State-of-Charge And State-of-Health Estimation Of Lithium-ion Battery For Electric Vehicle Base On EKF

Under the circumstance of energy shortage and environment pollution,which threat people's normal life,the rise of electric vehicles alleviates the energy and environment problem to some extent.As the optiaml power source of electric vehicles,lithium-ion battery has become the research focus of universities and enterprises.So the state of charge(SOC)and the state of Health(SOH)have great research value.Accurate estimate of SOC and SOH can extend the battery's lifetime,reduce the cost of electric vehicles production and improve the security of electric vehicles.The main work of this article is done as follows:1.In order to establish the model of the battery,a series of charge and discharge experiments of Lithium-ion batteries have been done,the temperature,discharge rate and the cyclic life of the battery are considered,the weighting factors of the model is analyzed,and a second-order RC Thevenin equivalent circuit model is proposed.In order to estimate the parameters of the model accurately,firstly the cftool of MATLAB is used to match parameters of model,then the parameters of the model is estimated by recursive least square adding forgetting factor,we know that the latter method is more suitable for real car after comparison.2.Aiming at accuracy problem of the model,the simulation of the established battery model under pulse charge-discharge working condition and custom changing electricity working condition is conducted,the simulation results and the experiment results are compared,the proposed model has great dynamic and static characteristics and high model accuracy.3.Based on the verified model,the extended Kalman filter(EKF)is taken to estimate the SOC and SOH.Choose internal resistance as the SOH's definition by analyzing the characteristic quantity of SOC.The simulation results and experimental data in different situations are compared,the result shows that the proposed SOC estimation algorithm can correct the error introduced by noise.Because the initial value is difficult to get in real car,the initial value is verified by this algorithm,we can know that the algorithm can correct the initial value constantly and have better robustness for the initial error.The Dual-EKF is chosen to estimate SOC and internal resistance simultaneously.Check the change of SOH by changing internal resistance periodically,the pulse charge-discharge working condition and UDDS working condition and the NEDC working condition of our country are chosen,and the effectiveness of the algorithm is verified.