DOI:
10.1109/sepoc67005.2025.11297616
Authors:
David Grasev, Dalibor Biolek, Viera Biolková, Zdeněk Kolka
Abstract:
This paper presents a method for identifying the parameters of Li-ion battery impedance models using measured data. The primary objective is to enhance the likelihood of locating the global minimum of the objective function, rather than converging to a local one. The proposed approach integrates two optimization algorithms: Particle Swarm Optimization (PSO) and the Nelder-Mead (NM) simplex method. PSO performs a broad exploration of the parameter space to identify a promising region near a potential global minimum, while NM is employed for local refinement. The degree of interaction between the two algorithms is parametrically adjustable, allowing users to balance accuracy and computational efficiency. The methodology is demonstrated using impedance measurements of a 13,450 mAh Li-ion battery, obtained with a BioLogic SP300 potentiostat.