基于无迹卡尔曼滤波的电池SOC估算与仿真  

Estimation and Simulation of Battery SOC Based on Unscented Kalman Filter

作　　者：张博远 李立伟 王越 刘含筱 ZHANG Boyuan;LI Liwei;WANG Yue;LIU Hanxiao(College of Electrical Engineering,Qingdao University,Qingdao 266071)

机构地区：[1]青岛大学电气工程学院,青岛266071

出　　处：《计算机与数字工程》2021年第3期589-594,共6页Computer & Digital Engineering

基　　金：山东省科技发展计划项目(编号:2011GGB01123);山东省重点研发计划项目(编号:2017GGX50114)资助。

摘　　要：基于一阶Thevenin模型的扩展卡尔曼滤波在实际工程应用中,因为要对系数求其雅各比矩阵,略去了高阶项所表示的部分电池特征,在电池电流变化剧烈情况下极易失真,不能真实地反映电池状态。论文提出了基于二阶Thevenin模型的无迹卡尔曼滤波算法,二阶模型本身就能更加真实地反映电池状态,同时该算法不是对数据进行切割处理,而是经过UT变化后数据简化的同时并没有丢失其本来特征。在不损失精度的前提下提高了算法对数据的处理效率。经实验验证其误差控制在2%以内,且算法本身具有很好的鲁棒性和收敛性。Extended Kalman filter based on first-order Thevenin model is applied in practical engineering.Because the Jacobi⁃an matrix of the coefficient is calculated,some characteristics of batteries expressed by higher-order terms are omitted.It is easy to be distorted under the condition of severe change of battery current and can't reflect the real state of batteries.In this paper,an un⁃scented Kalman filtering algorithm based on the second-order Thevenin model is proposed.The second-order model itself can more truly reflect the state of the battery.At the same time,the algorithm does not cut the data,but simplifies the data after UT changes without losing its original characteristics.The algorithm improves the efficiency of data processing without losing accuracy.The ex⁃perimental results show that the error is controlled within 2%,and the algorithm itself has good robustness and convergence.

关 键 词：锂电池 BMS 二阶Thevenin模型 无迹卡尔曼滤波 荷电状态 

分 类 号：TM912.2[电气工程—电力电子与电力传动]