A Novel ADEKF Method for State-of-Charge Estimation of Liion Batteries

The capacity and state of charge (SOC) of a battery are essential for the battery management system (BMS). To perform high-accuracy online estimation of a battery throughout its life cycle, a joint estimator of SOC and capacity with low computational complexity based on an adaptive extended Kalman filter with a subregional decay element (ADEKF) is proposed in this paper. First, the parameters of the first-order circuit model are identified using the varied forgetting factor with recursive least squares (VFFRLS), and the simplified reduced-order equation of state for the battery is derived. Subsequently, the innovative ADEKF algorithm is proposed to adapt the simplified reduced-order state equation to achieve stable convergence of the SOC and capacity. Finally, the accuracy and adaptability of joint estimation with different initial states are verified using federal city driving program tests at 0 degrees C, 25 degrees C, and 45 degrees C for the optimal battery operating charge interval. In addition, compared with the other joint estimation methods, the results show that the proposed algorithm has much better performance in terms of joint estimation accuracy and convergence speed.