Fractional modeling and parameter identification of lithium-ion battery

Abstract

To simulate and control the lithium-ion battery system more effectively, it is necessary to establish a specific physical model of lithium-ion battery. The partnership for a new generation of vehicle (PNGV) model is a kind of equivalent circuit models which has low-complexity. Firstly, this paper introduces the PNGV model, and then derives the fractional PNGV model improved by fractional-order impedance elements. Furthermore, a random mutation ant colony optimization (RMACO) adapted to the fractional parameter identification is proposed, which uses the collected voltage and current data to perform parameter identification of the fractional PNGV model. Finally, the proposed algorithm is compared with the particle swarm optimization (PSO) algorithm, the absolute error and the average relative error of the RMACO are all less than the PSO. The results show that the RMACO has better parameter estimation effectiveness.