Research On Joint Online Estimation Method For Model Parameters And State Of Charge Of Lithium-ion Batteries

Due to the increasingly serious energy and environmental problems in China,electric vehicles have been promoted and widely concerned with their advantages of low emissions,low pollution and low noise.The research on key technologies of electric vehicles such as battery management system(BMS)and its core algorithms has become a hot research topic.The online accurate estimation of state of charge(SOC)is the key issue of BMS research.This project relies on the National Natural Science Foundation of China project "Lithium-ion power battery state and parameter adaptive joint estimation theory research" and focuses on the joint estimation method for model parameters and state of charge based on dual extended Kalman filter(DEKF)and its development and implementation in BMS.The full text is summarized as follows:Firstly,five equivalent circuit models of the lithium-ion battery were introduced.By comparison,the Thevenin equivalent circuit model with simple structure and easy parameter identification was selected as the battery model for research.The parameters of the battery were obtained by offline parameter identification experiment.The variation characteristics of each parameter with SOC,temperature and rate were analyzed.The discrete state space equation of the lithium-ion battery was derived.Secondly,in order to solve the problem that the SOC estimation accuracy is reduced due to the variation of model parameters in complex environment,this paper proposed the parameter online identification method based on forgetting factor recursive least squares(FFRLS)and the parameter online identification method based on extended Kalman filter(EKF).The dynamic condition tests of wide temperature range and multi-current condition were carried out.By comparison,the parameter online identification method based on EKF had higher precision.Then,the parameter online identification and SOC estimation were combined to construct the joint estimation method for model parameters and state of charge based on DEKF.Real-time online estimation of model parameters and SOC was achieved.The simulation model was constructed in Matlab/Simulink.Using the dynamic condition test data,starting from three error sources,the robustness of the DEKF method in complex operating environment was discussed comprehensively and systematically.It was concluded that the DEKF algorithm has strong robustness and high SOC estimation accuracy,the absolute value of the SOC estimation error is within 3%,and the convergence time is within 100s.Finally,the development and implementation of DEKF algorithm in BMS was carried out.The experimental platform was built to carry out bench experiments.The applicability and reliability of the DEKF algorithm in the real system were verified by simulating various complex conditions.It was concluded that the DEKF algorithm in BMS shows good SOC estimation performance under the complex operating environment of wide temperature range and multi-current condition.The absolute value of the SOC estimation error is always within 3%,and the convergence time is within 100s.This paper first transplanted the DEKF algorithm into the actual BMS,and made a positive and beneficial exploration of the engineering application of the algorithm.