State-of-charge estimation of lithium-ion batteries based on fractional-order modeling and adaptive square-root cubature Kalman filter

This paper mainly studies the state of charge (SOC) estimation of lithium batteries based on a fractional-order adaptive square-root cubature Kalman filter (FO-ASRCKF). Firstly, a fractional-order model (FOM) of lithium battery is established by using fractional-order derivative theory. In order to meet the identification accuracy, an improved adaptive genetic algorithm is applied to the process of multi-parameter model identification. Then, the FO-ASRCKF algorithm based on FOM and adaptive rules is proposed, and a comparative experiment with Fractional-order adaptive iterative extended Kalman filter (FO-AIEKF) and Integer-order adaptive square-root cubature Kalman filter (IO-ASRCKF) is carried out. The experimental results show that the proposed FO-ASRCKF can work normally under various working conditions, and it has higher SOC estimation accuracy, with the mean absolute error (MAE) being less than 0.5%. Moreover, it can also overcome the divergence caused by noise and wrong initial values, indicating a better robustness.