Unknown Input Observer for Battery Open Circuit Voltage Estimation: an LMI Approach

Abstract

Open Circuit Voltage (OCV) is a vital component in Battery Management System which can be utilized to determine the battery condition namely state of charge (SOC) and state of health (SOH). Since OCV can not be measured when a battery is continuously connected to load, the only way to determine its value is through estimation. In this paper, the battery OCV is estimated using Unknown Input Observer (UIO). The parameters of the UIO are formulated into linear matrix inequality (LMI) to satisfy the stability of a chosen Lyapunov function. By solving the LMI, one possible solution for UIO parameters is obtained. To demontrate the effectiveness of the designed UIO in estimating OCV, simulations of a battery discharged using pulse load profile and varying load profile are performed. The results show that the OCV estimation faithfully tracks the actual OCV where the OCV estimation error tends to zero for a given initial OCV estimation error. Furthermore, the employment of UIO may decrease computational complexity since there is no need to include nonlinear SOC-OCV relationship in the OCV estimation.