Battery multi-time scale fractional-order modeling method for state of charge estimation adaptive to full parameters updating

Abstract

The fractional-order theory has been successfully applied to battery modeling and state of charge (SOC) estimation thanks to the rapid development of smart energy storage and electric vehicles. The fractional-order model (FOM) has high nonlinearity, which makes it difficult to identify the parameters of the FOM, especially the online identification of the order. Aiming at the problem of parameter identification and SOC estimation of the FOM of battery, a multi-time scale fractional-order modeling method is proposed in this paper. Then, a multi-time scale parameter identification strategy based on feature separation is proposed, and two sub-filters are used to complete the online identification of parameters. Finally, a fractional-order multi-innovation unscented Kalman filtering (FO-MI-UKF) algorithm is proposed for SOC estimation to utilize the value of historical information better. Under dynamic stress test (DST) and Beijing bus dynamic stress test conditions (BBDST), compared with the single-time scale parameter identification algorithm and the single-innovation SOC estimation algorithm, the root mean square error of the estimation results is reduced by 13.3 % and 8.7 %, respectively. The experimental results verify the effectiveness of the modeling method and provide a new idea for fractional-order modeling.