A Hybrid Estimation Method for SOC of Lithium Batteries in Electrical Vehicles Considering Vehicle Operating Condition Recognition

摘要

Owing to its advantages of being closed-loop and online as well as small memory storage length, adaptive extended Kalman filter (AEKF) based state of charge (SOC) estimation method for lithium batteries in electric vehicles (EVs) has been widely used. However, EVs often have to operate under complex conditions, such as sudden acceleration and deceleration. Under the strong nonlinear input, the AEKF based method can not track the true SOC rapidly. In order to solve this problem, an estimator for EVs operating condition is used to identify the working conditions in this paper, including stable condition and non-stationary condition. AEKF based method is used to estimate SOC when EVs run under stable condition, and look-up table based method is employed to estimate SOC in case of the non-stationary condition. The experimental results show that the hybrid estimation method proposed in this paper has higher SOC estimation accuracy and better convergence rate.