Exponential Moving Average Extended Kalman Filter for Robust Battery State-of-Charge Estimation

This paper proposes a exponential moving average extended Kalman filter (EMAF-EKF) for the purpose of battery state-of-charge (SOC) estimation. The proposed approach relies on a second-order RC circuit model of the Lithium-ion battery. In the proposed approach, first, exponential moving average filters are applied to the various input signals that are battery current, temperature and terminal voltage. Then, the filtered signals are feed to the conventional extended Kalman filter for the state-of-charge estimation. Unlike the convention extended Kalman filter, the proposed EMAF-EKF is very robust to measurement noises in the input signals. Moreover, compared to other noise robust extended Kalman filter, our solution is very simple and easy to implement. Experimental results with the Turnigy graphene 5000mAh65C li-ion battery dataset are provided to show the performance improvement in SOC estimation by the proposed approach over the conventional counterpart.