Performance Analysis of State of Charge Estimation Techniques for Two-Wheeler Electric Vehicles

Abstract

In this article, a comparative study of the various state of charge (SOC) estimation methods, for example, coulomb count (CC), Kalman filter (KF), and extended Kalman filter (EKF), has been presented for a two-wheeler electric vehicle's Lithium-ion (Li-ion) battery. Accurate estimation of SOC is a critical requirement in EV for its range prediction, charging mode selection, remaining useful life estimation, etc. However, SOC prediction is a challenging task due to the nonlinearities of the battery. Therefore, an electrical equivalent circuit model of Li-ion battery is devised in this work, and the above-said methods, i.e., CC, KF, and EKF, are developed in MATLAB Simulink for battery's SOC estimation under different discharge test profiles. A 39 Ah capacity Li-ion battery with a nominal voltage rating of 72 V, suitable for two-wheeler EV application, is considered for study with all the estimation techniques. It has been found that the EKF method provides minimal error (< ±0.l %) in SOC estimation under all loading conditions since it can incorporate the nonlinearities of battery in its estimation technique.