A fractional-order model of lithium-ion batteries and multi-domain parameter identification method

Abstract

A multi-domain parameter identification method for a fractional-order model of lithium-ion batteries is presented. The fractional-order model is studied, and twenty-five identification parameters are determined. An intelligent optimization method named the genetic algorithm-particle swarm optimization algorithm is used to identify the parameters. Based on electrochemical impedance spectroscopy in the frequency domain and the terminal voltage of the dynamic stress test in the time domain, a multi-domain identification method is proposed. In synthetic experiment, the proposed genetic algorithm-particle swarm optimization algorithm has higher accuracy and faster convergence speed than the traditional optimization methods, and the proposed multi-domain identification method has more accurate parameter identification results than the frequency-domain identification and time-domain identification. In experiment on lithium-ion batteries, the model parameters are identified by electrochemical impedance spectroscopy and dynamic stress test data, and the parameter identification results are verified by verification test data. The results demonstrate that the genetic algorithm-particle swarm optimization algorithm and multi-domain identification method can be used as robust and reliable tools for parameter identification of lithium-ion batteries. A MATLAB application with the proposed method is also published on the community MATLAB website, providing researchers with a more convenient and effective tool.