Study On The State Of Charge Estimation Of Power Lithium-ion Battery

Although the technology of pure electric vehicles has become more and more mature in recent years,there are still many urgent problems to be solved,such as the cruising ability of pure electric vehicles,the service life and energy utilization of battery,etc.The main method to solve these problems is to design an efficient and reasonable battery management system(BMS).Estimating the state of charge(SOC)of the battery is one of the most basic and important functions of the battery management system and the accuracy of the SOC estimation is related to total system performance.In this paper,with 2000 m Ah ternary power lithium-ion battery as the research object,the pros and cons of SOC estimated by Extended Kalman Filter(EKF)are analyzed.Then aiming at improving the SOC estimation accuracy,the SOC estimation is studied.The main contributions are as follows:(1)The working principle of lithium-ion battery is analyzed,and several factors that affect SOC estimation,such as temperature and charge-discharge rate,are analyzed through experiments.(2)Through the analysis and comparison of several typical equivalent circuit models,we determine the Thevenin equivalent circuit model as the battery model of the lithium battery.The functional relationship between SOC and open circuit voltage is confirmed though experiment,and the parameter identification of the battery equivalent circuit model is also studied.Two different parameter identification methods are used to identify battery model parameters,namely offline parameter identification and online parameter identification based on recursive least squares algorithm.Finally,the battery models established using the two different parameter identification methods are analyzed and compared through experiments.The test results show that the battery model established using online parameter identification based on the recursive least squares algorithm with forgetting factor has higher precision.(3)We analyize the principle of Kalman filtering algorithm,and summarize the advantages and disadvantages of EKF algorithm in the process of estimating SOC.Aiming at the shortcomings of the EKF algorithm in the process of estimating the SOC,that is reducing the estimation accuracy due to introducing the linearization error,the IEKF algorithm is proposed to reduce the influence of linearization errors in the EKF estimation process.Levenberg-Marquardt algorithm is used to correct the error covariance in the iterative process of IEKF algorithm to ensure that the estimation error of IEKF is uniformly reduced during the iteration.Finally,the SOC-based estimation of the LM-IEKF algorithm is verified via a simulation test,and the estimation results are compared with the EKF estimation ones.We find that the LM-IEKF algorithm has higher accuracy in estimating SOC.(4)Taking into account the accuracy of the battery model and the statistics of the process noise and measurement noise of the battery system all have a greater impact on the estimation SOC of EKF,in order to reduce the impact of above two aspects on SOC estimation so as to improve the SOC estimation accuracy,we propose to use much robust H_∞ filtering to estimate battery SOC.The performance of H_∞ filter for estimating the SOC of the lithium battery is analyzed and the estimation result is compared with the EKF result.The result show that the H_∞ filtering estimates the SOC of lithium battery with better accuracy and robustness than EKF.