Research on battery SOC estimation method by combining optimization algorithm and multi-model Kalman filtering

Zhiming Chen, Changqi Zhu, Lei Liu

Abstract: With the rapid growth of electric vehicles and energy storage systems, accurate state of charge (SOC) estimation has become a critical component of battery management systems (BMS), essential for preventing overcharging and over-discharging, enhancing operational safety, and extending battery life. This paper proposes a novel SOC estimation method based on an enhanced self-correcting (ESC) model incorporating a second-order RC circuit, enabling a more accurate simulation of battery response time and dynamic behavior. To improve model reliability, a genetic algorithm-particle swarm optimization (GA-PSO) approach is employed for parameter identification. Additionally, a multi-model adaptive extended Kalman filter (AEKF) algorithm is introduced to achieve precise SOC estimation. MATLAB simulations using constant current discharge and automotive driving cycle data demonstrate that the proposed method outperforms traditional AEKF algorithms, with faster convergence and higher estimation accuracy, particularly in scenarios with varying initial estimation accuracies. The results highlight the potential of this approach to significantly enhance SOC estimation in BMS, contributing to safer operation and prolonged battery life in electric vehicles and energy storage systems.

Submission Number: 171