An adaptive fractional-order extended Kalman filtering approach for estimating state of charge of lithium-ion batteries

This paper proposes an adaptive fractional-order extended Kalman filter (AFEKF) approach for estimating state of charge (SOC) of a lithium-ion battery. Firstly, the fractional-order model (FOM) with a constant phase element module is established to describe the fractional-order characteristics inside a lithium-ion battery. Then, the augmented state equation including SOC is built by using the augmented vector approach. To avoid the calculation of the coefficients of the measurement equation, a linear adaptive integer-order Kalman filter is adopted. The AFEKF approach with the Sage-Husa filter is proposed to update the unknown parameters and unknown noises on the basis of augmented state equations. Finally, the comparison experiment between AFEKF approach and adaptive integer-order extended Kalman filter (AIEKF) approach is designed. Besides, the applicability of the AFEKF approach under different working conditions is also tested. The experimental tests indicate that the SOC estimation accuracy of the AFEKF approach is higher than that of the AIEKF approach, and the AFEKF approach can also be applicable in complex environments.