A Study of Recursive Techniques for Robust Identification of Time-Varying Electrical Equivalent Circuit Models of Li-Ion Batteries

Abstract

This article presents results of a comparative study of recursive techniques for robust identification of time-varying electrical equivalent circuit models of li-ion batteries. Two such methods are studied, a direct continuous time system identification method and an indirect discrete time technique. The results of this study indicate that although both methods work equally well for identification of time-invariant circuit models from clean voltage-current data, the direct continuous time method outperforms indirect discrete time technique for identification of time-varying circuit models. Similar conclusions can also be drawn for identification of equivalent circuit models in the presence of noise and/or unmodeled dynamics.