State of Charge Estimation in Lithium‐Ion Batteries Using Adaptive Fractional‐Order Cubature Particle Filter

ABSTRACT In this study, an adaptive fractional‐order cubature particle filter (AFCPF) method is introduced for the estimation of state of charge (SOC) in lithium‐ion batteries (LIBs), utilizing a fractional‐order model. Initially, a fractional‐order equivalent circuit model incorporating a constant phase element module is formulated based on the Caputo definition, providing a precise depiction of the dynamic behavior of LIBs. Subsequently, the coefficients are adaptively estimated using the linear Kalman filter. Then, an augmented state equation incorporating the SOC is formulated using the augmented vector approach. To mitigate the effect of noise disturbance, adaptive estimations of the noise covariance matrices are implemented. Comparative experiments are carried out with the AFCPF method and the adaptive fractional‐order cubature Kalman filter (AFCKF). Furthermore, experiments under varying operational conditions are performed to validate the versatility of the AFCPF method. The experimental results indicate that the proposed AFCPF algorithm exhibits a notable superiority in SOC estimation accuracy at 25°C with 0.8 C. Compared to the AFCKF algorithm, the mean absolute error (MAE) and root mean square error (RMSE) are reduced by 13.81% and 9.8%, respectively. Furthermore, the AFCPF algorithm maintains its applicability for diverse environmental conditions.