CN106372272B - Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis - Google Patents

Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis Download PDF

Info

Publication number
CN106372272B
CN106372272B CN201610620760.1A CN201610620760A CN106372272B CN 106372272 B CN106372272 B CN 106372272B CN 201610620760 A CN201610620760 A CN 201610620760A CN 106372272 B CN106372272 B CN 106372272B
Authority
CN
China
Prior art keywords
degradation
capacity
prediction
data
generalized
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
Application number
CN201610620760.1A
Other languages
Chinese (zh)
Other versions
CN106372272A (en
Inventor
刘红梅
李连峰
吕琛
马剑
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.)
Beijing University of Aeronautics and Astronautics
Original Assignee
Beijing University of Aeronautics and Astronautics
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 Beijing University of Aeronautics and Astronautics filed Critical Beijing University of Aeronautics and Astronautics
Priority to CN201610620760.1A priority Critical patent/CN106372272B/en
Publication of CN106372272A publication Critical patent/CN106372272A/en
Application granted granted Critical
Publication of CN106372272B publication Critical patent/CN106372272B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/20Design optimisation, verification or simulation
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation, e.g. linear programming, "travelling salesman problem" or "cutting stock problem"

Abstract

The invention provides a lithium battery capacity and service life prediction method based on a generalized degradation model and multi-scale analysis, which comprises the following steps: firstly, performing phase space reconstruction on historical capacity degradation data, and fitting a generalized degradation model by using the reconstructed data to obtain model parameters and fitting a root mean square error; secondly, carrying out forward one-step estimation on future capacity in a plurality of scales, and obtaining a forward one-step prediction result through weighted integration; then, adding the prediction result into a known sequence, and carrying out forward iterative prediction; and finally, judging whether the multi-step prediction result reaches a failure threshold value or not, and calculating the residual service life. The generalized degradation model provided by the invention can fit a complex combined degradation rule, and has the advantages of strong model universality, wide application range and small fitting error; the multi-scale prediction is carried out on the basis of phase space reconstruction, and then the prediction result is obtained through weighting integration, so that the prediction result is not easily interfered by noise and has high accuracy.

Description

Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis
Technical Field
The invention relates to the technical field of lithium battery health management, in particular to a lithium battery capacity and remaining service life prediction method based on a generalized degradation model and multi-scale analysis.
Background
The lithium ion battery is a new generation of green high-energy rechargeable battery, has the outstanding advantages of high voltage, large energy density, good cycle performance, small self-discharge, no memory effect and the like, can keep excellent temperature characteristic, good leakage-proof performance, stable working voltage and longer working life even after long-time discharge, and is widely applied to aerospace, ships, automobiles and consumer electronics products.
During long-term use of lithium ions, the performance of the lithium ions is gradually degraded and finally completely fails. The performance degradation is caused by the gradual deterioration of physical and chemical structural properties of the positive and negative active materials, the bonding strength of the coating, the quality of the diaphragm and the like in the cyclic charge and discharge process. An unexpected end of battery life can cause the whole functionality of the system to be invalid, a typical example is that a mars ring detector emitted by the U.S. aviation and space administration sends out an error instruction in the operation process to command the solar panel to act towards the sun, the instruction is executed without considering the battery performance state, so that the battery is overdischarged, the temperature rise is too high, the recharging capability is lost, and the whole system loses the power supply and loses the connection. The battery pack of the electric automobile often generates overcharge and overdischarge phenomena due to overlong charging and discharging time in the working process, thereby not only influencing the service performance of the battery and shortening the service life of the battery, but also reducing the driving range of the electric automobile and lowering the cost performance of the whole automobile; breakthrough of battery in service life, tolerance and driver series technology will eventually lead to development of lithium ion batteries with low cost, durability and good tolerance. However, the performance of the system of the lithium ion battery decreases with the lapse of time, aging, environmental factors, and operational conditions. Therefore, it is very important to manage the health of the battery system. The health state of the battery is monitored in real time, and the residual service life is predicted, so that on one hand, the safe and reliable operation of the system can be ensured, the serious consequences caused by the failure of the battery are reduced, and on the other hand, decision support can be provided for the maintenance and repair of the system.
Lithium ion battery health management technology has attracted research interest of numerous scholars in recent years. Hundreds of papers on theory and applications in this field appear in academic journals, meeting paper collections and technical reports each year. Currently, methods for lithium battery health assessment and prediction mainly focus on three directions: a mechanism model based approach, a data driven approach and a hybrid approach. The main challenge of the mechanism model based approach is the lack of a theoretical circuit model unique to lithium batteries and the numerical estimation of circuit virtual components requires bulky and expensive equipment such as EIS equipment. The greatest advantage of the data-driven approach is that no expert knowledge about the battery chemistry and material properties is required, and the key to success is the relevance and availability of the extracted performance-degrading features. The invention provides a novel data-driven prediction method, which can capture performance degradation characteristics in historical capacity data in a self-adaptive manner and can be used for predicting future states without assuming a specific performance degradation model in advance; meanwhile, the predicted value is predicted from multiple angles by using the idea of ensemble learning, and then the result is obtained by weighting integration, so that the predicted robustness is good, and the accuracy is high.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the existing lithium battery prediction method based on the degradation evolution model needs to assume in advance that a battery to be predicted follows a specific degradation mode, one model can only be suitable for similar batteries with the same degradation rule, the universality of the model is poor, no dynamic adjustment capability exists in different degradation stages, and the accuracy of the prediction result is poor.
The technical scheme adopted by the invention for solving the technical problems is as follows: a lithium battery capacity and service life prediction method based on a generalized degradation model and multi-scale analysis comprises the following steps:
(1) performing phase space reconstruction based on historical capacity degradation data to obtain performance degradation data of the lithium battery in multiple scales;
(2) the generalized performance degradation model can adaptively extract performance degradation rules from historical data and can fit more complex combined degradation rules, and the specific model is as follows:
in the formula, tiIs time,. epsiloniThe model is Gaussian noise, a, b, c and d are model parameters, and a, b, c and d are more than 0, and the model parameters can be determined through test data;
(3) predicting the future capacity of the previous step in a plurality of analysis scales, and carrying out integrated weighting on the root-mean-square error value fitted based on the generalized degradation model to obtain a capacity value to be predicted;
(4) in multi-step prediction, the newly added data is used for dynamically updating the parameters of the degradation model so as to keep better dynamic prediction capability.
Compared with the prior art, the invention has the advantages that:
(1) the generalized degradation model provided by the invention can fit a complex combination degradation rule, and has the advantages of strong model universality, wide application range and small fitting error;
(2) the multi-scale prediction is carried out on the basis of phase space reconstruction, and then the prediction result is obtained through weighted integration, so that the prediction result is not easily interfered by noise and has high accuracy.
Drawings
FIG. 1 is a flow chart of lithium battery capacity and life prediction;
FIG. 2 is a linear degradation curve under different parameters;
FIG. 3 is a quadratic degradation curve under different degradation parameters;
FIG. 4 is a root function degradation curve under different parameters;
FIG. 5 is a generalized model degradation curve under different degradation parameters;
FIG. 6 is a schematic diagram of the reconstruction principle of raw data;
FIG. 7 is a diagram of a forward multi-step prediction principle;
FIG. 8 is a CS2_36 battery prediction, where FIG. 8(a) is 300 cycle positions, FIG. 8(b) is 350 cycle positions, and FIG. 8(c) is 400 cycle positions;
fig. 9 shows the CX2 — 38 battery prediction result, where fig. 9(a) shows 400 cycle positions, fig. 9(b) shows 550 cycle positions, and fig. 9(c) shows 600 cycle positions.
Detailed Description
The invention is further described with reference to the following figures and detailed description.
The invention relates to a lithium battery capacity and service life prediction method based on a generalized degradation model and multi-scale analysis, which specifically comprises the following steps:
1. generalized degradation model
The product or device gradually degrades in performance due to physical or chemical changes during long-term use, eventually leading to failure. The common degradation model mainly has three types, i.e., constant-speed degradation in which the degradation speed is substantially constant, degradation in which the degradation speed gradually increases, and degradation in which the degradation speed gradually decreases. The degradation speed refers to the performance degradation amount of a product or equipment in unit time, and reflects the rapid degree of the product tending to fail. The degradation speed of the product with the constant-speed degradation rule is the same in different time periods of the life cycle, and the degradation rule is expressed by a linear function. However, many mathematical models can be used for modeling non-uniform degradation (degradation speed is time-varying), and the invention selects a quadratic curve and a root function curve to respectively model the degradation of two different degradation laws. The selection has the advantages that different types of degradation rules can be comprehensively represented, meanwhile, the necessary model parameters are few, and the model parameters can be separated from other parts in the model, so that the optimization processing is facilitated.
Three different types of degradation laws are described in detail below:
(1) linear degradation model
The linear degradation function is shown in equation (1),
yi=a-b·tii(1)
in the formula, tiThe time or cycle number corresponding to the ith measurement; y isiPerformance indicators such as capacity, health index, etc.; a and b are model parameters, a is intercept, b is degradation speed, and a and b are more than 0; epsiloni~N(0,σ2) I is 1,2, n is gaussian noise, and n is the number of sampling points. FIG. 2 shows the degradation parameters under different conditionsLinear degradation curve.
According to the formula (1), for the performance degradation conforming to the linear degradation rule, the degradation model of the system can be obtained only by identifying two parameters of the system through degradation data, and the future state and the residual life can be predicted based on the degradation model. In any case, the law of the evolution of the degradation of the actual product is often not simply linear.
(2) Secondary degradation model
In the process of degradation, the degradation speed of a large number of products or equipment in practical use is increasing. Equation (2) can model this acceleration degradation,
yi=c-d·ti 2i(2)
wherein c and d are model parameters c, d is more than 0 and is determined by experimental data. Fig. 3 shows the degradation curves for different degradation parameters.
(3) Root function degradation model
Some products or devices, or some degradation stage of the products, may show a phenomenon of slow degradation, and this type of decay law is described by formula (3).
In the formula, e and f are model parameters e, f is more than 0 and is determined by experimental data. Fig. 4 shows the degradation curves for different degradation parameters.
(4) Generalized degradation model
The models show three typical performance decay rules of products or equipment, and a single model is only suitable for corresponding products with unchanged performance decay rules. However, the actual product usually runs under complex working conditions and environments, and it is not easy to know which degradation model to follow in different use stages, so a general degradation model is needed to automatically capture the degradation rule of the product, and the parameters of the model are determined by test data without defining the specific type of the model in advance. The performance decay rule of the actual product is a combination form of the three, as shown in formula (4).
Wherein a, b, c and d are model parameters, and a, b, c and d are more than 0 and can be determined by experimental data. The different combinations of the three parameters b, c, d determine the shape of the degradation curve. Fig. 5 shows the degradation curves for different degradation parameters.
As can be seen from fig. 5, the generalized degradation model can obtain different types of degradation rules by adjusting parameters of the degradation model, and can provide various types of evolution rules common in the product performance degradation process. According to the method, the generalized degradation model is used, and the performance degradation model parameters are identified according to the degradation data of the lithium battery in different degradation stages, so that the capacity is predicted.
2. Lithium battery capacity and service life prediction based on generalized degradation model and multi-scale analysis
The method comprises the steps of firstly, fitting a generalized degradation model through the historical capacity degradation data of the lithium battery, obtaining parameters of the model, and then inputting future time points into the model to obtain the future performance state of the system. In order to improve the robustness of prediction, the future state of the system is predicted at a plurality of time scales, and then weighting is performed based on the fitting residual error to obtain a final result, and the specific principle is described in detail as follows.
(1) One step forward capacity prediction
And recording the capacity data sequence of the lithium battery which is P cycles ahead by taking the time of the predicted point as a reference as { x }1,x2,...,xPAnd forward one-step prediction is carried out to estimate the capacity of the lithium battery in P +1 cyclesTo predict the capacity at time P +1 from multiple scales, the original data is reconstructed in phase space as follows, the principle of which is shown in FIG. 6.
Where Q is the embedding dimension and D is the highest analysis scale.
For convenience of representation, the reconstructed data is recorded as:
use ofRespectively carrying out least square fitting analysis to obtain generalized degradation modelsEstimation of respective parameters ofThe Root Mean Square Error (RMSE) of the fit is:
the capacity of a P +1 cycle lithium battery is estimated as:
the root mean square error of the generalized regression model fitted on different analysis scales is recorded as { rmse1,rmse2,...,rmseDThe prediction result of the fitting model on the P +1 circulation capacity isFirst, a weighted weight is generated based on the fitted root mean square error sequence:
the final estimate of the battery capacity for the P +1 cycle is then:
the method applies the idea of ensemble learning during prediction, carries out multiple times of analysis for obtaining a final result, and finally integrates the analysis result to obtain the final result. The algorithm has stronger robustness, stronger anti-noise interference capability and higher prediction precision.
(2) Forward multi-step capacity prediction and life prediction
In order to provide decision support for product maintenance and component replacement, in practice, multi-step prediction or residual life prediction is often required to be performed on the future capacity of the lithium battery, so that maintenance activities are arranged at proper time, the use value of the product is fully exerted on the premise of ensuring the safe operation of equipment, and the life cycle cost is reduced.
Using a prediction window sliding mode to perform multi-step prediction on the future capacity of the lithium battery, as shown in fig. 7, during the first-step prediction, using known data to respectively obtain parameters of each-order model through multi-scale regression analysis, and then extending a point forward to obtain a prediction result of the previous step; when the second step of prediction is carried out, the prediction result of the first step is used as the input of a training model, and the prediction result of the second step is obtained through the model; by analogy, the result of the prediction n steps ahead can be obtained.
By extending the sequence to a given failure threshold, the predicted number of steps required to reach the failure threshold is the current Remaining Useful Life (RUL) of the battery.
(3) Evaluation of prediction results
The accuracy of the predicted results is evaluated using Root Mean Square Error (RMSE), which can quantify the average level of difference between the predicted values and the actual values. The result of the capacity prediction of the previous n steps is recorded asThe value of the real capacity is { yP+1,yP+2,...,yP+nAnd RMSE calculation formula of the predicted result is as follows:
3. test verification
The invention uses the lithium ion battery cycle life data provided by the CalCE center of the university of Maryland to carry out the verification of the algorithm. The Life test was for two different types of LiCoO2The batteries (CS2 and CX2) were subjected to a cyclic charge-discharge test, and voltage, current, temperature, capacity, impedance, and the like during charge and discharge were measured and stored. Table 1 gives specification information for two types of batteries.
TABLE 1CS2 and CX2 Battery Specifications
The test cell is subjected to a life cycle test under a standard constant current-constant voltage profile: firstly, charging a battery in a constant current state, keeping the voltage constant after the voltage reaches a preset value (charging cut-off voltage) until the charging current drops to a set value (20mA), and finishing the charging process; the battery is discharged under a constant current state, and when the discharge voltage drops to a preset value (discharge termination voltage), the discharge process is finished. In the experiment, the charge cut-off voltage and the discharge end voltage were set to 4.2V and 2.7V, respectively, and constant current discharge was performed at a 1C rate, that is, the discharge current was constant at 1A.
The discharge capacity is selected as a health index of the lithium battery, and the failure threshold value is set to be 80% of the rated capacity, namely when the capacity of the test battery is degraded to be 0.8 times of the initial capacity, the battery is considered to be failed. The test results of batteries CS2_36 and CX2_38 are shown in table 2.
TABLE 2 results of the CS2_36 and CX2_38 tests on batteries
The prediction method provided by the invention is used for predicting the capacity and the residual service life of the CS2_36 battery and the CX2_38 battery at different cycle positions at the future moment, the fitting data length Q is set to be 200, the analysis scale D is set to be 2, and the prediction results are respectively shown in FIGS. 8 and 9.
The accuracy of the predictions was quantified using Root Mean Square Error (RMSE), which is shown in table 3 for the two-cell predictions.
TABLE 3 root mean square error of predicted results
As can be seen from fig. 8, fig. 9 and table 3, in the early stage of prediction, the known data is less, the battery performance degradation rule is not sufficiently represented, and thus the accuracy of the capacity and the remaining life prediction is low. With the increase of known data, the prediction result is more and more accurate; near the failure moment, the difference between the predicted capacity and the actual capacity is small, and meanwhile, the prediction result of the residual service life is accurate. Test analysis results show that the method can effectively and accurately predict the future capacity and the residual service life of the lithium battery, and the accuracy of the prediction result is high.

Claims (1)

1. A lithium battery capacity and service life prediction method based on a generalized degradation model and multi-scale analysis is characterized by comprising the following steps: the method comprises the following steps:
(1) performing phase space reconstruction based on historical capacity degradation data to obtain performance degradation data of the lithium battery in multiple scales;
(2) the generalized performance degradation model can adaptively extract performance degradation rules from historical data and can fit more complex combined degradation rules, and the specific model is as follows:
in the formula, tiIs time,. epsiloniThe model is Gaussian noise, a, b, c and d are model parameters, and a, b, c and d are more than 0, and the model parameters can be determined through test data;
(3) predicting the future capacity of the previous step in a plurality of analysis scales, and carrying out integrated weighting on the root-mean-square error value fitted based on the generalized degradation model to obtain a capacity value to be predicted; wherein the content of the first and second substances,
the method comprises the steps of firstly, fitting a generalized degradation model through the historical capacity degradation data of the lithium battery to obtain parameters of the model, then inputting future time points into the model to obtain the future performance state of the system, predicting the future state of the system at a plurality of time scales in order to improve the predicted robustness, and then weighting based on fitting residual errors to obtain a final result, wherein the method specifically comprises the following steps:
(1) one step forward capacity prediction
And recording the capacity data sequence of the lithium battery which is P cycles ahead by taking the time of the predicted point as a reference as { x }1,x2,...,xPAnd forward one-step prediction is carried out to estimate the capacity of the lithium battery in P +1 cyclesIn order to predict the capacity at time P +1 from multiple scales, the original data is reconstructed in phase space,
wherein Q is the embedding dimension and D is the highest analysis scale;
for convenience of representation, the reconstructed data is recorded as:
use ofRespectively carrying out least square fitting analysis to obtain generalized degradation modelsAll the ingredients of (A)Estimation of numbersThe Root Mean Square Error (RMSE) of the fit is:
the capacity of a P +1 cycle lithium battery is estimated as:
the root mean square error of the generalized regression model fitted on different analysis scales is recorded as { rmse1,rmse2,...,rmseDThe prediction result of the fitting model on the P +1 circulation capacity isFirst, a weighted weight is generated based on the fitted root mean square error sequence:
the final estimate of the battery capacity for the P +1 cycle is then:
(4) in multi-step prediction, the newly added data is used for dynamically updating the parameters of the degradation model so as to keep better dynamic prediction capability.
CN201610620760.1A 2016-08-01 2016-08-01 Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis Active CN106372272B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610620760.1A CN106372272B (en) 2016-08-01 2016-08-01 Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610620760.1A CN106372272B (en) 2016-08-01 2016-08-01 Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis

Publications (2)

Publication Number Publication Date
CN106372272A CN106372272A (en) 2017-02-01
CN106372272B true CN106372272B (en) 2020-02-07

Family

ID=57878658

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610620760.1A Active CN106372272B (en) 2016-08-01 2016-08-01 Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis

Country Status (1)

Country Link
CN (1) CN106372272B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107292025B (en) * 2017-06-21 2019-08-06 北京航空航天大学 The sealing life prediction technique of soft bag lithium ionic cell
CN107273632B (en) * 2017-06-28 2019-09-27 北京航空航天大学 The non-accumulated impact pectus part lifetime of system prediction technique of one kind and device

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101620045A (en) * 2009-07-31 2010-01-06 北京航空航天大学 Method for evaluating reliability of stepping stress quickened degradation experiment based on time sequence
CN102279928A (en) * 2011-07-20 2011-12-14 北京航空航天大学 Product performance degradation interval prediction method based on support vector machine and fuzzy information granulation
CN102789545A (en) * 2012-07-12 2012-11-21 哈尔滨工业大学 Method for predicating remaining life of turbine engine based on degradation model matching
CN103439666A (en) * 2013-09-02 2013-12-11 北京航空航天大学 Geometric method for evaluating capacity recession of lithium ion battery
CN103678938A (en) * 2013-12-27 2014-03-26 北京航空航天大学 Degradation model consistency testing method catering to space shapes and error ranges
CN103778295A (en) * 2014-01-26 2014-05-07 南京航空航天大学 Method for evaluating operating reliability of multi-model integrated aero-engine under multiple failure modes
CN105117550A (en) * 2015-08-26 2015-12-02 电子科技大学 Product multidimensional correlation-oriented degradation failure modeling method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8467984B2 (en) * 2009-09-30 2013-06-18 Battelle Energy Alliance, Llc Systems, methods and computer readable media for estimating capacity loss in rechargeable electrochemical cells

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101620045A (en) * 2009-07-31 2010-01-06 北京航空航天大学 Method for evaluating reliability of stepping stress quickened degradation experiment based on time sequence
CN102279928A (en) * 2011-07-20 2011-12-14 北京航空航天大学 Product performance degradation interval prediction method based on support vector machine and fuzzy information granulation
CN102789545A (en) * 2012-07-12 2012-11-21 哈尔滨工业大学 Method for predicating remaining life of turbine engine based on degradation model matching
CN103439666A (en) * 2013-09-02 2013-12-11 北京航空航天大学 Geometric method for evaluating capacity recession of lithium ion battery
CN103678938A (en) * 2013-12-27 2014-03-26 北京航空航天大学 Degradation model consistency testing method catering to space shapes and error ranges
CN103778295A (en) * 2014-01-26 2014-05-07 南京航空航天大学 Method for evaluating operating reliability of multi-model integrated aero-engine under multiple failure modes
CN105117550A (en) * 2015-08-26 2015-12-02 电子科技大学 Product multidimensional correlation-oriented degradation failure modeling method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
"A Multi-Factor Battery Cycle Life Prediction Methodology for Optimal Battery Manageement";Valentin Muenzel et al;《Proceedings of the 2015 ACM Sixth International Conference on Future Energy systems》;20150717;第57-64页 *
"Prognostics and health monitoring for lithium-ion battery";Qiang Miao et al.;《Proceedings of 2011 IEEE International Conference on Intelligence and Security Informatics》;20110815;第242-246页 *
"时变应力下蓄电池循环寿命预测方法研究";刘天宇 等;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20160315;第8-18页 *
"锰酸锂动力电池寿命测试方法";时玮 等;《汽车工程》;20150115;第37卷(第1期);第67-71页 *

Also Published As

Publication number Publication date
CN106372272A (en) 2017-02-01

Similar Documents

Publication Publication Date Title
Pan et al. State of charge estimation of lithium-ion batteries using a grey extended Kalman filter and a novel open-circuit voltage model
Yang et al. A neural network based state-of-health estimation of lithium-ion battery in electric vehicles
CN109031153B (en) Online health state estimation method for lithium ion battery
CN107861075B (en) Method for determining SOP of power battery
CN107797067B (en) A kind of lithium ion battery service life migration prediction technique based on deep learning
Chen et al. State of health estimation for lithium-ion batteries based on fusion of autoregressive moving average model and elman neural network
CN108919129A (en) When a kind of under variable working condition power battery life-span prediction method
CN107015157A (en) The lithium battery cycles left life-span online fast test method of fragment is risen based on constant current equipressure
CN108872866B (en) Dynamic evaluation and long-acting prediction fusion method for charge state of lithium ion battery
Kim et al. Real-time state of charge and electrical impedance estimation for lithium-ion batteries based on a hybrid battery model
CN106372272B (en) Lithium battery capacity and service life prediction method based on generalized degradation model and multi-scale analysis
CN110398697B (en) Lithium ion health state estimation method based on charging process
CN108896914B (en) Gradient lifting tree modeling and prediction method for health condition of lithium battery
CN104965179A (en) Lithium ion storage battery temperature combinational circuit model and parameter identification method thereof
CN110161425B (en) Method for predicting remaining service life based on lithium battery degradation stage division
Chen et al. Analysis and prediction of the discharge characteristics of the lithium–ion battery based on the Grey system theory
Jiang et al. An aging-aware soc estimation method for lithium-ion batteries using xgboost algorithm
De Sousa et al. Comparison of different approaches for lead acid battery state of health estimation based on artificial neural networks algorithms
Kim et al. A real-time condition monitoring for lithium-ion batteries using a low-price microcontroller
Chen et al. State of health estimation for lithium-ion battery based on long short term memory networks
Tong et al. Comprehensive battery equivalent circuit based model for battery management application
Zhang et al. A novel method for estimating state-of-charge in power batteries for electric vehicles
Wu et al. Health Prognosis with Optimized Feature Selection for Lithium-ion Battery in Electric Vehicle Applications
Vatani et al. Cycling lifetime prediction model for lithium-ion batteries based on artificial neural networks
Kim et al. A new hybrid filter-based online condition monitoring for lithium-ion batteries

Legal Events

Date Code Title Description
PB01 Publication
C06 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