A real time state of charge estimation using Harris Hawks optimization‐based filtering approach for electric vehicle power batteries

Summary The State of Charge (SOC) estimation of the battery is used to display the operating conditions of the battery to make charging/discharging decisions to reduce range anxiety. For this purpose, precision in SOC estimation is always essential. The Extended Kalman Filter (EKF) algorithm is an effective method for accurate SOC estimation of nonlinear batteries. However, the performance of the filter strongly depends on process noise covariance matrix (Q) and measurement noise covariance matrix (R) values. The improper value of these matrices reduces the convergence rate and increases the estimation error. In this paper, the impact of these matrices on SOC estimation is analyzed by the trial & error method, then the Harris Hawks Optimization (HHO) algorithm based EKF is proposed. As a first step, HHO is used to obtain the optimal values for covariance matrices. In the second step, optimized matrix values are injected into the EKF, which confirms the filter accuracy in the SOC estimation. To justify the necessity of an optimized EKF, a model‐based SOC estimation algorithm is executed for a Li‐ion battery discharging scenario. The effectiveness of the HHO based EKF is compared with the traditional particle swarm optimization‐based EKF and the Adaptive Extended Kalman filter (AEKF) for SOC estimation. The simulation results show the proposed HHO based EKF algorithm gives an accurate SOC estimation of the battery even when the initial SOC is incorrect. The numerical results show that the performance indices such as maximum error, Mean Absolute Error, Mean Square Error, and Root Mean Square Error of both SOC and voltage obtained by the proposed HHO‐EKF are low compared with AEKF. The robustness of the proposed algorithm is also verified with the static load conditions and the dynamic profile of an electric vehicle. In this paper, the impact of Noise Covariance matrices (Q and R) on SOC estimation is analyzed, then the Harris Hawks Optimization (HHO) algorithm based EKF is proposed. As a first step, HHO is employed to obtain the optimal values of noise covariance matrices. In the second step, optimized matrix values are injected into the EKF, which confirms the filter accuracy in the SOC estimation.