Robust Gain-scheduled Observer Design for State-of-charge Estimation of Rechargeable Battery with Hysteresis Characteristics

Abstract

In this paper, we consider a model-based state-of-charge estimation of a rechargeable battery with hysteresis characteristics. One approach to the estimation problem is to utilize a linear observer designed based on an approximately derived linear time-invariant model. However, due to the nonlinearity of the hysteresis characteristics, this cannot achieve high estimation performance. To improve the estimation performance, we focus on that the hysteresis is nonlinear only to a single variable and describe the battery system as a linear parameter-varying (LPV) model. Then, a robust gain-scheduled observer against the modeling errors is designed for the LPV system. The observer design is formulated as a convex optimization problem under linear matrix inequality (LMI) constraints. The effectiveness of the proposed method is illustrated through numerical simulations.