State of Charge Estimation for Power Battery Base on Improved Particle Filter

Abstract

In this paper, an improved particle filter (Improved Particle Swarm Optimized Particle Filter, IPSO-PF) algorithm is proposed to estimate the state of charge (SOC) of lithium-ion batteries. It solves the problem of inaccurate posterior estimation due to particle degradation. The algorithm divides the particle population into three parts and designs different updating methods to realize self-variation and mutual learning of particles, which effectively promotes global development and avoids falling into local optimum. Firstly, a second-order RC equivalent circuit model is established. Secondly, the model parameters are identified by the particle swarm optimization algorithm. Finally, the proposed algorithm is verified under four different driving conditions. The results show that the root mean square error (RMSE) of the proposed algorithm is within 0.4% under different driving conditions, and the maximum error (ME) is less than 1%, showing good generalization. Compared with the EKF, PF, and PSO-PF algorithms, the IPSO-PF algorithm significantly improves the estimation accuracy of SOC, which verifies the superiority of the proposed algorithm.

Keywords

• lithium-ion battery
• state of charge
• particle filter
• particle swarm optimization