Improved particle swarm optimization-adaptive dual extended Kalman filtering for accurate battery state of charge and state of energy joint estimation with efficient core factor feedback correction

With the rapid development of electric vehicles, the accuracy requirement for lithium-ion battery state feedback is increasing. However, traditional algorithms cannot achieve the desired accuracy. For this purpose, this article focuses on the ternary lithium-ion battery as the research object to achieve high-precision feedback. The results show that the proposed second-order resistor capacitance-partnership for a new generation vehicle (RC-PNGV) equivalent circuit model for ternary lithium-ion batteries and gradually decaying memory recursive least squares method improve the online parameter identification accuracy of battery equivalent models, successfully reducing the overall precision error of State of charge (SOC) estimation from 7.99 % to only 0.35 %. The improved particle swarm optimization algorithm-adaptive dual extended Kalman filter method effectively improves the accuracy and stability of joint estimation of SOC and SOE for ternary lithium-ion batteries. The error in joint estimation is reduced from 3.16 % to 0.89 %, demonstrating that the improved algorithm has high precision, adaptability, and correction capability. This study uses the improved particle swarm optimization - adaptive dual extended Kalman filter algorithm to research the joint estimation of SOC and SOE for lithium-ion batteries, aiming to provide efficient state feedback for batteries to ensure their operational efficiency and safety.