On-line state of charge estimation of lithium-ion power battery pack using optimized unscented Kalman filtering

Abstract

Power battery pack is the energy source of pure electric vehicles, hybrid vehicles and battery energy storage system for micro-grid energy source. In the practical work, The main factor affecting the stable operation of power battery pack is the cells' inconsistency, Inconsistencies will make the whole battery pack service life reduced greatly; The main factor affecting the efficient operation of the whole battery pack is the accuracy of state of charge (SOC) estimation, The accuracy of SOC estimation affects the battery pack energy utilization directly. In this paper, in order to eliminate inconsistencies impact on cycle life of the battery pack, A novel equalization technology is proposed, In the equalization circuit, equalization sub-circuits can achieve DC-DC converter function, by controlling the sub-circuit we can achieve the energy transfer from single cell to part series battery pack or transfer from the part battery pack to the single cell. In order to improve the accuracy of the battery SOC estimation, Build the simulation system for charging and discharging the battery pack, adaptive unscented kalman filter(AUKF) algorithm is proposed to estimate the battery pack SOC, the simulation results show that the SOC estimation accuracy with the adaptive unscented Kalman filter algorithm is higher than the ordinary unscented Kalman filter(UKF) algorithm.