Application-specific parameterization of reduced order equivalent circuit battery models for improved accuracy at dynamic load

Abstract

Reduced order equivalent circuit battery models are widely used for modeling batteries as a part of complex system or as a basis for battery monitoring algorithms. These models are often parameterized in frequency domain using impedance spectroscopy or in time domain by applying current profile and measuring respective voltage response. This paper shows how the frequency characteristic of a typical battery load in a given application can be considered during parameterization to improve the accuracy of the model when it is used in this application with dynamic load. The method uses impedance-based parameterization technique by adopting additional weighting coefficients to the complex nonlinear least squares (CNLS) fitting procedure without introducing any restrictions on its applicability. As an example, modeling of lithium-ion battery in electric vehicles is considered. The proposed techniques reduce the inaccuracy of modeled dynamic battery voltage response by 40% in the considered example.