State of Charge Estimation of Vanadium Redox Flow Battery Based on Online Equivalent Circuit Model

State of charge (SOC) is an important index to ensure the safe and stable operation of vanadium redox flow battery (VRB) and its accurate estimation can protect the battery from overcharge and over-discharge. Time-varying model parameters and nonlinear estimation algorithm are the main challenges for accurate estimation of SOC. This paper proposes the SOC estimation method for VRB based on online equivalent circuit model. Firstly, the recursive least squares with forgetting factor (FRLS) is used to online identify the model parameters. Then, based on this online model, the unscented Kalman filter (UKF) is employed to online estimate SOC. Finally, a hybrid pulse current experiment is conducted to verify the proposed the estimation method on a 5kW/3kWh VRB system in the laboratory. Validated with experimental data, the online identified model parameters and the estimated SOC present excellent accuracy, the root mean square error (RMSE) of the model and SOC is 0.0165V and 0.01 respectively. The proposed SOC estimation method is capable of estimating SOC accurately.