State of charge estimation of vehicle lithium-ion battery based on unscented Kalman filter

Accurate estimation of battery state of charge (SOC) is an important safety guarantee during battery charging and discharging. Aiming at the estimation of SOC of lithium-ion battery, the hybrid pulse power characteristic experiment (HPPC) and dynamic stress experiment (DST) are designed. Firstly, A straightforward Thevenin model was selected as an equivalent circuit model(ECM), and a genetic algorithm(GA) was utilised to ascertain the parameters of this model. Secondly, estimation of battery SOC is done using Extended Kalman Filter (EKF) and Unsigned Kalman Filter (UKF). Finally, the effectiveness of EKF and UKF algorithms is verified in dynamic stress test at varying temperatures. The results demonstrate that UKF algorithm exhibits higher accuracy than EKF algorithm, and its estimation error can be maintained at a level of 1.25%.