CN111999649A - XGboost algorithm-based lithium battery residual life prediction method - Google Patents

XGboost algorithm-based lithium battery residual life prediction method Download PDF

Info

Publication number
CN111999649A
CN111999649A CN202010845212.5A CN202010845212A CN111999649A CN 111999649 A CN111999649 A CN 111999649A CN 202010845212 A CN202010845212 A CN 202010845212A CN 111999649 A CN111999649 A CN 111999649A
Authority
CN
China
Prior art keywords
model
data
lithium battery
cart
xgboost
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
CN202010845212.5A
Other languages
Chinese (zh)
Inventor
马正阳
刘明威
程钏
徐凡
娄维尧
杨克允
沈伟健
林韩波
蔡姚杰
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.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
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 Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN202010845212.5A priority Critical patent/CN111999649A/en
Publication of CN111999649A publication Critical patent/CN111999649A/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]
    • G01R31/3644—Constructional arrangements
    • G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/385—Arrangements for measuring battery or accumulator variables
    • G01R31/387—Determining ampere-hour charge capacity or SoC
    • G01R31/388—Determining ampere-hour charge capacity or SoC involving voltage measurements
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/392—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]
    • G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Secondary Cells (AREA)

Abstract

The invention discloses a method for predicting the residual life of a lithium battery based on an XGboost algorithm, which comprises the following steps of: 1) feature extraction: monitoring data generated in the operation process of the lithium battery, extracting voltage time sequence data of lithium battery charging from the data, and performing feature generation on the extracted voltage time sequence data according to needs; 2) constructing a model: establishing an XGboost model based on a decision tree, wherein the decision tree adopts a classification and regression tree CART; continuously generating CARTs according to the training data to fit the residual error generated by the last CART, and finally integrating all the CARTs to obtain a final XGboost integration model; 3) training and predicting: and sending the extracted features into a decision tree CART-based XGboost model for training, and using the trained model for predicting the RUL of the lithium battery. According to the invention, by introducing an integration algorithm and taking CART in a decision tree algorithm as a basic learner, a series of CART are trained and integrated, so that the prediction performance of the residual service life of the lithium battery is improved.

Description

XGboost algorithm-based lithium battery residual life prediction method
Technical Field
The invention belongs to the technical field of lithium battery residual life prediction, and particularly relates to a method for predicting the residual life of a lithium battery based on an XGboost algorithm.
Background
With the global energy crisis getting worse and environmental protection being more and more important, people prefer to select electric vehicles for travel to reduce the use of fossil energy and reduce carbon emission. The lithium ion battery has the advantages of high specific energy, high working voltage, wide temperature range, low self-charging rate, long cycle life, good safety and the like, and is the most common power source for the electric automobile. However, as the number of charging and discharging cycles increases, the capacity of the lithium ion battery may undergo oscillatory attenuation, and the battery itself may age, which eventually leads to the degradation of the battery performance until the battery is discarded.
At present, two types of prediction methods for the residual life (RUL) of the lithium battery exist, one is a method based on a model, the prediction is carried out by modeling through a physical or chemical method, but the model is difficult to establish due to the complicated electrochemical characteristics in the battery; the other is a data-driven method, which can estimate the remaining life of the battery by using the monitoring parameters and sample data of the battery and with the help of a related algorithm. However, the existing data driving method mostly adopts a single algorithm for prediction, and because the performance of the single algorithm is limited, it is generally difficult to obtain a high-quality prediction result.
The integration algorithm combines a plurality of basic algorithms to make up for deficiencies, so that the comprehensive performance of the basic algorithms is improved. There are many types of integration algorithms, wherein the Boosting type of integration algorithm obtains a series of weaker basis learners through continuous training, and then combines the basis learners into a strong learner. The extreme gradient lifting tree algorithm (XGboost) provided by Chentianqi is used as a Boosting type integration algorithm, a large number of decision tree models can be integrated, and the accuracy of model prediction results is greatly improved.
The existing data driving method mostly adopts data such as voltage, current and the like generated in the battery operation process as auxiliary information to predict the remaining life of the battery. According to the method, the charging voltage information generated in the operation process of the lithium battery is learned through the XGboost algorithm based on the decision tree, the correlation between the charging voltage information and the residual life is constructed, and the prediction of the residual life of the lithium battery based on the XGboost algorithm is realized.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a lithium battery residual life prediction method based on an XGboost algorithm.
The technical scheme of the invention is as follows:
a lithium battery residual life prediction method based on an XGboost algorithm is characterized by comprising the following steps:
1) feature extraction: monitoring data generated in the operation process of the lithium battery, extracting voltage time sequence data of lithium battery charging from the data, performing feature generation on the extracted voltage time sequence data according to needs, and sorting voltage change data in the charging process according to the principle of equal voltage difference;
2) constructing a model: establishing an XGboost model based on a decision tree, wherein the decision tree CART adopts classification and regression trees; continuously generating CARTs according to the training data to fit the residual error generated by the last CART, and finally integrating all the CARTs to obtain a final XGboost integration model;
3) training and predicting: and sending the extracted features into a decision tree CART-based XGboost model for training, and using the trained model for predicting the RUL of the lithium battery.
The method for predicting the remaining life of the lithium battery based on the XGboost algorithm is characterized in that the step 1) specifically comprises the following steps: respectively selecting multiple groups of charging time sequence data (V) with equal voltage difference1,V2) And obtaining a plurality of groups of charging time sequence characteristics with equal voltage difference after data processing.
The method for predicting the remaining life of the lithium battery based on the XGboost algorithm is characterized in that the step 2) specifically comprises the following steps: firstly, generating a CART tree, wherein the CART decision tree is a binary tree, and nodes are divided into two types: internal nodes and leaf nodes, wherein a leaf node has no children; if X and Y are assumed to represent input and output variables, respectively, the training data set can be represented as:
D＝{(x1,y1),(x2,y2),…,(xN,yN)} (1)
if a constructed CART regression tree has K leaves, it means that CART divides the input space into K cells: r1，R2，···，RKAnd each unit RKCorresponding to a fixed output value ckSequentially traversing all segmentation variables j and segmentation points s, and finding an optimal segmentation pair (j, s) which enables the following formula value to be minimum while segmenting the sample space:
Figure BDA0002642816580000021
wherein, c1And c2Respectively representing the output values of the data set after the data set is divided into two parts;
finally establishing a CART model by traversing the values of the characteristic variables and carrying out segmentation of the data set;
the mechanism of the XGboost algorithm is that a new CART tree is continuously trained to fit a residual error calculated by a previous tree, then the predicted values of all decision trees are added to obtain a final output result, and the result of the t-th training iteration is expressed as:
Figure BDA0002642816580000031
wherein the content of the first and second substances,
Figure BDA0002642816580000032
the final model is represented as a result of the model,
Figure BDA0002642816580000033
represents the sum of all previously generated models, ft(xi) Representing the newly generated tree model, t representing the total number of generated base learners;
the XGboost controls overfitting through a regularization method, and the objective function of the XGboost is as follows:
Figure BDA0002642816580000034
Figure BDA0002642816580000035
wherein the objective function J(t)The first item in the method is a loss function of the model, measures the difference between a predicted value and a true value and reflects the capability of the model for fitting training data; omega in the second term is the complexity function of the model, where gamma is the penalty factor for the leaf node, pkIs the number of leaf nodes of the tree, wkIs the weight value of the leaf node, λ is the regularization coefficient of the leaf node weight;
the final objective function can be derived and simplified by the following second order taylor expansion from equation (4):
Figure BDA0002642816580000036
Figure BDA0002642816580000037
Figure BDA0002642816580000038
wherein, giAnd hiFirst and second derivatives of the loss function, respectively;
and continuously training the XGboost algorithm to produce a new decision tree model, accumulating the prediction results of the residual error, and finally integrating all generated decision tree models.
The method for predicting the remaining life of the lithium battery based on the XGboost algorithm is characterized in that the step 1) specifically comprises the following steps:
the selected characteristic data is processed, and the magnitude order and dimension of the selected characteristic data are greatly different from the magnitude order and dimension of the residual capacity of the lithium battery, so that the extracted characteristic data needs to be standardized, the model is convenient to process and operate, and a specific formula is as follows:
Figure BDA0002642816580000041
wherein x' is the normalized data, x is the raw data that is not normalized, μ is the mean of the data, and σ is the standard deviation of the data;
establishing a lithium battery RUL prediction model by using a decision tree CART-based XGboost algorithm and selected feature data, and evaluating a prediction result of the model by using a mean square error RMSE;
Figure BDA0002642816580000042
wherein the content of the first and second substances,
Figure BDA0002642816580000043
representing true data, yiThe output of the model is represented, and k represents the number of samples contained in the test set.
The invention has the beneficial effects that: according to the invention, an integration algorithm is introduced, CART in a decision tree algorithm is used as a basic learner, a series of CARTs are trained and integrated, and the constructed XGboost integration model improves the prediction performance of the residual service life of the lithium battery by performing learning training on voltage information in the charging process of the lithium battery.
Drawings
FIG. 1 is a graph comparing the predicted value and the authenticity of the prediction of the RUL of the lithium battery according to the present invention.
Detailed Description
The invention is further described with reference to the drawings and examples.
As shown in fig. 1: a lithium battery residual life prediction method based on an XGboost algorithm comprises the following steps:
1) feature extraction:
monitoring data generated in the operation process of the lithium battery, extracting voltage time sequence data of lithium battery charging from the data, performing feature generation on the extracted voltage time sequence data according to needs, and sorting voltage change data in the charging process according to the principle of equal voltage difference.
Specifically, a plurality of groups of charging time sequence data (V) with equal voltage difference are respectively selected1,V2) And obtaining a plurality of groups of charging time sequence characteristics with equal voltage difference after data processing.
2) Constructing a model:
an XGboost model based on a decision tree is established, and the decision tree adopts a classification and regression tree (CART). And continuously generating CARTs according to the training data to fit the residual error generated by the last CART, and finally integrating all the CARTs to obtain the final XGboost integration model. The method comprises the following steps:
firstly, generating a CART tree, wherein the CART decision tree is a binary tree, and nodes are divided into two types: internal nodes and leaf nodes, where a leaf node has no children. If X and Y are assumed to represent input and output variables, respectively, the training data set can be represented as:
D＝{(x1,y1),(x2,y2),…,(xN,yN)} (1)
if a constructed CART regression tree has K leaves, it means that CART divides the input space into K cells: r1，R2，···，RKAnd each unit RKCorresponding to a fixed output value ck. Sequentially traversing all segmentation variables j and segmentation points s, and finding an optimal segmentation pair (j, s) which enables the following formula value to be minimum while segmenting the sample space:
Figure BDA0002642816580000051
wherein, c1And c2Respectively representing the output values of the data set after the data set is divided into two parts.
And finally establishing a CART model by traversing the values of the characteristic variables and carrying out segmentation of the data set.
The mechanism of the XGBoost algorithm is to continuously train a new CART tree to fit a residual calculated from a previous tree, then add up the predicted values of all decision trees to obtain a final output result, and the result of the t-th training iteration can be expressed as:
Figure BDA0002642816580000052
wherein the content of the first and second substances,
Figure BDA0002642816580000053
the final model is represented as a result of the model,
Figure BDA0002642816580000054
represents the sum of all previously generated models, ft(xi) Representing the newly generated tree model and t representing the total number of generated basis learners.
The XGboost controls overfitting through a regularization method, and the objective function of the XGboost is as follows:
Figure BDA0002642816580000055
Figure BDA0002642816580000056
wherein the objective function J(t)The first item in the method is a loss function of the model, measures the difference between a predicted value and a true value and reflects the capability of the model for fitting training data; omega in the second term is the complexity function of the model, where gamma is the penalty factor for the leaf node, pkIs the number of leaf nodes of the tree, wkIs the weight value of the leaf node, and λ is the regularization coefficient of the leaf node weight.
The final objective function can be derived and simplified by the following second order taylor expansion from equation (4):
Figure BDA0002642816580000061
Figure BDA0002642816580000062
Figure BDA0002642816580000063
wherein, giAnd hiThe first and second derivatives of the loss function are respectively.
And continuously training the XGboost algorithm to produce a new decision tree model, accumulating the prediction results of the residual error, and finally integrating all generated decision tree models.
3) Training and predicting:
and sending the extracted features into a decision tree CART-based XGboost model for training, and using the trained model for predicting the RUL of the lithium battery.
The specific method is to process the selected characteristic data, and because the magnitude order and the dimension of the selected characteristic data are greatly different from the magnitude order and the dimension of the residual capacity of the lithium battery, the extracted characteristic data need to be standardized, so that the model can be conveniently processed and operated, and the specific formula is shown as follows
Figure BDA0002642816580000064
Where x' is the normalized data, x is the raw data that was not normalized, μ is the mean of the data, and σ is the standard deviation of the data.
And establishing a RUL prediction model of the lithium battery by using the selected characteristic data through an XGboost algorithm based on the CART decision tree, and evaluating the prediction result of the model by using a mean square error (RMSE).
Figure BDA0002642816580000065
Wherein the content of the first and second substances,
Figure BDA0002642816580000066
representing true data, yiThe output of the model is represented, and k represents the number of samples contained in the test set. Example (b):
1) feature extraction
Monitoring data generated in the operation process of the lithium battery, extracting voltage time sequence data of lithium battery charging from the data, and selecting an isoelectronic voltage difference charging time sequence in a range of (3.5V, 3.6V) as an input characteristic of the model according to the principle of equal voltage difference according to experience.
2) Building models
The method comprises the steps of constructing an XGboost integration model by taking CART as a base learner, setting input characteristics as one-dimensional characteristics of charging time with equal voltage difference, and outputting the one-dimensional characteristics as the residual service life of a lithium battery.
3) Training and predicting
The equal voltage difference time series characteristics of the selected charging voltage (3.5V, 3.6V) interval are fed into the model for training, and the prediction results are evaluated using RMSE as an index. Table 1 shows the evaluation of the model prediction results, and fig. 1 shows the comparison of the model prediction results with the true values. It can be seen that the model can accurately predict the trend of the lithium battery RUL.
TABLE 1 model prediction result evaluation
Figure BDA0002642816580000071

Claims (4)

1. A lithium battery residual life prediction method based on an XGboost algorithm is characterized by comprising the following steps:
1) feature extraction: monitoring data generated in the operation process of the lithium battery, extracting voltage time sequence data of lithium battery charging from the data, performing feature generation on the extracted voltage time sequence data according to needs, and sorting voltage change data in the charging process according to the principle of equal voltage difference;
2) constructing a model: establishing an XGboost model based on a decision tree, wherein the decision tree CART adopts classification and regression trees; continuously generating CARTs according to the training data to fit the residual error generated by the last CART, and finally integrating all the CARTs to obtain a final XGboost integration model;
3) training and predicting: and sending the extracted features into a decision tree CART-based XGboost model for training, and using the trained model for predicting the RUL of the lithium battery.
2. The XGboost algorithm-based lithium battery residual life prediction method as claimed in claim 1, wherein the step 1) specifically comprises: respectively selecting multiple groups of charging time sequence data (V) with equal voltage difference1,V2) And obtaining a plurality of groups of charging time sequence characteristics with equal voltage difference after data processing.
3. The XGboost algorithm-based lithium battery residual life prediction method as claimed in claim 1, wherein the step 2) specifically comprises: firstly, generating a CART tree, wherein the CART decision tree is a binary tree, and nodes are divided into two types: internal nodes and leaf nodes, wherein a leaf node has no children; if X and Y are assumed to represent input and output variables, respectively, the training data set can be represented as:
D＝{(x1,y1),(x2,y2),…,(xN,yN)} (1)
if a constructed CART regression tree has K leaves, it means that CART divides the input space into K cells: r1，R2，···，RKAnd each unit RKCorresponding to a fixed output value ckSequentially traversing all segmentation variables j and segmentation points s, and finding an optimal segmentation pair (j, s) which enables the following formula value to be minimum while segmenting the sample space:
Figure FDA0002642816570000011
wherein, c1And c2Respectively representing the output values of the data set after the data set is divided into two parts;
finally establishing a CART model by traversing the values of the characteristic variables and carrying out segmentation of the data set;
the mechanism of the XGboost algorithm is that a new CART tree is continuously trained to fit a residual error calculated by a previous tree, then the predicted values of all decision trees are added to obtain a final output result, and the result of the t-th training iteration is expressed as:
Figure FDA0002642816570000021
wherein the content of the first and second substances,
Figure FDA0002642816570000022
the final model is represented as a result of the model,
Figure FDA0002642816570000023
represents the sum of all previously generated models, ft(xi) Representing the newly generated tree model, t representing the total number of generated base learners;
the XGboost controls overfitting through a regularization method, and the objective function of the XGboost is as follows:
Figure FDA0002642816570000024
Figure FDA0002642816570000029
wherein the objective function J(t)The first item in the method is a loss function of the model, measures the difference between a predicted value and a true value and reflects the capability of the model for fitting training data; omega in the second term is the complexity function of the model, where gamma is the penalty factor for the leaf node, pkIs the number of leaf nodes of the tree, wkIs the weight value of the leaf node, λ is the regularization coefficient of the leaf node weight;
the final objective function can be derived and simplified by the following second order taylor expansion from equation (4):
Figure FDA0002642816570000025
Figure FDA0002642816570000026
Figure FDA0002642816570000027
wherein, giAnd hiFirst and second derivatives of the loss function, respectively;
and continuously training the XGboost algorithm to produce a new decision tree model, accumulating the prediction results of the residual error, and finally integrating all generated decision tree models.
4. The XGboost algorithm-based lithium battery residual life prediction method as claimed in claim 1, wherein the step 1) specifically comprises: the selected characteristic data is processed, and the magnitude order and dimension of the selected characteristic data are greatly different from the magnitude order and dimension of the residual capacity of the lithium battery, so that the extracted characteristic data needs to be standardized, the model is convenient to process and operate, and a specific formula is as follows:
Figure FDA0002642816570000028
wherein x' is the normalized data, x is the raw data that is not normalized, μ is the mean of the data, and σ is the standard deviation of the data;
establishing a lithium battery RUL prediction model by using a decision tree CART-based XGboost algorithm and selected feature data, and evaluating a prediction result of the model by using a mean square error RMSE;
Figure FDA0002642816570000031
wherein the content of the first and second substances,representing true data, yiThe output of the model is represented, and k represents the number of samples contained in the test set.
CN202010845212.5A 2020-08-20 2020-08-20 XGboost algorithm-based lithium battery residual life prediction method Pending CN111999649A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010845212.5A CN111999649A (en) 2020-08-20 2020-08-20 XGboost algorithm-based lithium battery residual life prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010845212.5A CN111999649A (en) 2020-08-20 2020-08-20 XGboost algorithm-based lithium battery residual life prediction method

Publications (1)

Publication Number Publication Date
CN111999649A true CN111999649A (en) 2020-11-27

Family

ID=73473680

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010845212.5A Pending CN111999649A (en) 2020-08-20 2020-08-20 XGboost algorithm-based lithium battery residual life prediction method

Country Status (1)

Country Link
CN (1) CN111999649A (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112990284A (en) * 2021-03-04 2021-06-18 安徽大学 Individual trip behavior prediction method, system and terminal based on XGboost algorithm
CN113204857A (en) * 2021-03-15 2021-08-03 北京锐达芯集成电路设计有限责任公司 Method for predicting residual life of electronic device based on extreme gradient lifting tree algorithm
CN113537770A (en) * 2021-07-15 2021-10-22 广西电网有限责任公司电力科学研究院 Decision tree configuration life prediction method and system based on cloud computing
CN113657646A (en) * 2021-07-15 2021-11-16 广西电网有限责任公司电力科学研究院 Method and system for acquiring equipment life prediction in real time based on cloud computing
CN113779884A (en) * 2021-09-14 2021-12-10 慧镕电子系统工程股份有限公司 Detection method for service life of recovered chip
CN113848480A (en) * 2021-09-23 2021-12-28 广东恒翼能科技有限公司 Method for predicting total discharge capacity of lithium battery capacity grading process
CN114169603A (en) * 2021-12-04 2022-03-11 湖北第二师范学院 XGboost-based regional primary school entrance academic degree prediction method and system
CN114167284A (en) * 2021-11-02 2022-03-11 江苏博强新能源科技股份有限公司 Lithium battery RUL prediction method and device based on BMS big data and ensemble learning
CN117572250A (en) * 2024-01-17 2024-02-20 山东工商学院 Method for estimating SOH of battery based on multi-feature fusion and XGBoost

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
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
CN108199795A (en) * 2017-12-29 2018-06-22 北京百分点信息科技有限公司 The monitoring method and device of a kind of equipment state
CN108896914A (en) * 2018-05-11 2018-11-27 浙江工业大学 A kind of modeling of gradient boosted tree and prediction technique of lithium battery health status
US20190113577A1 (en) * 2017-10-17 2019-04-18 The Board Of Trustees Of The Leland Stanford Junior University Data-driven Model for Lithium-ion Battery Capacity Fade and Lifetime Prediction
CN110058178A (en) * 2019-05-14 2019-07-26 郑州大学 A kind of lithium battery method for detecting health status and system
CN110757510A (en) * 2019-10-31 2020-02-07 广东工业大学 Method and system for predicting remaining life of robot
CN111044928A (en) * 2019-12-31 2020-04-21 福州大学 Lithium battery health state estimation method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
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
US20190113577A1 (en) * 2017-10-17 2019-04-18 The Board Of Trustees Of The Leland Stanford Junior University Data-driven Model for Lithium-ion Battery Capacity Fade and Lifetime Prediction
CN108199795A (en) * 2017-12-29 2018-06-22 北京百分点信息科技有限公司 The monitoring method and device of a kind of equipment state
CN108896914A (en) * 2018-05-11 2018-11-27 浙江工业大学 A kind of modeling of gradient boosted tree and prediction technique of lithium battery health status
CN110058178A (en) * 2019-05-14 2019-07-26 郑州大学 A kind of lithium battery method for detecting health status and system
CN110757510A (en) * 2019-10-31 2020-02-07 广东工业大学 Method and system for predicting remaining life of robot
CN111044928A (en) * 2019-12-31 2020-04-21 福州大学 Lithium battery health state estimation method

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112990284A (en) * 2021-03-04 2021-06-18 安徽大学 Individual trip behavior prediction method, system and terminal based on XGboost algorithm
CN112990284B (en) * 2021-03-04 2022-11-22 安徽大学 Individual trip behavior prediction method, system and terminal based on XGboost algorithm
CN113204857A (en) * 2021-03-15 2021-08-03 北京锐达芯集成电路设计有限责任公司 Method for predicting residual life of electronic device based on extreme gradient lifting tree algorithm
CN113537770A (en) * 2021-07-15 2021-10-22 广西电网有限责任公司电力科学研究院 Decision tree configuration life prediction method and system based on cloud computing
CN113657646A (en) * 2021-07-15 2021-11-16 广西电网有限责任公司电力科学研究院 Method and system for acquiring equipment life prediction in real time based on cloud computing
CN113779884A (en) * 2021-09-14 2021-12-10 慧镕电子系统工程股份有限公司 Detection method for service life of recovered chip
CN113848480A (en) * 2021-09-23 2021-12-28 广东恒翼能科技有限公司 Method for predicting total discharge capacity of lithium battery capacity grading process
CN113848480B (en) * 2021-09-23 2022-08-23 广东恒翼能科技有限公司 Method for predicting total discharge capacity of lithium battery capacity grading process
CN114167284A (en) * 2021-11-02 2022-03-11 江苏博强新能源科技股份有限公司 Lithium battery RUL prediction method and device based on BMS big data and ensemble learning
CN114167284B (en) * 2021-11-02 2023-12-22 江苏博强新能源科技股份有限公司 Lithium battery RUL prediction method and equipment based on BMS big data and integrated learning
CN114169603A (en) * 2021-12-04 2022-03-11 湖北第二师范学院 XGboost-based regional primary school entrance academic degree prediction method and system
CN117572250A (en) * 2024-01-17 2024-02-20 山东工商学院 Method for estimating SOH of battery based on multi-feature fusion and XGBoost

Similar Documents

Publication Publication Date Title
CN111999649A (en) XGboost algorithm-based lithium battery residual life prediction method
CN108896914B (en) Gradient lifting tree modeling and prediction method for health condition of lithium battery
CN110082682B (en) Lithium battery state of charge estimation method
Mamo et al. Long short-term memory with attention mechanism for state of charge estimation of lithium-ion batteries
CN113917334B (en) Battery health state estimation method based on evolution LSTM self-encoder
CN110544177A (en) Load identification method based on power fingerprint and computer readable storage medium
CN115469227B (en) Aggregation variation self-encoder and dynamic regular lithium battery abnormality detection method
Yuan et al. Evoq: Mixed precision quantization of dnns via sensitivity guided evolutionary search
CN111999650A (en) Lithium battery residual life prediction method based on support vector regression algorithm
CN116609668B (en) Lithium ion battery health state and residual life prediction method
CN114545274A (en) Lithium battery residual life prediction method
CN114726045B (en) Lithium battery SOH estimation method based on IPEA-LSTM model
CN115015760B (en) Lithium battery health state assessment method based on neural network and migration integrated learning
CN113484774B (en) Lithium battery pack capacity estimation method based on temperature calibration and neural network
CN115730635A (en) Electric vehicle load prediction method
CN113687242A (en) Lithium ion battery SOH estimation method for optimizing and improving GRU neural network based on GA algorithm
CN115994441A (en) Big data cloud platform online battery life prediction method based on mechanism information
CN115586444A (en) Lithium battery residual life prediction method based on VMD and BP neural network
CN116106761A (en) Real-time lithium ion battery electric quantity estimation method based on typical correlation analysis
CN114280490B (en) Lithium ion battery state of charge estimation method and system
CN112327169B (en) Lithium battery residual life prediction method
CN114460481A (en) Energy storage battery thermal runaway early warning method based on Bi-LSTM and attention mechanism
CN113657030A (en) Method for predicting remaining service life of lithium battery based on Gaussian process regression
CN116736133A (en) Early prediction method for capacity degradation track of lithium ion battery in full life cycle
CN117289167A (en) Battery remaining life prediction method, device and medium based on multiple neural network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20201127

RJ01 Rejection of invention patent application after publication