基于ESP模型的锂电池参数辨识和SOC估计  

Parameters identification and soc estimation of lithium batteries based on ESP model

作　　者：吴波[1] 谢锋 卢佩航 徐劲力[1] WU Bo;XIE Feng;LU Pei-hang;XU Jin-li(School of Mechanical and Eletrical Engineering,Wuhan University of Technology,Wuhan Hubei 430070,China)

机构地区：[1]武汉理工大学机电工程学院,湖北武汉430070

出　　处：《电源技术》2020年第6期832-835,874,共5页Chinese Journal of Power Sources

摘　　要：精准的电池模型是电动汽车电池管理系统的关键,它为准确估计电池荷电状态(SOC)提供保证。基于电化学反应机理,建立扩展单粒子(ESP)模型。基于ESP模型参数较多的特点,利用改进遗传算法分不同SOC阶段进行多组参数辨识,通过充放电实验验证模型的准确性。基于ESP模型和卡尔曼滤波算法,引入强跟踪滤波器和自适应滤波方法,且对模型状态方程进行修正,建立自适应强跟踪Sigma点卡尔曼滤波算法来进行SOC估计。结果表明,ESP模型有较高的精度,且基于此模型和所建立的算法可以实现对SOC的精确估计,其最大误差在2.3%以内。The accurate battery model is the key of the battery management system of electric vehicle,which provides guarantee for the accurate estimation of battery state of charge(SOC).Based on the mechanism of electrochemical reaction,the extended single particle(ESP)model was established.According to the characteristics of ESP model with many parameters,the improved genetic algorithm was used to identify the parameters in different SOC stages,and the accuracy of the model was verified by charging and discharging experiments.Based on ESP model and Kalman filter algorithm,the strong tracking filter and adaptive filtering method were introduced,and the state equation of the model was modified.An adaptive strong tracking Sigma point Kalman filter algorithm was established to estimate SOC.The results show that the ESP model has high accuracy,and based on the model and the algorithm,the SOC can be accurately estimated with the maximum error of less than 2.3%.

关 键 词：扩展单粒子模型 参数辨识 改进遗传算法 荷电状态估计 卡尔曼滤波 

分 类 号：TM912[电气工程—电力电子与电力传动]