Improved parameters identification and state of charge estimation for lithium-ion battery with real-time optimal forgetting factor

Abstract

Accurate estimation of state of charge (SOC) is extremely essential for energy management of electric vehicles, and precise identification of model parameters will directly affect the results of SOC estimation. However, the traditional recursive least squares (RLS) method cannot accurately track the changes of model parameters in actual complex conditions. To solve these problems, a new parameters identification method combining real-time variable forgetting factor recursive least squares (VFFRLS) and adaptive extended Kalman filter (AEKF) is proposed, and the unscented Kalman filter (UKF) method is used to calculate SOC in real time. Through comparative verification and analysis, this method owns good accuracy of model parameters identification and robustness in three commonly used equivalent circuit models. Finally, experiments under dynamic stress test (DST) cycles show that the root mean square error of terminal voltage and SOC are 0.19% and 0.07% respectively in dual polarization model with VFFRLS, which proves that the proposed method can significantly improve the estimation accuracy of SOC.