right of the set of parameters-2,10]And (4) carrying out secondary fine search by adopting a small step pitch 1 to find out the final global optimal parameters (C, g).
7. The method according to claim 1, wherein in step S4, the average absolute error MAE:
Figure FDA0003137069360000023
mean square error RMSE:
Figure FDA0003137069360000031
capacity prediction Error:
Figure FDA0003137069360000032
wherein n is the sequence length,
Figure FDA0003137069360000033
to predict capacity, y is the actual capacity.
8. The method of claim 1, wherein in step S4, the fitting degree coefficient R is determined2Comprises the following steps:
Figure FDA0003137069360000034
wherein y is the observed value of the capacity of the lithium battery,
Figure FDA0003137069360000035
is the average value of the observed values of the capacity of the lithium battery,
Figure FDA0003137069360000036
the method comprises the steps of obtaining a lithium battery capacity estimated value, y is a lithium battery capacity value of the i cycle at the first time, SST is a total square sum, SSR is a regression square sum, and SSE is a residual square sum.
9. The method according to claim 1, wherein in step S4, the charge and discharge voltage, the charge and discharge current, and the charge and discharge temperature of the lithium battery are collected; processing the acquired data, and storing the original data into historical data for real-time updating; putting the processed data into a computing unit to predict the RUL of the lithium battery; and visualizing the RUL prediction result of the lithium battery.
10. A lithium ion battery remaining service life prediction system based on IGS-SVM is characterized by comprising:
the extraction module is used for constructing a health factor HI with the same degradation capacity as the capacity of the lithium battery by using the voltage of the discharge end;
the data module is used for establishing a data set for predicting the RUL of the lithium battery according to the health factor HI established by the extraction module and the extracted historical data, and dividing the data set into a training set and a test set;
the updating module is used for optimizing the parameters of the support vector machine on the training set obtained by the data module by using an improved grid searching method to obtain optimal parameters, and updating the parameters of the support vector machine model by using the optimal parameters to obtain an IGS-SVM model;
the prediction module is used for putting the test set of the data module into the IGS-SVM model trained by the updating module to obtain the average absolute error value, the root mean square error value and the fitting degree coefficient R of the test set in model training2And (3) measuring the capacity prediction error between the predicted value and the true value, and predicting the residual service life of the lithium battery in real time.
CN202110722824.XA 2021-06-28 2021-06-28 Lithium ion battery remaining service life prediction method and system based on IGS-SVM Pending CN113281671A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110722824.XA CN113281671A (en) 2021-06-28 2021-06-28 Lithium ion battery remaining service life prediction method and system based on IGS-SVM

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110722824.XA CN113281671A (en) 2021-06-28 2021-06-28 Lithium ion battery remaining service life prediction method and system based on IGS-SVM

Publications (1)

Publication Number Publication Date
CN113281671A true CN113281671A (en) 2021-08-20

Family

ID=77285949

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110722824.XA Pending CN113281671A (en) 2021-06-28 2021-06-28 Lithium ion battery remaining service life prediction method and system based on IGS-SVM

Country Status (1)

Country Link
CN (1) CN113281671A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114239373A (en) * 2021-12-17 2022-03-25 上海空间电源研究所 Particle swarm algorithm-based aerospace BDR module residual life prediction method
CN114290960A (en) * 2021-12-30 2022-04-08 中国第一汽车股份有限公司 Method and device for acquiring battery health degree of power battery and vehicle
CN114355197A (en) * 2021-12-07 2022-04-15 暨南大学 Method and device for rapidly detecting complementary energy of power battery
CN114372417A (en) * 2022-01-12 2022-04-19 青岛科技大学 Electric vehicle battery health state and remaining life evaluation method based on charging network
CN114420983A (en) * 2022-03-31 2022-04-29 北京英博新能源有限公司 Method, device and system for evaluating health of fuel cell stack and electronic equipment
CN114881316A (en) * 2022-04-24 2022-08-09 上海玫克生储能科技有限公司 Lithium battery residual life prediction method and system, terminal device and storage medium
CN116227087A (en) * 2023-03-29 2023-06-06 南京理工大学 Method for predicting tensile characteristics of welded joint of thin-wall part based on finite element method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106908736A (en) * 2017-03-17 2017-06-30 哈尔滨工业大学 Based on the lithium battery method for predicting residual useful life that depth confidence net and Method Using Relevance Vector Machine are merged
US20190205621A1 (en) * 2018-01-04 2019-07-04 Shutterfly, Inc. High precision additive pattern recognition for image and other applications
CN110161425A (en) * 2019-05-20 2019-08-23 华中科技大学 A kind of prediction technique of the remaining life divided based on lithium battery catagen phase
CN110443377A (en) * 2019-06-24 2019-11-12 南方电网调峰调频发电有限公司信息通信分公司 A kind of support vector machines life of storage battery prediction technique based on immune algorithm optimization
CN110687451A (en) * 2019-08-28 2020-01-14 武汉科技大学 Error compensation-based method for predicting residual life of lithium battery of support vector machine

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106908736A (en) * 2017-03-17 2017-06-30 哈尔滨工业大学 Based on the lithium battery method for predicting residual useful life that depth confidence net and Method Using Relevance Vector Machine are merged
US20190205621A1 (en) * 2018-01-04 2019-07-04 Shutterfly, Inc. High precision additive pattern recognition for image and other applications
CN110161425A (en) * 2019-05-20 2019-08-23 华中科技大学 A kind of prediction technique of the remaining life divided based on lithium battery catagen phase
CN110443377A (en) * 2019-06-24 2019-11-12 南方电网调峰调频发电有限公司信息通信分公司 A kind of support vector machines life of storage battery prediction technique based on immune algorithm optimization
CN110687451A (en) * 2019-08-28 2020-01-14 武汉科技大学 Error compensation-based method for predicting residual life of lithium battery of support vector machine

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
JIANFANG JIA，等: "SOH and RUL Prediction of Lithium-Ion Batteries Based on Gaussian Process Regression with Indirect Health Indicators", 《ENERGIES》, pages 1 - 20 *
季方芳，等: "基于IGS-SVM模型的牛肉生理成熟度预测方法", 《食品科学》, pages 71 - 77 *
徐莹: "纯电动汽车锂离子电池剩余寿命预测研究", 《中国优秀硕士学位论文全文数据库》, pages 3 - 4 *
王树坤;黄妙华;刘安康;刘泽明;: "基于GA-SVR模型的锂离子电池剩余容量预测", 汽车技术, no. 10, pages 57 - 60 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114355197A (en) * 2021-12-07 2022-04-15 暨南大学 Method and device for rapidly detecting complementary energy of power battery
CN114239373A (en) * 2021-12-17 2022-03-25 上海空间电源研究所 Particle swarm algorithm-based aerospace BDR module residual life prediction method
CN114290960A (en) * 2021-12-30 2022-04-08 中国第一汽车股份有限公司 Method and device for acquiring battery health degree of power battery and vehicle
CN114290960B (en) * 2021-12-30 2023-10-27 中国第一汽车股份有限公司 Method and device for acquiring battery health of power battery and vehicle
CN114372417A (en) * 2022-01-12 2022-04-19 青岛科技大学 Electric vehicle battery health state and remaining life evaluation method based on charging network
CN114420983A (en) * 2022-03-31 2022-04-29 北京英博新能源有限公司 Method, device and system for evaluating health of fuel cell stack and electronic equipment
CN114881316A (en) * 2022-04-24 2022-08-09 上海玫克生储能科技有限公司 Lithium battery residual life prediction method and system, terminal device and storage medium
CN114881316B (en) * 2022-04-24 2023-06-16 上海玫克生储能科技有限公司 Lithium battery remaining life prediction method, system, terminal equipment and storage medium
CN116227087A (en) * 2023-03-29 2023-06-06 南京理工大学 Method for predicting tensile characteristics of welded joint of thin-wall part based on finite element method

Similar Documents

Publication Publication Date Title
CN113281671A (en) Lithium ion battery remaining service life prediction method and system based on IGS-SVM
Guo et al. Online estimation of SOH for lithium-ion battery based on SSA-Elman neural network
Lipu et al. A review of state of health and remaining useful life estimation methods for lithium-ion battery in electric vehicles: Challenges and recommendations
CN112092675B (en) Battery thermal runaway early warning method, system and server
Tian et al. Deep neural network battery charging curve prediction using 30 points collected in 10 min
Zhang et al. Research on electric vehicle charging safety warning model based on back propagation neural network optimized by improved gray wolf algorithm
Kaur et al. Deep learning networks for capacity estimation for monitoring SOH of Li‐ion batteries for electric vehicles
Chen et al. State of health estimation for lithium-ion batteries based on fusion of autoregressive moving average model and elman neural network
Liu et al. Future ageing trajectory prediction for lithium-ion battery considering the knee point effect
Song et al. Remaining life prediction of lithium-ion batteries based on health management: A review
Zhang et al. Aging performance characterization and state-of-health assessment of retired lithium-ion battery modules
CN109604186A (en) Power battery performance flexibility assesses method for separating
CN116609676B (en) Method and system for monitoring state of hybrid energy storage battery based on big data processing
Lan et al. In-depth bibliometric analysis on research trends in fault diagnosis of lithium-ion batteries
Liu et al. Comparative study of incremental capacity curve determination methods for lithium-ion batteries considering the real-world situation
Zou et al. Advancements in Artificial Neural Networks for health management of energy storage lithium-ion batteries: A comprehensive review
CN111460656B (en) Method and system for evaluating operation life of communication power supply of electric power machine room
Xu et al. Safety warning analysis for power battery packs in electric vehicles with running data
Ouyang et al. Prognostics and health management of lithium-ion batteries based on modeling techniques and Bayesian approaches: A review
Zhao et al. Data-driven lithium-ion battery degradation evaluation under overcharge cycling conditions
Singh et al. State of charge estimation techniques of Li-ion battery of electric vehicles
Diao et al. Research on Electric Vehicle Charging Safety Warning based on A-LSTM Algorithm
CN115128468A (en) Chemical energy storage battery PHM undervoltage fault prediction method
Hou et al. Remaining useful cycle life prediction of lithium-ion battery based on TS fuzzy model
CN114720901A (en) Method for estimating health state of lithium ion battery

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