基于改进的EKF法估计电动汽车电池SOC

Estimation of SOC based on improved EKF method for electric vehicle batteries

基于传统卡尔曼滤波算法的电池组荷电状态(state of charge,SOC)估计方法适合于电流变化比较剧烈的电动汽车动力电池SOC估计,但由于电池模型以及系统噪声、量测噪声统计特性的不确定性,容易引起滤波发散。在研究与分析极化效应、库仑效率、内阻、温度、老化等对电池可用容量的影响实验的基础上,对扩展的卡尔曼滤波(expended kalman filter,EKF)算法进行改进。实验结果表明:改进后的EKF方法对随机的量测噪声具有较强的抑制能力,提高了估算精度,更适用于实际应用。

The state of charge（SOC） estimation of battery pack based on traditional Kalman filter method is suitable for estimating the SOC of electric vehicle batteries where the current fluctuates drastically. However, the uncertainty due to battery model and statistical information of the system and measurement noise will result in filtering divergence. Based on the analysis of factors affecting the SOC such as polarization effect, coulombic efficiency, internal resistance, temperature and ageing, the expended Kalman filter method was improved. Accordingly, the accuracy of the estimate system was improved. Matlab simulation and experiments were carried out. The comparison indicates that the improved EKF method performs well when disturbance happens.