High accuracy temperaure-dependent SOC estimation based on real-time parameter identification for rechargeable Li-Ion battery pack

The model-based adaptive control methods such as extended Kalman filter (EKF) are widely used as a typical state of charge (SOC) estimation method for a battery pack. The EKF estimates and corrects the SOC by properly adjusting the control gain, according to a state of the battery pack. However, the performance of the EKF cannot be guaranteed when a temperature of the battery pack is sharply changed. The accuracy of the equivalent circuit model is varied according to the temperature, therefore the performance of the EKF estimation is not consistent. In order to increase the performance of the SOC estimation at various temperatures when using the adaptive control method, techniques for designing accurate systems and the Kalman gain are required. Therefore, in this paper, an observer for accurate parameter identification of the battery pack is designed and the EKF is used to improve the SOC estimation performance at low temperatures.