Improved Particle Swarm Optimization-Kalman Particle Filtering for High-Precision State-of-Energy Evaluation of Lithium-Ion Batteries Over Wide Temperature Ranges based on Fusion Dual Factor Parameter Identification

In recent years, lithium-ion batteries have become an important power source for new energy electric vehicles. State-of-energy (SOE) is an important parameter in the battery management system, which determines the current maximum possible range of electric vehicles. In this study, an improved particle swarm optimization-Kalman particle filtering for high-precision SOE evaluation method is proposed. First, the improved recursive least squares algorithm is designed by fusion of the forgetting factor and the memory length to improve the accuracy and robustness of online parameter identification. Then, aiming at the problems of particle degradation and diversity loss, the extended Kalman filtering and particle swarm optimization strategies are introduced, which effectively improve the accuracy and stability of the particle filter (PF). Finally, to further improve the filtering performance, this study proposes an adaptive mechanism of noise variance, which effectively enhances the convergence of PF. The experimental results under complex working conditions show that the mean absolute error of the parameter identification method proposed in this study is between 0.63% and 0.75%, and the root mean square error of the proposed SOE estimation method is between 0.46% and 0.65%, indicating that the method proposed in this study has high robustness and accuracy.