CN110687451A - Error compensation-based method for predicting residual life of lithium battery of support vector machine - Google Patents

Error compensation-based method for predicting residual life of lithium battery of support vector machine Download PDF

Info

Publication number
CN110687451A
CN110687451A CN201910802577.7A CN201910802577A CN110687451A CN 110687451 A CN110687451 A CN 110687451A CN 201910802577 A CN201910802577 A CN 201910802577A CN 110687451 A CN110687451 A CN 110687451A
Authority
CN
China
Prior art keywords
time
svr
lithium battery
training
battery
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201910802577.7A
Other languages
Chinese (zh)
Other versions
CN110687451B (en
Inventor
张永
陈廖格豪
谢林柏
郑英
袁烨
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.)
Wuhan University of Science and Engineering WUSE
Original Assignee
Wuhan University of Science and Engineering WUSE
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 Wuhan University of Science and Engineering WUSE filed Critical Wuhan University of Science and Engineering WUSE
Priority to CN201910802577.7A priority Critical patent/CN110687451B/en
Publication of CN110687451A publication Critical patent/CN110687451A/en
Application granted granted Critical
Publication of CN110687451B publication Critical patent/CN110687451B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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]
    • G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/392—Determining battery ageing or deterioration, e.g. state of health
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/10—Energy storage using batteries

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Tests Of Electric Status Of Batteries (AREA)
  • Secondary Cells (AREA)

Abstract

The invention provides a method for predicting the residual life of a lithium battery of a support vector machine based on error compensation. The method comprises the following specific processes: firstly, a discharge voltage difference health index with equal time intervals is extracted on the basis of the battery capacity to form a mixed health index. Various noises are often included in the lithium ion battery index measurement process, and in order to reduce the influence of the noises, the EEMD algorithm is used for carrying out noise reduction treatment on the lithium ion battery data set; then, using a PSR algorithm to obtain the time delay and embedding dimension of a training set in the process of training the SVR model; and finally, the error correction model is used for reducing SVR prediction errors and improving the prediction accuracy of the RUL. The superiority of the method in the aspect of RUL prediction is proved by a NASA lithium ion battery data set.

Description

Error compensation-based method for predicting residual life of lithium battery of support vector machine
Technical Field
The invention belongs to the technical field of batteries, and relates to a method for predicting the residual life of a lithium battery of a support vector machine based on error compensation.
Background
The battery industry develops to the present, and the lithium ion battery becomes a research hotspot by virtue of the advantages of high voltage, high energy density, low self-discharge rate, long cycle life, good safety performance and the like. Because of various outstanding advantages of lithium batteries, lithium batteries have been widely used in the fields of airplanes, electric vehicles, mobile phones, notebook computers, and even aerospace, etc. However, the performance of the lithium ion battery is reduced with the passage of time, which causes the power equipment and the system to be in failure, reduces the production efficiency, and seriously causes property loss and casualties. For example, in 2013, several boeing 787 airliners were on fire due to lithium ion battery failure resulting in aircraft stopping; in 1999, the failure of the battery inside caused the failure of the space experiment in the united states. Therefore, the accurate prediction of the remaining life of the lithium ion battery plays an important role in lithium ion battery fault Prediction and Health Management (PHM), and the PHM can enable the health management function to have autonomy and reduce human intervention, thereby greatly reducing the cost. The PHM technology estimates the state of the lithium ion battery by detecting various indexes of the lithium ion battery, wherein the residual life performance index is the core, and the safety and the reliability of the lithium ion battery are greatly improved. According to the invention, a PSR-SVR-EC algorithm is adopted, EEMD is firstly used for noise reduction processing on the measured data, the addition of PSR and an error correction idea greatly improves the prediction precision of SVR, and accurate prediction on the residual life of the lithium battery is realized.
Disclosure of Invention
The invention aims to provide a method for accurately predicting the residual service life of a lithium iron phosphate battery.
In order to achieve the above purpose, the solution of the invention is:
the invention provides a method for predicting the residual life of a lithium battery of a support vector machine based on error compensation, which is characterized by comprising the following steps:
the method comprises the following steps: selecting two indexes of the capacity of the lithium battery and the voltage difference of the discharge at equal time intervals as health factors of the lithium battery, selecting the T time as a prediction starting point, namely a time sequence before the T time as a training set, training a regression model to predict the capacity of the battery after the T time,
step two: the integrated empirical mode decomposition (EEMD) algorithm is used for carrying out noise reduction on the health factor,
step three: when training the SVR model, the phase space reconstruction PSR is used to respectively obtain the time delay and embedding dimension of two health factor time sequences, the multivariable phase space reconstruction is carried out, thereby defining the input and output of the training SVR,
step four: the SVR model SVR-EC based on error compensation can effectively reduce the prediction error of SVR, the battery capacity after T time is predicted by using the model established by PSR-SVR-EC, and if the predicted capacity reaches the failure threshold value, the residual service life RUL is the time from T time to the failure threshold value point.
Furthermore, the method for predicting the residual life of the lithium battery of the support vector machine based on the error compensation also has the following characteristics: the selected lithium battery capacity data is a NASA public data set, three groups of the NASA data sets, namely B0005, B0006 and B0007 are selected as experimental data sets, different initial prediction points are selected for different data sets, data before the initial prediction points are used as training sets, and a regression model is trained to predict the future battery capacity.
Furthermore, the method for predicting the residual life of the lithium battery of the support vector machine based on the error compensation also has the following characteristics:
the procedure for PSR using the C-C method was as follows:
Figure BDA0002182746960000021
wherein, N is the data length, m is the embedding dimension, r is the standard deviation, t is the reconstruction time delay, when x is less than 0, θ (x) is 0; when x > 0, θ (x) is 1, the distance between points is represented by the infinite norm of the difference between vectors, the correlation integral belongs to the cumulative distribution function, and represents the probability that the distance between any two points in the phase space is less than r, defining the test statistic:
S(m,N,r,t)＝C(m,N,r,t)-Cm(m,N,r,t)
in the calculation, the time sequence needs to be split into t disjoint subsequences, where t is the reconstruction time delay, that is:
Figure BDA0002182746960000031
calculating S (m, N, r, t) of each sequence by adopting a block average strategy
Figure BDA0002182746960000032
Let N → ∞ then:
Figure BDA0002182746960000033
the local maximum time interval is S (m, r, t) crossing zero or the time point with the minimum difference among all the radiuses r, the points in the reconstruction phase space are closest to the uniform distribution, the minimum value and the maximum value corresponding to the radiuses r are selected, and the difference is defined as:
ΔS1(m,t)＝max{S1(m,rj,t,N)}-min{S1(m,rj,t,N)}
the above formula measures Delta S1Maximum deviation of (m, t) -t from all radii r, so the optimum delay can be taken as Δ S1First local minimum point or Δ S of (m, t) -t1The first zero point of (m, r, t) -t is taken as m-2, 3,4,5, rjI σ/2, i 1,2,3,4, σ is the standard deviation of the time series, calculated as follows:
looking for S1_cor(t) the global minimum point is used to obtain the optimal delay time window tauωIn the C-C method, there is a quantitative relation τ between the delay time window, the delay time and the embedding dimensionωFrom (m-1) τ, the embedding dimension can be derivedA number m.
Furthermore, the method for predicting the residual life of the lithium battery of the support vector machine based on the error compensation also has the following characteristics:
and (3) solving the time delay and embedding dimension of the two HIs after noise reduction by using the C-C method, and performing multivariate phase space reconstruction to define input and output when the SVR is trained:
Figure BDA0002182746960000041
Figure BDA0002182746960000042
the selected SVR kernel function is a radial basis kernel function, and the specific expression is as follows:
Figure BDA0002182746960000043
and predicting the future capacitance by using the trained model, and comparing the future capacitance with the previously defined threshold value to finally estimate the residual service life of the lithium battery. The method and the device can estimate the RUL of the lithium battery more accurately.
Drawings
Fig. 1 is a graph of capacity decay curves and failure thresholds corresponding to three groups of lithium batteries used in the present invention.
Fig. 2 is a flow chart of the error correction concept used in the present invention.
FIG. 3 is a PSR-SVRER flow chart proposed by the present invention.
Fig. 4 is a graph showing the result of the B05 battery data prediction starting point being 100, which is used in the present invention.
Fig. 5 is a graph showing the result of the B06 battery data prediction starting point being 100, which is used in the present invention.
Fig. 6 is a graph showing the result of the B07 battery data prediction starting point being 100, which is used in the present invention.
Detailed Description
The technical solution of the present invention will be described in detail below with reference to the accompanying drawings. Simulations were performed on three sets of battery data to verify the validity of the algorithm.
As shown in FIG. 3, the NASA data set obtained by the present invention is firstly screened, and three groups of B0005, B0006 and B0007 are correspondingly selected as experimental data sets. Two health factors of battery capacitance and equal time interval discharge voltage difference are extracted, and a training set and a test set are divided.
And carrying out EEMD decomposition on the extracted health factors to obtain IMFs and a residual error term Res.
And removing IMFs with large vibration frequency, and recombining the rest IMFs and a residual error term Res to obtain a health factor after noise reduction.
Respectively calculating the time delay tau of two health factors after noise reduction by using a C-C method in a training set1,τ2And the embedding dimension m1,m2Selecting tau as max (tau)1,τ2),m＝max(m1,m2) Performing multivariate phase space reconstruction:
Figure RE-GDA0002304285850000051
the training sample defining SVR is (X)j,Yj) X is input and Y is output.
The error correction idea is processed in such a way that ① error time series are obtained by subtracting the known value and the predicted value in the training part.
② obtaining the error time sequence in the training part, and establishing a model by using a phase space reconstruction-support vector regression method.
③ the error of the test portion is predicted by using the established model.
④ combining the preliminary prediction result and the error prediction result to obtain the final predicted value of the battery capacity.
And predicting the capacitance at the future moment by using a regression model trained by a PSR-SVR-EC algorithm, if the predicted capacitance reaches a defined threshold value, determining the RUL as the time from the prediction starting moment to a battery failure threshold value point, comparing the obtained predicted capacitance value with real capacity data and noise reduction data serving as a test set by using three indexes of RMSE, MAPE and RUL errors, and verifying the effectiveness of the method.
After the scheme is adopted, the lithium battery capacity and the extracted equal-time-interval discharge voltage difference are selected as the lithium battery health factors, and noise reduction processing is carried out on noise contained in the measurement process of each index of the lithium battery by using EEMD; obtaining the time delay tau of the health factor after noise reduction by using a C-C method1,τ2And embedding dimension m1,m2Selecting tau as max (tau)1,τ2),m＝max(m1,m2) Performing multivariate phase space reconstruction; thus defining the inputs and outputs when training the SVR. Finally, accuracy of SVR prediction is greatly improved by adding an error correction idea.
The specific processes of the embodiments of the present invention are detailed below:
the obtained NASA data set is firstly screened, and three groups B0005, B0006 and B0007 are correspondingly selected as experimental data sets. Two health factors of battery capacitance and discharge voltage difference at equal time intervals are extracted, and a training set and a test set are divided.
As shown in FIG. 1, the invention carries out simulation experiments based on three groups of battery data of B0005, B0006 and B0007 in NASA, and verifies the effectiveness of the algorithm on the three groups of battery data.
Adding a random white noise omega to the extracted health factor x (t)i(t)：
xi(t)＝x(t)+ωi(t),i＝1,2,3,...,M。
Wherein x isi(t) is the health factor for the ith trial with random white noise added.
Adding x of random white noise to each experimenti(t) is EMD decomposed into several IMFs components ci,j(t) and residual term ri(t)：
Figure BDA0002182746960000071
Wherein the subscript j represents xiAnd (t) the jth IMF component, L being the number of IMFs decomposed.
Adding random white noise to M xi(t) EMD decomposition was performed, and the results were stored for each time.
Until the M trials are completed, the result of the integrated average of the IMFs and residual terms is:
Figure BDA0002182746960000072
wherein the content of the first and second substances,
Figure BDA0002182746960000073
for the ensemble averaging of the jth IMFs,
Figure BDA0002182746960000074
is the integrated average of the residual terms.
The results of EEMD are expressed as:
Figure BDA0002182746960000075
and removing IMFs with large vibration frequency, and recombining the rest IMFs and a residual error term Res to obtain a health factor after noise reduction.
Respectively calculating the time delay tau of two health factors after noise reduction by using a C-C method in a training set1,τ2And the embedding dimension m1,m2Selecting tau as max (tau)1,τ2),m＝max(m1,m2) Performing multivariate phase space reconstruction:
Figure RE-GDA0002304285850000081
the training sample defining SVR is (X)j,Yj) X is input and Y is output.
As shown in fig. 2, the present invention provides an error correction concept, which comprises the following steps:
① the error time series is derived from the difference between the known and predicted values in the training portion.
② obtaining the error time sequence in the training part, and establishing a model by using a phase space reconstruction-support vector regression method.
③ the error of the test portion is predicted by using the established model.
④ combining the preliminary prediction result and the error prediction result to obtain the final predicted value of the battery capacity.
And predicting the capacitance at the future moment by using a regression model trained by a PSR-SVR-EC algorithm, if the predicted capacitance reaches a defined threshold value, determining the RUL as the time from the prediction starting moment to a battery failure threshold value point, comparing the obtained predicted capacitance value with real capacity data and noise reduction data serving as a test set by using three indexes of RMSE, MAPE and RUL errors, and verifying the effectiveness of the method.
Fig. 4,5 and 6 show the predicted results of three different batteries at 100 as the predicted starting point. As can be seen, the prediction results have a higher accuracy than either the noise reduction data or the raw data.
Tables 1,2 and 3 show that the method of the present invention has good performance for RUL prediction in terms of accuracy index.
The RMSE1 and the like refer to the precision indexes between the prediction result and the noise reduction data, and the RMSE2 and the like refer to the precision indexes between the prediction result and the original data.
Table 1: the invention discloses a prediction accuracy index table when the starting point of three groups of battery data prediction is 100.
Predicting effect when 100 time is starting point
Figure BDA0002182746960000091
Table 2: the other method uses three sets of battery data to predict the prediction accuracy index table when the starting point is 100.
Other method | predicting effect when 100 time is the starting point
Figure BDA0002182746960000092
Table 3: the invention uses three groups of battery data to predict the accuracy index table of the initial point differently.
Prediction effect of different training sample lengths
Figure BDA0002182746960000093
Aiming at the problem of predicting the residual life of the lithium iron phosphate battery, the invention provides a PSR-SVR-EC algorithm, a simulation experiment is realized based on the actual discharge data of the NASA lithium battery, and the effectiveness of the PSR-SVR-EC algorithm in the invention is proved through the experiment.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (4)

1. The method for predicting the residual life of the lithium battery of the support vector machine based on error compensation is characterized by comprising the following steps:
the method comprises the following steps: selecting two indexes of the capacity of the lithium battery and the voltage difference of the discharge at equal time intervals as the health factors of the lithium battery, selecting the T time as a prediction starting point, namely selecting a time sequence before the T time as a training set, training a regression model to predict the capacity of the battery after the T time,
step two: the integrated empirical mode decomposition (EEMD) algorithm is used for carrying out noise reduction on the health factor,
step three: when training the SVR model, using the phase space reconstruction PSR to respectively obtain the time delay and embedding dimension of two health factor time sequences, performing multivariate phase space reconstruction, thereby defining the input and output of the training SVR,
step four: the SVR model SVR-EC based on error compensation can effectively reduce the prediction error of SVR, the battery capacity after T time is predicted by using the model established by PSR-SVR-EC, and if the predicted capacity reaches the failure threshold value, the residual service life RUL is the time from T time to the failure threshold value point.
2. The method for predicting the remaining life of a lithium battery of a support vector machine based on error compensation according to claim 1, wherein:
the selected lithium battery capacity data is NASA public data set,
b0005, B0006 and B0007 in the NASA data set are selected as experimental data sets, different initial prediction points and data before the initial prediction points are selected as training sets for different data sets, and a regression model is trained to predict the future battery capacity.
3. The method for predicting the remaining life of a lithium battery of a support vector machine based on error compensation according to claim 1, wherein:
the procedure for PSR using the C-C method was as follows:
wherein, N is the data length, m is the embedding dimension, r is the standard deviation, t is the reconstruction time delay, when x is less than 0, θ (x) is 0; when x > 0, θ (x) is 1, the distance between points is represented by the infinite norm of the difference between vectors, the correlation integral belongs to the cumulative distribution function, and represents the probability that the distance between any two points in the phase space is less than r, defining the test statistic:
S(m,N,r,t)＝C(m,N,r,t)-Cm(m,N,r,t)
in the calculation, the time sequence is divided into t disjoint subsequences, where t is the reconstruction time delay, that is:
Figure FDA0002182746950000021
calculating S (m, N, r, t) of each sequence by adopting a block average strategy
Figure FDA0002182746950000022
Let N → ∞ then:
the local maximum time interval is taken as the time point when S (m, r, t) crosses the zero point or the difference between all the radiuses r is minimum, the points in the reconstruction phase space are closest to uniform distribution, the minimum value and the maximum value corresponding to the radiuses r are selected, and the difference is defined as:
ΔS1(m,t)＝max{S1(m,rj,t,N)}-min{S1(m,rj,t,N)}
the above formula measures Delta S1(m, t) -t maximum deviation from all radii r, the optimum delay is taken as Δ S1First local minimum point or Δ S of (m, t) -t1The first zero point of (m, r, t) -t is taken as m-2, 3,4,5, rjI σ/2, i 1,2,3,4, σ is the standard deviation of the time series, calculated as follows:
Figure FDA0002182746950000031
Figure FDA0002182746950000032
looking for S1_cor(t) the global minimum point is used to obtain the optimal delay time window tauωIn the C-C method, there is a quantitative relation τ between the delay time window, the delay time and the embedding dimensionωFrom this, the embedding dimension m can be derived (m-1).
4. The method for predicting the remaining life of a lithium battery of a support vector machine based on error compensation according to claim 1, wherein:
and (3) solving the time delay and embedding dimension of the two HIs after noise reduction by using the C-C method, and performing multivariate phase space reconstruction to define input and output when the SVR is trained:
Figure RE-FDA0002304285840000033
Figure RE-FDA0002304285840000034
the selected SVR kernel function is a radial basis kernel function, and the specific expression is as follows:
Figure RE-FDA0002304285840000035
CN201910802577.7A 2019-08-28 2019-08-28 Error compensation-based method for predicting residual life of lithium battery of support vector machine Active CN110687451B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910802577.7A CN110687451B (en) 2019-08-28 2019-08-28 Error compensation-based method for predicting residual life of lithium battery of support vector machine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910802577.7A CN110687451B (en) 2019-08-28 2019-08-28 Error compensation-based method for predicting residual life of lithium battery of support vector machine

Publications (2)

Publication Number Publication Date
CN110687451A true CN110687451A (en) 2020-01-14
CN110687451B CN110687451B (en) 2022-10-25

Family

ID=69108415

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910802577.7A Active CN110687451B (en) 2019-08-28 2019-08-28 Error compensation-based method for predicting residual life of lithium battery of support vector machine

Country Status (1)

Country Link
CN (1) CN110687451B (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111443294A (en) * 2020-04-10 2020-07-24 华东理工大学 Method and device for indirectly predicting remaining life of lithium ion battery
CN111983470A (en) * 2020-08-24 2020-11-24 哈尔滨理工大学 Lithium ion power battery safety degree evaluation method and device based on correlation dimension
CN111999650A (en) * 2020-08-20 2020-11-27 浙江工业大学 Lithium battery residual life prediction method based on support vector regression algorithm
CN112327169A (en) * 2020-11-09 2021-02-05 上海工程技术大学 Lithium battery residual life prediction method
CN112487702A (en) * 2020-10-26 2021-03-12 湖州师范学院 Method for predicting remaining service life of lithium ion battery
CN113111572A (en) * 2021-03-23 2021-07-13 合肥工业大学 Method and system for predicting residual life of aircraft engine
CN113156325A (en) * 2021-03-18 2021-07-23 吉林大学 Method for estimating state of health of battery
CN113281671A (en) * 2021-06-28 2021-08-20 长安大学 Lithium ion battery remaining service life prediction method and system based on IGS-SVM
CN113569460A (en) * 2021-06-08 2021-10-29 北京科技大学 Real vehicle fuel cell system state multi-parameter prediction method and device
US11507922B1 (en) * 2021-06-24 2022-11-22 Coupang Corp. Computer-implemented systems and methods for artificial intelligence (AI)-based inbound plan generation using fungibility logic

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102692241A (en) * 2012-06-11 2012-09-26 华中科技大学 Error compensation method of wireless positioning system
CN102778653A (en) * 2012-06-20 2012-11-14 哈尔滨工业大学 Data-driven lithium ion battery cycle life prediction method based on AR (Autoregressive) model and RPF (Regularized Particle Filtering) algorithm
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
CN106126906A (en) * 2016-06-22 2016-11-16 重庆科技学院 Short-term wind speed forecasting method based on C C Yu ELM
CN107004164A (en) * 2014-11-27 2017-08-01 株式会社博甘斯 The predictive maintenance system and method for equipment
CN109901075A (en) * 2017-12-11 2019-06-18 福特全球技术公司 Method for predicting battery life

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102692241A (en) * 2012-06-11 2012-09-26 华中科技大学 Error compensation method of wireless positioning system
CN102778653A (en) * 2012-06-20 2012-11-14 哈尔滨工业大学 Data-driven lithium ion battery cycle life prediction method based on AR (Autoregressive) model and RPF (Regularized Particle Filtering) algorithm
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
CN107004164A (en) * 2014-11-27 2017-08-01 株式会社博甘斯 The predictive maintenance system and method for equipment
CN106126906A (en) * 2016-06-22 2016-11-16 重庆科技学院 Short-term wind speed forecasting method based on C C Yu ELM
CN109901075A (en) * 2017-12-11 2019-06-18 福特全球技术公司 Method for predicting battery life

Non-Patent Citations (7)

* Cited by examiner, † Cited by third party
Title
G. SANTAMARÍA-BONFIL 等: "Wind speed forecasting for wind farms: A method based on support vector regression", 《RENEWABLE ENERGY》 *
LINXIA LIAO 等: "Review of Hybrid Prognostics Approaches for Remaining Useful Life Prediction of Engineered Systems, and an Application to Battery Life Prediction", 《IEEE TRANSACTIONS ON RELIABILITY》 *
SONG LIANGONG 等: "Support Vector Machine Prediction Model Based on Chaos Theory", 《INTERNATIONAL JOURNAL OF MULTIMEDIA AND UBIQUITOUS ENGINEERING》 *
YANG CHANG 等: "A new hybrid method for the prediction of the remaining useful life of a lithium-ion battery", 《APPLIED ENERGY》 *
印学浩 等: "基于多时间尺度的锂离子电池状态联合估计", 《仪器仪表学报》 *
张洋 等: "锂离子电池剩余寿命在线预测", 《机械科学与技术》 *
韩晓霞 等: "混沌与支持向量机结合的多相催化建模与优化研究", 《中国博士学位论文全文数据库 信息科技辑》 *

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111443294A (en) * 2020-04-10 2020-07-24 华东理工大学 Method and device for indirectly predicting remaining life of lithium ion battery
CN111999650A (en) * 2020-08-20 2020-11-27 浙江工业大学 Lithium battery residual life prediction method based on support vector regression algorithm
CN111983470B (en) * 2020-08-24 2022-06-24 哈尔滨理工大学 Lithium ion power battery safety degree evaluation method and device based on correlation dimension
CN111983470A (en) * 2020-08-24 2020-11-24 哈尔滨理工大学 Lithium ion power battery safety degree evaluation method and device based on correlation dimension
CN112487702A (en) * 2020-10-26 2021-03-12 湖州师范学院 Method for predicting remaining service life of lithium ion battery
CN112487702B (en) * 2020-10-26 2024-04-05 湖州师范学院 Method for predicting residual service life of lithium ion battery
CN112327169A (en) * 2020-11-09 2021-02-05 上海工程技术大学 Lithium battery residual life prediction method
CN112327169B (en) * 2020-11-09 2023-03-24 上海工程技术大学 Lithium battery residual life prediction method
CN113156325A (en) * 2021-03-18 2021-07-23 吉林大学 Method for estimating state of health of battery
CN113156325B (en) * 2021-03-18 2024-04-19 吉林大学 Method for estimating health state of battery
CN113111572A (en) * 2021-03-23 2021-07-13 合肥工业大学 Method and system for predicting residual life of aircraft engine
CN113569460A (en) * 2021-06-08 2021-10-29 北京科技大学 Real vehicle fuel cell system state multi-parameter prediction method and device
US11507922B1 (en) * 2021-06-24 2022-11-22 Coupang Corp. Computer-implemented systems and methods for artificial intelligence (AI)-based inbound plan generation using fungibility logic
CN113281671A (en) * 2021-06-28 2021-08-20 长安大学 Lithium ion battery remaining service life prediction method and system based on IGS-SVM

Also Published As

Publication number Publication date
CN110687451B (en) 2022-10-25

Similar Documents

Publication Publication Date Title
CN110687451B (en) Error compensation-based method for predicting residual life of lithium battery of support vector machine
Liu et al. Prognostics for state of health estimation of lithium-ion batteries based on combination Gaussian process functional regression
He et al. A new model for State-of-Charge (SOC) estimation for high-power Li-ion batteries
CN110687450B (en) Lithium battery residual life prediction method based on phase space reconstruction and particle filtering
US20230029810A1 (en) Battery model construction method and battery degradation prediction device
CN112415414A (en) Method for predicting remaining service life of lithium ion battery
He et al. An improved coulomb counting approach based on numerical iteration for SOC estimation with real-time error correction ability
CN109376892B (en) Equipment state prediction method based on life cycle stage of equipment
Mian Qaisar A proficient Li-ion battery state of charge estimation based on event-driven processing
CN112069727B (en) Intelligent transient stability evaluation system and method with high reliability for power system
CN110806541A (en) AD-BAS-based lithium battery model parameter identification method
CN110308366B (en) Harmonic source positioning method based on orthogonal matching pursuit algorithm
CN116306798A (en) Ultra-short time wind speed prediction method and system
CN113791351B (en) Lithium battery life prediction method based on transfer learning and difference probability distribution
Zhao et al. State of charge estimation of lithium-ion battery based on improved adaptive boosting algorithm
Liu et al. A key-term separation based least square method for Hammerstein SOC estimation model
Hong et al. State-of-health estimation of lithium-ion batteries using a novel dual-stage attention mechanism based recurrent neural network
CN109799454A (en) Generator stator insulation residual breakdown strength prediction technique based on particle group optimizing
CN110991741A (en) Section constraint probability early warning method and system based on deep learning
CN105868544A (en) A method of analyzing insulating material performance by using three-parameter Weibull distribution to process flashover voltages
CN113933718B (en) Retired battery capacity sorting method, retired battery capacity sorting device, retired battery capacity sorting equipment and storage medium
Pan et al. An optimization algorithm for locomotive secondary spring load adjustment based on artificial immune
CN115018133A (en) Lithium battery capacity prediction method based on chaos theory and nonlinear system
Jia et al. An adaptive spherical square‐root double unscented Kalman filtering algorithm for estimating state‐of‐charge of lithium‐ion batteries
CN111126659A (en) Power load prediction method and system

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