Enhancing the estimation accuracy in low state-of-charge area: A novel onboard battery model through surface state of charge determination

In order to predict the battery remaining discharge energy in electric vehicles, an accurate onboard battery model is needed for the terminal voltage and state of charge (SOC) estimation in the whole SOC range. However, the commonly-used equivalent circuit model (ECM) provides limited accuracy in low-SOC area, which hinders the full use of battery remaining energy. To improve the low-SOC-area performance, this paper presents an extended equivalent circuit model (EECM) based on single-particle electrochemical model. In EECM, the solid-phase diffusion process is represented by the SOC difference within the electrode particle, and the terminal voltage is determined by the surface SOC (SOCsurf) representing the lithium concentration at the particle surface. Based on a large-format lithium-ion battery, the voltage estimation performance of ECM and EECM is compared in the entire SOC range (0-100%) under different load profiles, and the genetic algorithm is implemented in model parameterization. Results imply that the EECM could reduce the voltage error by more than 50% in low-SOC area. The SOC estimation accuracy is then discussed employing the extended Kalman filter, and the EECM also exhibits significant advantage. As a result, the EECM is very potential for real-time applications to enhance the voltage and SOC estimation precision especially for low-SOC cases. (C) 2014 Elsevier B.V. All rights reserved.