CN103487759A - Hybrid electric vehicle battery SOC prediction method - Google Patents
Hybrid electric vehicle battery SOC prediction method Download PDFInfo
- Publication number
- CN103487759A CN103487759A CN201310428151.2A CN201310428151A CN103487759A CN 103487759 A CN103487759 A CN 103487759A CN 201310428151 A CN201310428151 A CN 201310428151A CN 103487759 A CN103487759 A CN 103487759A
- Authority
- CN
- China
- Prior art keywords
- battery
- prediction
- battery soc
- model
- electric vehicle
- 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
Links
Images
Landscapes
- Secondary Cells (AREA)
Abstract
The invention discloses a hybrid electric vehicle battery SOC prediction method. Accurate SOC value prediction is realized through external characteristic parameters of a battery. The hybrid electric vehicle battery SOC prediction method comprises the steps that (1) an input variable and an output variable needed by battery SOC prediction modeling are determined according to the external characteristic parameters of the battery; (2) hybrid electric vehicle simulation software is used for conducting development, modeling and simulation of a complete hybrid electric vehicle system, and input sample data of battery SOC prediction and output sample data of battery SOC prediction are collected; (3) a battery SOC prediction model is designed by means of the least squares support vector machine method under the bayesian evidence framework; (4) the established model is used for conducting prediction and error analysis. According to the hybrid electric vehicle battery SOC prediction method, the blindness of artificial parameter selection is avoided and the prediction effect of the model is enhanced. The designed prediction model has higher precision, an SOC value can be predicted accurately in real time especially under the conditions that battery energy feedback is considered and working conditions change continuously, the practicability is strong, and the effectiveness is high.
Description
Technical field
The present invention relates to a kind of power battery for hybrid electric vehicle SOC Forecasting Methodology that is applied to, adopt least square method supporting vector machine under Bayesian evidence framework.
Background technology
Battery is one of energy of hybrid vehicle, for guaranteeing that battery performance is good, extending battery, and the fuel economy that improves hybrid vehicle, must carry out rationally effectively management and control battery, but prerequisite be must be accurately and obtain reliably the existing capacity parameter of battery.SOC cannot directly be measured it as the bulk properties of battery, can only be by some external characteristic parameter predictions of directly measuring such as voltage, electric current, temperature are obtained.
Battery SOC Forecasting Methodology commonly used can be divided into following six classes: (1) is based on empirical equation and mathematical model, or the method for estimation of equivalent electrical circuit.The parameter of these mathematical models mainly obtains by the constant current charge-discharge characteristic, and this steady-state model can not reflect the dynamic perfromance of battery entirely truely.(2) Forecasting Methodology based on Ah counting.The Ah counting method is simple in structure, easy to operate, but has the defect that precision is not high in application.(3) Forecasting Methodology based on open-circuit voltage.The open-circuit voltage method is one of at present the most frequently used SOC Forecasting Methodology, by the stable current capacity of open-circuit voltage direct representation accumulator, simple to operate, but when measuring open-circuit voltage, need to consider galvanochemistry and the thermodynamic equilibrium of battery, simultaneously the stable long time that needs of open-circuit voltage.(4) Forecasting Methodology based on the accumulator internal resistance characteristic.The internal resistance method is that alternating current is injected into to accumulator, then the relation by internal resistance and capacity judges the current capacity of accumulator, during prediction SOC extreme value, precision is higher, but internal resistance is subject to battery temp, time of repose and discharges and recharges the impact of the factors such as original state, unstable with the relation of SOC, and accumulator internal resistance measuring instrument price is high, volume is large.(5) Forecasting Methodology based on the Kalman filter recursive algorithm.Kalman filtering method is regarded accumulator as dynamic system, and SOC is as a quantity of state of internal system, and the method need to be selected the descriptive equation of dynamic system, and recursive process also relates to complicated matrix inversion operation.Simultaneously, Kalman filter is as recursive algorithm, very responsive to the selection of initial value, and the initial value of mistake causes the continuous deterioration of estimating, it is given that initial SOC0 can be used open-circuit voltage to carry out, but for other initial values of recursion not preferably mode determine.(6) Forecasting Methodology based on neural network.Utilize non-linear mapping capability that neural network is stronger to realize the prediction of accumulator SOC.The method has been avoided the dependence of classic method to model and parameter, does not need impressed current and signal to process, and has improved system robustness and antijamming capability.But also there is at present over-fitting in neural network, easily is absorbed in local extremum, structural design depends on the defects such as experience.
Summary of the invention
Shortcoming for current SOC Forecasting Methodology, least square method supporting vector machine (least square support vector machines, LS-SVM) method provides effective solution, this mainly relies on two characteristics of LS-SVM: (1) empirical risk minimization, structural parameters are automatically determined according to sample data, are not had the over-fitting phenomenon in training process; (2) it is converted into the solve linear equations problem by the problem concerning study of standard SVM, has accelerated to solve speed, has overcome the defect of neural network.For improving matching and the precision of prediction of forecast model, further adopt the Bayesian evidence framework optimized algorithm to carry out parameter optimization to the LS-SVM model.
The technical solution adopted for the present invention to solve the technical problems is:
(1) according to the external characteristic parameter of battery as voltage, electric current, temperature and internal resistance etc., determine required input, the output variable of battery SOC prediction modeling;
(2) ADVISOR of hybrid vehicle simulation software is 2002 times, and hybrid electric vehicle complete vehicle system is developed, and modeling and simulation, gather input, output sample data that battery SOC is predicted;
(3) after input, output sample data are carried out to normalized, set up the input and output sample set for training and testing LS-SVM model;
(4), about the initial parameter value of LS-SVM forecast model, be also
with
initial value, the simulated annealing that we select the LS-SVM tool box to carry is here determined;
(5) after initial value is determined, utilize training sample set training LS-SVM, set up the SOC forecast model;
(6) adopt the Bayesian evidence framework optimized algorithm to model parameter
with
carry out optimizing;
(7) with required
with
again train LS-SVM, return to (5) step repeatedly, until select the optimum prediction model;
(8) with the model established, predicted and error analysis.
The invention has the beneficial effects as follows: the design forecast model has higher precision, especially in the situation that consider that battery energy regenerative and operating mode constantly change, still can dope real-time and accurately SOC value of battery, and practical, validity is high.
The accompanying drawing explanation
Below in conjunction with drawings and Examples, the present invention is further described.
Fig. 1 is the hybrid electric vehicle complete vehicle system development diagram;
Fig. 2 is vehicle speed data sample collection figure in the present invention;
Fig. 3 is the voltage data sample collection figure of battery in the present invention;
Fig. 4 is the current data sample collection figure of battery in the present invention;
Fig. 5 is the temperature data sample collection figure of battery in the present invention;
Fig. 6 is battery SOC data sample collection figure in the present invention;
Fig. 7 is the hybrid power automobile battery SOC prediction principle figure that adopts least square method supporting vector machine under Bayesian evidence framework.
Embodiment
Below in conjunction with accompanying drawing, the present invention is described in further detail.
Step of the present invention is as follows:
(1) according to the external characteristic parameter of battery as voltage, electric current, temperature and internal resistance etc., determine required input, the output variable of battery SOC prediction modeling;
(2) ADVISOR of hybrid vehicle simulation software is 2002 times, and hybrid electric vehicle complete vehicle system is developed, and modeling and simulation, gather input, output sample data that battery SOC is predicted;
(3) after input, output sample data are carried out to normalized, set up the input and output sample set for training and testing LS-SVM model;
(4), about the initial parameter value of LS-SVM forecast model, be also
with
initial value, the simulated annealing that we select the LS-SVM tool box to carry is here determined;
(5) after initial value is determined, utilize training sample set training LS-SVM, set up the SOC forecast model;
(6) adopt the Bayesian evidence framework optimized algorithm to model parameter
with
carry out optimizing;
(7) with required
with
again train LS-SVM, return to (5) step repeatedly, until select the optimum prediction model;
(8) with the model established, predicted and error analysis.
Fig. 1 is the hybrid electric vehicle complete vehicle system development diagram, and figure comprises engine, wheel, fuel tank, battery etc. parts, and we mainly develop battery here, selects rational performance parameter.
Fig. 2 to Fig. 6 is under Gary dynamic driving operating mode, the hybrid vehicle speed of a motor vehicle, cell voltage, electric current, temperature and corresponding SOC value, the input and output sample data of battery is proposed under MATLAB 7 environment, as training set and the test set of LS-SVM under Bayesian evidence framework by programming.
Fig. 7 is for adopting the hybrid power automobile battery SOC prediction principle figure of least square method supporting vector machine under Bayesian evidence framework, gather the input of the external characteristic parameter (voltage, electric current, temperature and SOC value) of battery as LS-SVM, adopt the Bayesian evidence framework optimized algorithm to optimize the parameter of LS-SVM,
with
, it should be noted that here
with
initial value have simulated annealing to try to achieve.In order fully to verify the validity of forecast model, Operation mode cycle is carried out to twice, the sample value obtained to be arranged, the odd term data that circulation is for the first time carried out in sample are used for training, and the even item data that circulation is carried out in sample are for the second time tested.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention.All any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (4)
1. a hybrid power automobile battery SOC Forecasting Methodology, it is characterized in that: by the external characteristic parameter of battery, realize SOC value Accurate Prediction, the external characteristic parameter of described battery comprises voltage, electric current, temperature and the internal resistance of battery, and the method specifically comprises the steps:
(1) determine required input, the output variable of battery SOC prediction modeling according to the external characteristic parameter of battery;
(2) utilize the hybrid vehicle simulation software, to hybrid electric vehicle complete vehicle system develop, modeling and simulation, gather input, the output sample data of battery SOC prediction;
(3) adopt the forecast model of least square method supporting vector machine method design battery SOC under Bayesian evidence framework;
(4) with the model established, predicted and error analysis.
2. a kind of hybrid power automobile battery SOC Forecasting Methodology according to claim 1, it is characterized in that: utilize the ADVISOR of hybrid vehicle simulation software 2002, hybrid electric vehicle complete vehicle system is developed to modeling and simulation, thereby the sample data of acquisition battery.
3. a kind of hybrid power automobile battery SOC Forecasting Methodology according to claim 1 and 2, it is characterized in that: the detailed process of described step (3) is:
(1) after input, output sample data are carried out to normalized, set up the input and output sample set for training and testing least square method supporting vector machine model;
(2), about the initial parameter value of LS-SVM forecast model, be also
with
initial value, select the simulated annealing that the LS-SVM tool box carries to determine;
(3) after initial value is determined, utilize training sample set training LS-SVM, set up the SOC forecast model;
(4) adopt the Bayesian evidence framework optimized algorithm to model parameter
with
carry out optimizing;
4. a kind of hybrid power automobile battery SOC Forecasting Methodology according to claim 1 and 2, it is characterized in that: Operation mode cycle is carried out twice, the sample value obtained is arranged, the odd term data that circulation is for the first time carried out in sample are used for training, the even item data that circulation is carried out in sample are for the second time tested, fully to verify the validity of forecast model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310428151.2A CN103487759A (en) | 2013-09-18 | 2013-09-18 | Hybrid electric vehicle battery SOC prediction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310428151.2A CN103487759A (en) | 2013-09-18 | 2013-09-18 | Hybrid electric vehicle battery SOC prediction method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103487759A true CN103487759A (en) | 2014-01-01 |
Family
ID=49828132
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310428151.2A Pending CN103487759A (en) | 2013-09-18 | 2013-09-18 | Hybrid electric vehicle battery SOC prediction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103487759A (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103983920A (en) * | 2014-05-26 | 2014-08-13 | 北京理工大学 | Method for establishing model of power battery of electric vehicle |
CN104156791A (en) * | 2014-08-29 | 2014-11-19 | 哈尔滨工业大学 | Lithium ion battery residual life predicting method based on LS-SVM probability ensemble learning |
CN105116343A (en) * | 2015-08-24 | 2015-12-02 | 桂林电子科技大学 | LS-SVM power cell SOC estimation method and system |
CN105678077A (en) * | 2016-01-07 | 2016-06-15 | 北京北交新能科技有限公司 | Online prediction method of power performance of lithium ion battery for hybrid power vehicle |
CN106021923A (en) * | 2016-05-19 | 2016-10-12 | 江苏理工学院 | Method and system for predicting state of charge of power battery of pure electric vehicle |
CN107290679A (en) * | 2017-07-03 | 2017-10-24 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric automobile |
CN108139446A (en) * | 2015-10-15 | 2018-06-08 | 江森自控科技公司 | For predicting the battery test system of cell testing results |
CN108828448A (en) * | 2018-06-08 | 2018-11-16 | 江苏大学 | Battery charge state estimation on line method based on charging voltage curve fusion Kalman filtering |
CN111157897A (en) * | 2019-12-31 | 2020-05-15 | 国网北京市电力公司 | Method and device for evaluating power battery, storage medium and processor |
CN113406500A (en) * | 2021-06-29 | 2021-09-17 | 同济大学 | Method for estimating residual electric quantity of power lithium battery |
CN113657459A (en) * | 2021-07-28 | 2021-11-16 | 合肥国轩高科动力能源有限公司 | Battery SOC prediction method and medium based on combination of principal component and support vector machine |
CN116819346A (en) * | 2023-08-29 | 2023-09-29 | 深圳凌奈智控有限公司 | Battery SOC estimation method, device, equipment and storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060212277A1 (en) * | 2005-03-16 | 2006-09-21 | Terry Hansen | Apparatus and method for estimating battery state of charge |
CN102680907A (en) * | 2012-05-31 | 2012-09-19 | 浙江大学 | Battery charging stress optical coefficient (SOC) detection method in photovoltaic system |
-
2013
- 2013-09-18 CN CN201310428151.2A patent/CN103487759A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060212277A1 (en) * | 2005-03-16 | 2006-09-21 | Terry Hansen | Apparatus and method for estimating battery state of charge |
CN102680907A (en) * | 2012-05-31 | 2012-09-19 | 浙江大学 | Battery charging stress optical coefficient (SOC) detection method in photovoltaic system |
Non-Patent Citations (5)
Title |
---|
BHASKAR SAHA等: "Prognostics Methods for Battery Health Monitoring Using a Bayesian Framework", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》 * |
刘勇智等: "最小二乘支持向量机在航空蓄电池剩余容量预测中的应用", 《蓄电池》 * |
阎威武等: "一种贝叶斯证据框架下支持向量机建模方法的研究", 《控制与决策》 * |
陈健美等: "基于LS-SVM的混合动力镍氢电池组SOC预测", 《中南大学学报(自然科学版)》 * |
鲍慧等: "新型电动汽车动力电池的荷电状态预测研究", 《绵阳师范学院学报》 * |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103983920A (en) * | 2014-05-26 | 2014-08-13 | 北京理工大学 | Method for establishing model of power battery of electric vehicle |
CN104156791A (en) * | 2014-08-29 | 2014-11-19 | 哈尔滨工业大学 | Lithium ion battery residual life predicting method based on LS-SVM probability ensemble learning |
CN105116343A (en) * | 2015-08-24 | 2015-12-02 | 桂林电子科技大学 | LS-SVM power cell SOC estimation method and system |
CN105116343B (en) * | 2015-08-24 | 2017-10-27 | 桂林电子科技大学 | The electrokinetic cell state of charge method of estimation and system of least square method supporting vector machine |
CN108139446A (en) * | 2015-10-15 | 2018-06-08 | 江森自控科技公司 | For predicting the battery test system of cell testing results |
CN105678077A (en) * | 2016-01-07 | 2016-06-15 | 北京北交新能科技有限公司 | Online prediction method of power performance of lithium ion battery for hybrid power vehicle |
CN106021923A (en) * | 2016-05-19 | 2016-10-12 | 江苏理工学院 | Method and system for predicting state of charge of power battery of pure electric vehicle |
CN107290679B (en) * | 2017-07-03 | 2019-11-12 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric car |
CN107290679A (en) * | 2017-07-03 | 2017-10-24 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric automobile |
CN108828448A (en) * | 2018-06-08 | 2018-11-16 | 江苏大学 | Battery charge state estimation on line method based on charging voltage curve fusion Kalman filtering |
CN108828448B (en) * | 2018-06-08 | 2020-08-28 | 江苏大学 | Battery state of charge on-line estimation method based on charging voltage curve fusion Kalman filtering |
CN111157897A (en) * | 2019-12-31 | 2020-05-15 | 国网北京市电力公司 | Method and device for evaluating power battery, storage medium and processor |
CN111157897B (en) * | 2019-12-31 | 2022-05-10 | 国网北京市电力公司 | Method and device for evaluating power battery, storage medium and processor |
CN113406500A (en) * | 2021-06-29 | 2021-09-17 | 同济大学 | Method for estimating residual electric quantity of power lithium battery |
CN113657459A (en) * | 2021-07-28 | 2021-11-16 | 合肥国轩高科动力能源有限公司 | Battery SOC prediction method and medium based on combination of principal component and support vector machine |
CN116819346A (en) * | 2023-08-29 | 2023-09-29 | 深圳凌奈智控有限公司 | Battery SOC estimation method, device, equipment and storage medium |
CN116819346B (en) * | 2023-08-29 | 2023-11-07 | 深圳凌奈智控有限公司 | Battery SOC estimation method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103487759A (en) | Hybrid electric vehicle battery SOC prediction method | |
Yang et al. | A fractional-order model-based battery external short circuit fault diagnosis approach for all-climate electric vehicles application | |
Mesbahi et al. | Dynamical modeling of Li-ion batteries for electric vehicle applications based on hybrid Particle Swarm–Nelder–Mead (PSO–NM) optimization algorithm | |
Guo et al. | A state-of-health estimation method of lithium-ion batteries based on multi-feature extracted from constant current charging curve | |
Dong et al. | A method for state of energy estimation of lithium-ion batteries based on neural network model | |
Castano et al. | Dynamical modeling procedure of a Li-ion battery pack suitable for real-time applications | |
Rezvanizaniani et al. | Review and recent advances in battery health monitoring and prognostics technologies for electric vehicle (EV) safety and mobility | |
Sun et al. | Model-based dynamic multi-parameter method for peak power estimation of lithium–ion batteries | |
JP5683175B2 (en) | An improved method for estimating the unmeasurable properties of electrochemical systems | |
CN102680795B (en) | Real-time on-line estimation method for internal resistance of secondary battery | |
JP5875037B2 (en) | Battery state prediction system, method and program | |
CN103632018B (en) | A kind of fuel cell modelling method based on Simscape platform | |
Li et al. | On-line estimation method of lithium-ion battery health status based on PSO-SVM | |
CN105807226A (en) | Lithium ion battery SOC prediction method and device based on equivalent circuit model | |
JP6305988B2 (en) | Device and method for determining energy status based on data derived from processing methods | |
CN104820189A (en) | Systems and methods for battery state estimation | |
CN105334462A (en) | Online estimation method for battery capacity loss | |
Xiong et al. | Research on an online identification algorithm for a thevenin battery model by an experimental approach | |
CN110568360B (en) | Lithium battery aging diagnosis method based on fuzzy logic algorithm | |
KR101996943B1 (en) | Apparatus and method for estimating of battery state-of-charge | |
Li et al. | Discharge capacity estimation for Li-ion batteries based on particle filter under multi-operating conditions | |
Zhang et al. | A multi-fault diagnosis method for lithium-ion battery pack using curvilinear Manhattan distance evaluation and voltage difference analysis | |
JP2011122917A (en) | Device for evaluating battery characteristics | |
CN103558555B (en) | A kind of realistic model for evaluating Vehicular dynamic battery performance parameter | |
Li et al. | An improved battery on-line parameter identification and state-of-charge determining method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140101 |