基于AEKF滤波与H_(∞)滤波的锂离子电池SOC联合估计  

State of charge estimation for lithium ion battery based on adaptive extended Kalman filter and H_(∞) filter algorithm

作　　者：王雨妍 李翔晟[1] 陈志峰 欧阳立芳 蒋宇阳 WANG Yuyan;LI Xiangsheng;CHEN Zhifeng;OUYANG Lifang;JIANG Yuyang(School of Mechanical and Electrical Engineering,Central South University of Forestry and Technology,Changsha Hunan 410004,China)

机构地区：[1]中南林业科技大学机电工程学院,湖南长沙410004

出　　处：《电源技术》2022年第5期536-540,共5页Chinese Journal of Power Sources

基　　金：湖南省自然科学基金省市联合项目(14JJ5014)。

摘　　要：锂电池荷电状态(SOC)的准确估计对提高电池的动态性能和能量利用率至关重要。针对现有卡尔曼滤波SOC估计方法存在估计精度低、鲁棒性差等问题,采用锂离子电池的二阶电阻-电容等效电路模型,通过HPPC循环脉冲实验和动态应力测试工况放电实验,结合带可变遗忘因子的递推最小二乘法(VFFRLS)及开窗理论,对等效电路模型参数进行在线辨识,提出利用自适应扩展卡尔曼滤波(AEKF)算法和H_(∞)滤波算法联合估计SOC的方法。结果表明:与AEKF算法相比,在DST工况下该算法可以使电池荷电状态估计的最大绝对误差减小3.9029%,平均绝对误差减小0.9622%,均方根误差减小0.5515%。与H_(∞)滤波算法相比,在DST工况下该算法可以使电池荷电状态估计最大绝对误差减小1.309%,平均绝对误差减小2.8934%,均方根误差减小2.6136%。The accurate estimation of the state of charge(SOC)of lithium batteries is essential to improve the dynamic performance and energy utilization of the battery.Aiming at the problems of low estimation accuracy and poor robustness existing in the existing Kalman filter SOC estimation methods,the second-order resistance-capacitance equivalent circuit model of lithium-ion battery was used,and the experimental data were obtained by HPPC cycle pulse experiment and dynamic stress test condition discharge experiment.Then combined with RLS with a variable forgetting factor and windowing theory,the equivalent circuit model parameters were identified online.Finally,the adaptive extended Kalman filter algorithm and the H_(∞)filter algorithm were used to jointly estimate the SOC.The results show that compared with the AEKF filtering algorithm,the algorithm can reduce the maximum absolute error of the battery state of charge estimation by 3.9029%,the average absolute error by 0.9622%,and the root mean square error by 0.5515%under different working conditions.Compared with the H_(∞)filtering algorithm,this algorithm can reduce the maximum absolute error of the battery state of charge estimation by 1.309%,the average absolute error by 2.8934%,and the root mean square error by 2.6136%under dynamic stress test(DST)working conditions.

关 键 词：锂离子电池 荷电状态 参数在线辨识 H无穷滤波 自适应扩展卡尔曼滤波 

分 类 号：TM912[电气工程—电力电子与电力传动]