Research on State of Charge Estimation of Lithium-Ion Battery Based on Improved Particle Swarm Optimization and Unscented Kalman Filter

Fast and accurate estimation of battery state of charge (SOC) is a key technology for battery maintenance and research on subsequent state of health and remaining useful life. Mathematical model of lithium-ion battery is established, considering its nonlinear system. The model parameters are identified online using the recursive least squares method with a forgetting factor. Relationship curves between open circuit voltage and SOC of battery in both charging and discharging states are established based on constant current charging and discharging condition of battery maintenance. To achieve more accurate SOC estimate, a SOC estimation method based on improved particle swarm optimization and unscented Kalman filter (IPSO-UKF) is proposed. Under constant current charge and discharge as well as hybrid pulse power characteristic (HPPC) tests, IPSO is employed to optimize the noise covariance matrix thereby enhancing the accuracy of SOC estimation. Experimental results validate the effectiveness of this method, demonstrating that the SOC estimation accuracy of the IPSOUKF surpasses that of UKF under constant current charge and discharge and HPPC tests. Furthermore, the SOC estimation convergence speed of IPSO-UKF is faster than that of PSOUKF.