A novel on-board state-of-charge estimation method for aged Li-ion batteries based on model adaptive extended Kalman filter

Abstract A battery management system needs to have an accurate inline estimation of SOC for each individual cell in the battery pack. This estimation process poses challenges after substantial battery aging. This article presents a novel method based on model adaptive extended Kalman filter (MAEKF) to estimate SOC for Li-ion batteries. Sensitivity analysis of the electrical model verifies that the accuracy of SOC estimated by EKF is sensitive to resistors used in the cell's electrical model. In order to get the best estimation, values of resistors are obtained in an optimization process in the MAEKF. This method uses the fact of two sudden changes in the cell's voltage derivative with respect to time while discharging current is constant. These two points are assumed as reference points in which their SOC can be determined from cell's chemistry. The optimization algorithm uses the derivative of the cell's measured terminal voltage to allocate SOC of 92% and 15% for two reference points in the Vcell equation and updates cell's electrical model. The algorithm's process is fast and computationally inexpensive, making on-board estimation practical. The obtained results demonstrate that by using this method the estimated SOC error for aged Li-ion cells does not exceed 4%.