State-of-Charge Estimation of Lithium-Ion Battery Using Multi-State Estimate Technic for Electric Vehicle Applications

Abstract

For reliable and safe operation of lithium-ion batteries in electric vehicles, the monitoring of the internal states of the batteries such as state-of-charge (SOC) is necessary. The purpose of this work is to present a novel SOC estimation algorithm. In this work, an equivalent circuit model (ECM) as well as the parameter identification method are studied. Then, the model structure of the battery in the state-space form is further investigated. Based on the model structure analysis, a novel SOC estimation algorithm is proposed using multi-state technic and Extend Kalman Filter (EKF). Some improvements are then introduced to improve the convergence and tracking performance of the algorithm in electric vehicle applications. The performances of the algorithm are validated through some experiments and simulations. Validation results show that the proposed SOC estimation algorithm can achieve an acceptable accuracy with the mean error being less than 2.72%.