State of Charge (SOC) Estimation Based on Extended Exponential Weighted Moving Average H∞ Filtering

When the classical H∞ algorithm (HIF) is applied to estimate state of charge (SOC) a lithium battery, influence historical data usually ignored, resulting in an increase estimation error. In order improve accuracy SOC estimation, this paper proposes extended exponential weighted moving average (EE-HIF) view data. By designing Gaussian function, distribution at different times can effectively reduce error caused by inaccuracy battery model. addition, when system contains white noise and alternating current input, proposed method achieve faster convergence speed better robustness. Simulation results show advantages algorithm, as compared HIF filtering exponentially (EWMA).