The Research On The On–Line SoC Estimation Method For Power Battery

At present,due to the lack of fossil energy and environmental problems,electric vehicles are widely used in daily life,and the power supply of electric vehicles is also more and more concerned.In order to improve the reliability of the power system of electric vehicles and expand the use of electric vehicles,the real-time accurate estimation of the SoC(State of Charge)of the power battery has great significance.Because the lithium-ion battery is currently the most widely used and the most promising power battery,so the lithium-ion battery is used as the research object,and its structure and performance are introduced in this study.The method of SoC estimation and the estimation accuracy are analyzed.The main contents of this study can be divided into the following parts.Firstly,the internal structure and working principle of the lithium-ion battery are analyzed,and the polarization effect of the battery is explained and divided into three categories according to the causes of the polarization effect.The factors affecting the available capacity of the battery such as charge and discharge rate,self-discharge rate,ambient temperature and battery aging degree are carried out.Through the correlation test,the SoC-EMF(Electromotive Force)curve of the battery is obtained,and the function of the battery SoC and EMF is obtained according to the curve fitting.Secondly,the advantages and disadvantages of the four equivalent circuit models commonly used of the battery are analyzed.And the parameters of Rint model,Thevenin model and DP model are identified based on the test results of the HPPC(Hybrid Pulse Power Characteristic)test by the least squares method and the simulation circuit of the three models is built.The accuracy of the models is compared by the error between the measured open circuit voltage and the simulated open circuit voltage of the battery.And the simulation results are analyzed.The average error of the open circuit voltage simulated by the Rint model is larger than that of the other two models obviously.The error of the Thevenin model and the DP model are very close to each other,but considering the accuracy and calculation of the model,Thevenin model is finally used as the equivalent circuit model of battery SoC estimation.Thirdly,the EKF(Extended Kalman Filter)algorithm and the Fireworks algorithm are introduced,and the Fireworks algorithm is applied to the estimation of the battery SoC,and an improved EKF algorithm is proposed.After that,the experiments are designed to verify the performance of the improved EKF algorithm.The SoC estimation results of the EKF algorithm and the improved EKF algorithm are compared with the real SoC under the DST test operating mode.Then,a certain degree of Gaussian noise is added to the input current of the two algorithms to verify the anti-jamming capability of the algorithms.Finally,different values are used as the initial value of the battery SoC,but the subsequent estimation of the SoC is still completed to verify the robustness of the algorithms.The experimental results show that the two algorithms both have good robustness.Compared with the EKF algorithm,the improved EKF algorithm has higher accuracy of the SoC estimation,and the improved EKF algorithm still has a high accuracy under the case of noise interference.