Parameter identification and SOC estimation for power battery based on multi-timescale double Kalman filter algorithm

Abstract

Accurate modelling and state of charge (SOC) estimation are of great significance to improve efficiency of power batteries and extend their life. In order to solve the issue of time-varying model parameters resulting in inaccurate SOC estimation, a combined online identification of model parameters and SOC estimation method for lithium-ion batteries based on extended Kalman filter (EKF) and unscented Kalman filter (UKF) with different timescales is proposed. A second-order RC circuit model is established and model parameters are identified online by UKF on a macroscopic timescale, and the battery SOC is estimated by EKF on a microscopic timescale. Compared with conventional SOC estimation methods in which model parameters are identified offline, the proposed method can obtain more accurate SOC estimation. The SOC mean absolute error (MAE) and root mean square error (RMSE) are both significantly reduced under the urban dynamometer driving schedule (UDDS) test. The SOC estimation results demonstrate the accuracy and robustness of the proposed method.