FRA and EKF Based State of Charge Estimation of Zinc-nickel Single Flow Batteries

The reliable state of charge (SOC) estimation is indispensable for flow batteries to maintain the safe and reliable operation. The widely adopted Extended Kalman filter (EKF) algorithm is a model-based method, however, the uncertainties in battery model will cause large errors in SOC estimation. An accurate battery model is the essence to capture the behaviors of batteries. In this paper, a novel framework for the SOC estimation of Zinc-nickel flow batteries is proposed based on the fast recursive algorithm (FRA) and extended Kalman filter (EKF). The FRA is firstly used to determine the model structure and identify the model parameters. Due to merits of FRA, a compact and accurate model of flow battery is built. Then, the SOC is estimated using the EKF based on the proposed linear-in-the-parameter model. Experimental studies and resultant simulations manifest the modelling accuracy of the proposed SOC estimation framework.