CN107479000B - A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation - Google Patents
A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation Download PDFInfo
- Publication number
- CN107479000B CN107479000B CN201710706002.6A CN201710706002A CN107479000B CN 107479000 B CN107479000 B CN 107479000B CN 201710706002 A CN201710706002 A CN 201710706002A CN 107479000 B CN107479000 B CN 107479000B
- Authority
- CN
- China
- Prior art keywords
- battery
- coefficient
- model
- rul
- offline
- 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.)
- Active
Links
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Tests Of Electric Status Of Batteries (AREA)
- Secondary Cells (AREA)
Abstract
The present invention proposes a kind of power battery RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation, it converts battery capacity using Box-Cox transformation, construct the linear model between capacity transformed value and cycle-index, and model parameter and model uncertainty are recognized using least-squares algorithm, the uncertainty application Monte-Carlo Simulation of remaining life generates.The algorithm can reduce online aging data needed for accurate predicting residual useful life, and when there is offline aging data, minimum online data amount required for accurate predicting residual useful life is only the 30% of battery complete attenuation data volume.
Description
Technical field
The present invention relates to the prognostics and health management fields of power battery, more particularly to establish the prediction of power battery
With health status administrative model, and the power battery management system based on established model.
Background technique
Lithium-ion-power cell can pass through its remaining life (Remaining useful in application on site
Life:RUL) prediction cut-off service life (End of Life:EOL).Under normal conditions, when the capacity attenuation of lithium battery holds to initial
Amount 80% when, that is, be considered the cut-off service life of lithium battery.Existing lithium-ion-power cell RUL prediction mainly has based on model
Method and method based on data-driven.Wherein, it usually using non-linear Ageing Model and is combined based on the method for model
Advanced particle filter technology such as particle filter technology predicts RUL, but on service life of lithium battery latter stage capacity attenuation slope
When spending very little, the accuracy of RUL prediction result is lower.It is dynamic to lithium ion that method based on data-driven is based primarily upon machine learning
Power cell degradation data carry out data mining, so that online URL fallout predictor is obtained, however this method is that acquisition is higher accurate
Degree, it usually needs a large amount of offline aging data.
Summary of the invention
For technical problem present in above-mentioned this field, the present invention provides one kind to be converted and be covered based on Box-Cox
The power battery RUL prediction technique of special Caro emulation, method includes the following steps:
Step 1 selects the lithium-ion-power cell of same size to carry out accelerated aging based on the condition that actually works online
Experiment, using the data of acquisition as the offline aging data of lithium-ion-power cell;
Step 2 carries out Box-Cox transformation to the offline aging data in step 1, obtains different dynamic battery
Offline transformation coefficient and capability value linear model;
The offline transformation coefficient of different dynamic battery in step 2 is averaged, and exists as power battery by step 3
The online transformation coefficient is applied to the capacity data observation that part obtains online, and recognizes step 2 by line transformation coefficient
The model coefficient of the capability value linear model of middle acquisition and the variance of the model coefficient.
Step 4, coefficient and the variance according to the model recognized in step 3, is calculated based on Monte Carlo
Method is emulated, and predicts the probability distribution of power battery RUL and RUL.
Further, the Box-Cox transformation carried out in the step 2, using following expression:
For C > 0
Wherein, C represents battery capacity observation, and λ represents the offline transformation coefficient of power battery.
Further, the capability value linear model of the power battery in the step 2, meet C (λ)~N (K β,
σ2) hypothesis, it may be assumed that
Wherein, C (λ)=(C1(λ),C2(λ),…,Cn(λ))T, λ represents the offline transformation coefficient of power battery, and K is design
Matrix meets K=(K1,K2,…,Kn)T, and Ki=(1, ki), β=(β0,β1)T, n is size, kiBattery is represented to follow
Number of rings, β0,β1Representative model coefficient, εiFor the independent random error for meeting normal distribution, mean value 0, variance σ2。
Further, the offline transformation coefficient of the power battery is obtained using maximum likelihood estimate, is taken so that following
Offline transformation coefficient λ of the maximum λ of formula intermediate value as power battery:
Wherein L*(λ) represents log-likelihood function,Expression formula is as follows:
Expression formula it is as follows:
It willAndExpression formula bring expression formula (3) into, that is, acquire corresponding λ value.
Further, the online transformation coefficient is applied to the capacity that part obtains online described in the step 3
Data, and the model coefficient of the capability value linear model obtained in step 2 and the variance of the model coefficient are recognized,
It specifically includes:
Based on the offline transformation coefficient λ of the power battery obtained in the step 2 to the capacity partially obtained online
Data observation value is converted, and utilizes least square method computation model factor beta0,β1:
Wherein
Model coefficient β0,β1Variance are as follows:
Wherein s2Representative errors item variances sigma2Estimated value: s2=SSR/ (n-2);SSR is the quadratic sum of residual error, expression formula
It is as follows:
Further, it is emulated in the step 4 based on Monte Carlo EGS4 method, predicts power battery RUL and RUL
Probability distribution, specifically include: multistep forward prediction carried out to the capability value linear model, for each simulation sample,
When the capacity transformed value of prediction is less than defined stale value, that is, think that battery reaches the cut-off service life, to export RUL;By institute
It states all RUL values that emulation obtains and carries out probability statistics acquisition RUL prediction and its probability distribution.
Based on method provided by aforementioned present invention, have following many utility model has the advantages that it can effectively eliminate battery declines
Subtract the lesser capacity attenuation situation of the latter stage gradient, is obviously improved the precision of prediction of battery RUL.Model is based on relative to existing
Or the method for data-driven, online data amount needed for constructing accurate linear Ageing Model and initialization Ageing Model from
Line data volume is all substantially reduced.
Detailed description of the invention
Fig. 1 is the RUL prediction principle figure of method provided by the present invention
Fig. 2 is lithium-ion-power cell accelerated life test data
Fig. 3 is the RUL prediction result of the battery A based on preceding 30% aging data
Specific embodiment
Method provided by the present invention is made with reference to the accompanying drawing and further illustrates and explains in detail.
As shown in Figure 1, a kind of power battery RUL based on Box-Cox transformation and Monte-Carlo Simulation of the invention is pre-
Survey method, method includes the following steps:
Step 1 selects the lithium-ion-power cell of same size to carry out accelerated aging based on the condition that actually works online
Experiment, using the data of acquisition as the offline aging data of lithium-ion-power cell;
Step 2 carries out Box-Cox transformation to the offline aging data in step 1, obtains different dynamic battery
Offline transformation coefficient and capability value linear model;
The offline transformation coefficient of different dynamic battery in step 2 is averaged, and exists as power battery by step 3
The online transformation coefficient is applied to the capacity data observation that part obtains online, and recognizes step 2 by line transformation coefficient
The model coefficient of the capability value linear model of middle acquisition and the variance of the model coefficient.
Step 4, coefficient and the variance according to the model recognized in step 3, is calculated based on Monte Carlo
Method is emulated, and predicts the probability distribution of power battery RUL and RUL.
In the preferred embodiment of the application, the Box-Cox carried out in the step 2 is converted, using as follows
Expression formula:
For C > 0
Wherein, C represents battery capacity observation, and λ represents the offline transformation coefficient of power battery.
In the preferred embodiment of the application, the linear mould of capability value of the power battery in the step 2
Type meets C (λ)~N (K β, σ2) hypothesis, it may be assumed that
Wherein, C (λ)=(C1(λ),C2(λ),…,Cn(λ))T, λ represents the offline transformation coefficient of power battery, and K is design
Matrix meets K=(K1,K2,…,Kn)T, and Ki=(1, ki), β=(β0,β1)T, n is size, kiBattery is represented to follow
Number of rings, β0,β1Representative model coefficient, εiFor the independent random error for meeting normal distribution, mean value 0, variance σ2。
In the preferred embodiment of the application, the offline transformation coefficient of the power battery uses maximal possibility estimation
Method obtains, and takes so that offline transformation coefficient λ of the maximum λ of following formula intermediate value as power battery:
Wherein L*(λ) represents log-likelihood function,Expression formula is as follows:
Expression formula it is as follows:
It willAndExpression formula bring expression formula (3) into, that is, acquire corresponding λ value.
In the preferred embodiment of the application, the online transformation coefficient is applied to described in the step 3
The capacity data that part obtains online, and recognize model coefficient and the institute of the capability value linear model obtained in step 2
The variance for stating model coefficient, specifically includes:
Based on the offline transformation coefficient λ of the power battery obtained in the step 2 to the capacity partially obtained online
Data observation value is converted, and utilizes least square method computation model factor beta0,β1:
Wherein
Model coefficient β0,β1Variance are as follows:
Wherein s2Representative errors item variances sigma2Estimated value: s2=SSR/ (n-2);SSR is the quadratic sum of residual error, expression formula
It is as follows:
It in the preferred embodiment of the application, is emulated, is predicted based on Monte Carlo EGS4 method in the step 4
The probability distribution of power battery RUL and RUL, specifically include: multistep forward prediction is carried out to the capability value linear model, it is right
In each simulation sample, when the capacity transformed value of prediction is less than defined stale value, that is, think that battery reaches the cut-off service life,
To export RUL;All RUL values that the emulation obtains are subjected to probability statistics and obtain RUL prediction and its probability distribution.
Fig. 2 and Fig. 3 show the power battery RUL based on the condition that actually works online to predict example.Fig. 2 is selected lithium
The experimental data of ion battery, experiment condition are as follows: experimental temperature is 25 DEG C;Constant current constant voltage is full of, i.e. constant current 0.5C fills
The supreme blanking voltage of electricity, then constant-voltage charge is until cut-off current 0.05C;Constant current 1C is discharged to lower blanking voltage, then with
0.5C size is discharged to lower blanking voltage again, and all capability values released are added the capability value as each cycle battery.Four
A battery number is respectively battery A, battery A1, battery A2, battery A3, and wherein battery A is the battery to work online, battery A1-
A3 is as the battery to work offline.It can be seen that battery A is smaller in the capacity attenuation gradient of end of lifetime.Table 1 shows four batteries
Transformation coefficient λ value,
Battery | A | A1 | A2 | A3 |
λ | -6.534 | -5.987 | -6.612 | -6.185 |
The λ value of battery A is the smallest in 4 batteries.Battery A1-A3 be offline battery, λ value be it is known that battery A be
Line working battery, since battery A only has part aging data it is known that therefore its λ value is unknown, estimated value is other three batteries
The mean value of λ, i.e. the λ estimated value of battery A are -6.113, and 0.421. bigger than the true λ value of battery A assumes that battery A only has 30%
For aging data it is known that being now to predict its remaining life, Fig. 3 shows the battery obtained based on preceding 30% aging data
The remaining life estimation condition of A, the lesser capacity attenuation of the end of lifetime gradient have avoided.λ of the capacity conversion value based on battery A
Estimated value converts to obtain to its capacity observation as Box-Cox, and preceding 30% data (150 circulations) are as the capacity acquired
Value, capacity predicted value are to carry out linear fit based on known capacity change data to obtain, it is seen that the mistake of predicted value and conversion value
Difference is smaller, and battery EOL predicted value is 512 circulations, more 12 circulations of than true value 500 circulations, so RUL is predicted at this time
Error is only -12 circulations, and the confidence interval of EOL prediction 95% is [502,522], it is shown that higher prediction accuracy and essence
Degree.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (3)
1. a kind of power battery RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation, it is characterised in that: tool
Body the following steps are included:
Step 1 selects the lithium-ion-power cell of same size to carry out accelerated life test based on the condition that actually works online,
Using the data of acquisition as the offline aging data of lithium-ion-power cell;
Step 2 carries out Box-Cox transformation to the offline aging data in step 1, obtains the offline of different dynamic battery
Transformation coefficient and capability value linear model;
The offline transformation coefficient of different dynamic battery in step 2 is averaged, and becomes online as power battery by step 3
Coefficient is changed, the online transformation coefficient is applied to the capacity data observation that part obtains online, and recognize in step 2 and obtain
The model coefficient of the capability value linear model taken and the variance of the model coefficient;
Step 4, coefficient and the variance according to the model recognized in step 3, based on Monte Carlo EGS4 method into
Row emulation, predicts the probability distribution of power battery RUL and RUL;
The Box-Cox transformation carried out in the step 2, using following expression:
For C > 0
Wherein, C represents battery capacity observation, and λ represents the offline transformation coefficient of power battery;
The capability value linear model of the power battery in the step 2 meets C (λ)~N (K β, σ2) hypothesis, it may be assumed that
Wherein, C (λ)=(C1(λ),C2(λ),…,Cn(λ))T, λ represents the offline transformation coefficient of power battery, and K is design matrix,
Meet K=(K1,K2,…,Kn)T, and Ki=(1, ki), β=(β0,β1)T, n is size, kiCirculating battery number is represented,
β0,β1Representative model coefficient, εiFor the independent random error for meeting normal distribution, mean value 0, variance σ2;
The online transformation coefficient is applied to the capacity data that part obtains online described in the step 3, and recognizes step
The model coefficient of the capability value linear model obtained in rapid two and the variance of the model coefficient, specifically include:
Based on the offline transformation coefficient λ of the power battery obtained in the step 2 to the capacity data partially obtained online
Observation is converted, and utilizes least square method computation model factor beta0,β1:
Wherein
Model coefficient β0,β1Variance are as follows:
Wherein s2Representative errors item variances sigma2Estimated value: s2=SSR/ (n-2);SSR is the quadratic sum of residual error, and expression formula is as follows:
2. the method as described in claim 1, it is characterised in that: the offline transformation coefficient of the power battery uses maximum likelihood
The estimation technique obtains, and takes so that offline transformation coefficient λ of the maximum λ of following formula intermediate value as power battery:
Wherein L*(λ) represents log-likelihood function,Expression formula is as follows:
Expression formula it is as follows:
It willAndExpression formula bring expression formula (8) into, that is, acquire corresponding λ value.
3. the method as described in claim 1, it is characterised in that: it is emulated in the step 4 based on Monte Carlo EGS4 method,
The probability distribution for predicting power battery RUL and RUL, specifically includes: it is pre- forward to carry out multistep to the capability value linear model
It surveys, when the capacity transformed value of prediction is less than defined stale value battery, which reaches cut-off, to be thought for each simulation sample
Service life, to export RUL;All RUL values that the emulation obtains are subjected to probability statistics and obtain RUL prediction and its probability point
Cloth.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710706002.6A CN107479000B (en) | 2017-08-17 | 2017-08-17 | A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710706002.6A CN107479000B (en) | 2017-08-17 | 2017-08-17 | A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107479000A CN107479000A (en) | 2017-12-15 |
CN107479000B true CN107479000B (en) | 2019-10-11 |
Family
ID=60600718
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710706002.6A Active CN107479000B (en) | 2017-08-17 | 2017-08-17 | A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107479000B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110110365A (en) * | 2019-04-01 | 2019-08-09 | 华南理工大学 | A kind of battery RUL prediction technique based on battery capacity decline track transformation to linearity |
CN110687450B (en) * | 2019-08-28 | 2022-10-25 | 武汉科技大学 | Lithium battery residual life prediction method based on phase space reconstruction and particle filtering |
CN113567876B (en) * | 2020-04-28 | 2024-05-28 | 北京新能源汽车股份有限公司 | Power supply testing method, device and equipment |
CN111882130B (en) * | 2020-07-30 | 2022-01-11 | 浙江大学 | Online dioxin emission prediction method based on generation path clustering and Box-Cox transformation |
CN113011012B (en) * | 2021-03-02 | 2023-11-28 | 傲普(上海)新能源有限公司 | Box-Cox change-based energy storage battery residual life prediction method |
CN113359048A (en) * | 2021-04-28 | 2021-09-07 | 中国矿业大学 | Indirect prediction method for remaining service life of lithium ion battery |
CN113391211B (en) * | 2021-06-11 | 2022-04-19 | 电子科技大学 | Method for predicting remaining life of lithium battery under small sample condition |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003329551A (en) * | 2002-05-15 | 2003-11-19 | Sysmex Corp | Price-putting method for standard sample |
CN103293487A (en) * | 2013-06-28 | 2013-09-11 | 哈尔滨工业大学 | Lithium ion battery service life forecasting method based on integrated model |
CN103675637A (en) * | 2013-11-14 | 2014-03-26 | 南京航空航天大学 | Power MOSFET health state assessment and residual life prediction method |
CN105445671A (en) * | 2015-12-29 | 2016-03-30 | 北京航天测控技术有限公司 | Lithium ion battery service life prediction method based on traceless particle filtering |
CN106908737A (en) * | 2017-03-31 | 2017-06-30 | 清远佳致新材料研究院有限公司 | A kind of lithium ion battery life-span prediction method based on electrochemical reaction mechanism emulation |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030018633A1 (en) * | 2001-03-30 | 2003-01-23 | University Of Cincinnati | Robust reference interval estimator |
-
2017
- 2017-08-17 CN CN201710706002.6A patent/CN107479000B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003329551A (en) * | 2002-05-15 | 2003-11-19 | Sysmex Corp | Price-putting method for standard sample |
CN103293487A (en) * | 2013-06-28 | 2013-09-11 | 哈尔滨工业大学 | Lithium ion battery service life forecasting method based on integrated model |
CN103675637A (en) * | 2013-11-14 | 2014-03-26 | 南京航空航天大学 | Power MOSFET health state assessment and residual life prediction method |
CN105445671A (en) * | 2015-12-29 | 2016-03-30 | 北京航天测控技术有限公司 | Lithium ion battery service life prediction method based on traceless particle filtering |
CN106908737A (en) * | 2017-03-31 | 2017-06-30 | 清远佳致新材料研究院有限公司 | A kind of lithium ion battery life-span prediction method based on electrochemical reaction mechanism emulation |
Non-Patent Citations (4)
Title |
---|
A Health Indicator Extraction and Optimization Framework for Lithium-Ion Battery Degradation Modeling and Prognostics;Datong Liu 等;《IEEE Transactions on Systems, Man, and Cybernetics: Systems》;20150630;第45卷(第6期);第915-928页 * |
基于RVM的锂离子电池剩余寿命预测方法研究;周建宝;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20150115(第01期);C042-30 * |
车用锂离子电池剩余使用寿命预测方法;王常虹 等;《汽车工程》;20151231;第37卷(第4期);第476-479页 * |
锂离子电池健康状态估计与剩余寿命预测;董汉成 等;《北京理工大学学报》;20151031;第35卷(第10期);第1074-1078页 * |
Also Published As
Publication number | Publication date |
---|---|
CN107479000A (en) | 2017-12-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107479000B (en) | A kind of lithium-ion-power cell RUL prediction technique based on Box-Cox transformation and Monte-Carlo Simulation | |
CN106772080B (en) | Time equivalence modeling method for accelerated degradation test of space lithium ion battery | |
Liu et al. | A health indicator extraction and optimization framework for lithium-ion battery degradation modeling and prognostics | |
Liu et al. | An integrated probabilistic approach to lithium-ion battery remaining useful life estimation | |
Orchard et al. | Risk measures for particle-filtering-based state-of-charge prognosis in lithium-ion batteries | |
He et al. | Prognostics of lithium-ion batteries based on Dempster–Shafer theory and the Bayesian Monte Carlo method | |
Zhou et al. | A fast screening framework for second-life batteries based on an improved bisecting K-means algorithm combined with fast pulse test | |
CN103926536A (en) | Method for predicting residual service life of lithium ion battery on basis of DST and BMC technologies | |
Yan et al. | Uncertainty management in Lebesgue-sampling-based diagnosis and prognosis for lithium-ion battery | |
Tang et al. | The capacity prediction of Li-ion batteries based on a new feature extraction technique and an improved extreme learning machine algorithm | |
CN109917292B (en) | Lithium ion battery life prediction method based on DAUPF | |
Selvaraj et al. | A comprehensive review of state of charge estimation in lithium-ion batteries used in electric vehicles | |
CN110095731B (en) | Method for directly predicting residual life of lithium ion battery applied to long-life space | |
Gasper et al. | Machine-learning assisted identification of accurate battery lifetime models with uncertainty | |
Cui et al. | State of health diagnosis and remaining useful life prediction for lithium-ion battery based on data model fusion method | |
CN113504483B (en) | Integrated prediction method for residual life of lithium ion battery considering uncertainty | |
CN106372272B (en) | Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis | |
Xia et al. | Lithium-ion battery ageing behavior pattern characterization and state-of-health estimation using data-driven method | |
CN113820604B (en) | Lithium battery SOH estimation method based on temperature prediction | |
CN109633470B (en) | Estimation method for battery real-time full charge time based on EKF-GPR and daily segment data | |
CN114839538A (en) | Method for extracting degradation characteristics of lithium ion battery for estimating residual life | |
Huang et al. | State of health prediction of lithium-ion batteries based on autoregression with exogenous variables model | |
CN109075397A (en) | The method and apparatus of electrochemical energy accumulator is optimally used for the service life | |
De Sousa et al. | Comparison of different approaches for lead acid battery state of health estimation based on artificial neural networks algorithms | |
Vilsen et al. | Transfer learning for adapting battery state-of-health estimation from laboratory to field operation |
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 |