基于FF-MILS的锂离子电池模型参数辨识  

Parameter identification of Li-ion battery model based on FF-MILS

作　　者：李争光 魏娟[1] 田海波[1] 侯效东 LI Zheng-guang;WEI Juan;TIAN Hai-bo;HOU Xiao-dong(College of Mechanical Engineering,Xi′an University of Science and Technology,Xi′an,Shaanxi 710054,China)

机构地区：[1]西安科技大学机械工程学院,陕西西安710054

出　　处：《电池》2021年第1期46-49,共4页Battery Bimonthly

基　　金：国家自然科学基金青年项目(51705412)。

摘　　要：以三元正极材料锂离子电池为研究对象,选用二阶RC电池模型,采用遗忘因子多新息最小二乘算法(FF-MILS)进行在线参数辨识。比较带有遗忘因子最小二乘算法(FFRLS)与遗忘因子多新息最小二乘算法辨识结果估计的端电压与实测端电压的绝对误差,以验证参数辨识效果。实验结果表明,在城市道路循环工况(UDDS)下,遗忘因子多新息最小二乘算法的平均绝对误差比未改进的算法减少了0.5%。Ternary cathode material Li-ion battery was used as the research object with the selection of second-order RC battery model,the forgetting factor multi-innovation least square algorithm(FF-MILS)was used for online parameter identification.The absolute error between terminal voltage and measured terminal voltage which identified by the forgetting factor least squares algorithm(FFRLS)and the forgetting factor multi-innovation least square algorithm were compared,the parameter identification effect was verified.The experiment results showed that the average absolute error of the algorithm was reduced by 0.5%compared with the original algorithm under the urban dynamometer driving schedule(UDDS).

关 键 词：电池模型 多新息 遗忘因子 参数辨识 

分 类 号：TM912.9[电气工程—电力电子与电力传动]