摘要
针对电池模型参数辨识不准确及扩展卡尔曼滤波(EKF)法无法正确确定外界噪声的影响,导致锂离子电池荷电状态(SOC)估计误差偏大的问题,提出一种遗忘因子递推最小二乘(FFRLS)-自适应迭代策略的EKF(AIEKF)算法。以双极化等效电路模型为基础,先利用FFRLS进行在线参数辨识,再将所辨识的各参数传给由EKF和迭代策略结合得到的AIEKF,完成对SOC估计。基于MATLAB进行仿真验证,用SOC估计的误差曲线、平均绝对误差及均方根误差的数值进行对比。相较于FFRLS-EKF算法,所提FFRLS-AIEKF算法的SOC估计精度更高,最大估计误差为1.6%。
Aiming at the problems of inaccurate identification of parameters in battery models and the inability of the extended Kalman filter(EKF)method to correctly determine the influence of external noise,which leads to a large error in the estimation of the state of charge(SOC)of Li-ion battery,a forgetting-factor recursive least squares(FFRLS)-adaptive iterative strategy EKF(AIEKF)algorithm is proposed.Based on the dual-polarization equivalent circuit model,the online parameter identification is carried out using the FFRLS,then the identified parameters are passed to the AIEKF obtained by combining the EKF and the iterative strategy,so as to complete the estimation of SOC.A simulation verification based on MATLAB is carried out to compare the error curves of SOC estimation and the values of the mean absolute error and the root mean square error.The proposed FFRLS-AIEKF algorithm estimates the SOC with a higher accuracy compared to the FFRLS-EKF algorithm,the maximum estimation error is 1.6%.
作者
阮爱国
史仰泽
王方钦
黄开义
陈太刚
梁大鸿
陈海波
陈思文
RUAN Aiguo;SHI Yangze;WANG Fangqin;HUANG Kaiyi;CHEN Taigang;LIANG Dahong;CHEN Haibo;CHEN Siwen(Guanling Branch,CGN Guizhou Duyun Wind Power Co.,Ltd.,Guiyang 550000,Guizhou,China)
出处
《电池》
北大核心
2025年第3期529-535,共7页
Battery Bimonthly
基金
中广核新能源控股有限公司科技项目(003-CLX-F120-2024-268)。
