Lithium-Ion Battery Pack SOC Estimation using Optimized ECM Parameters and Kalman Filter

Abstract

This paper presents an extended Kalman filter (EKF) to estimate the state of charge (SOC) of series connected battery pack considering different practical aspects. Modeling is done to determine how capacity and resistance changes at the cell level affect battery pack performance. Experimental current and voltage of Li-ion cell along with the nonlinear least square method are used to obtain optimized model parameters, which can reduce the computation time as compared to identifying them in real time. The proposed EKF can reduce computation and complexity from a hardware deployment point of view by using an algorithm iteratively for all cells in the battery pack. The efficiency of proposed method is evaluated by simulating different real time scenarios in MATLAB. Impact of unequal charge distribution among different cells to decide battery pack SOC is analyzed. Performance of the proposed EKF for SOC estimation is found to be improved with reduced complexity and computations.