基于RLS和EKF算法的全钒液流电池SOC估计  

Vanadium redox battery SOC estimation based on RLS and EKF algorithm

作　　者：邱亚[1] 李鑫[1] 陈薇[1] 魏达 段泽民[1] 

机构地区：[1]合肥工业大学电气与自动化工程学院,合肥230009 [2]湖南德沃普电气股份有限公司,湖南邵东422800

出　　处：《控制与决策》2018年第1期37-44,共8页Control and Decision

基　　金：湖南省科技重大专项项目(2016GK103);山西省重点研发计划项目(201603D112004)

摘　　要：针对全钒液流电池的荷电状态(SOC)估计精度低、估计成本较高等问题,提出一种基于递推最小二乘算法(RLS)与扩展卡尔曼滤波算法(EKF)相结合的估计方法.该方法通过RLS算法辨识所建立的钒电池数学模型参数,通过EKF算法估计钒电池的SOC,将二者结合实现电池参数发生变化时准确估计钒电池的SOC.以5 k W/30 k Wh的钒电池为对象,应用所提出的算法实现钒电池的SOC估计.结果表明,该算法可以准确估计钒电池的SOC,且可节省额外增加单片检测电池测量SOC的费用.In order to solve the problem of low accuracy and high cost of state of charge（SOC） estimation for the Vanadium redox battery, an estimation method is proposed based on the combination of the recursive least squares（RLS）algorithm and extended Kalman filtering（EKF） algorithm. The RLS algorithm is used to estimate the battery model parameters, and the EKF algorithm is used to estimated the battery SOC. The battery SOC is estimated accurately by the combination of the RLS and EKF when the battery parameters are changed. Taking the 5 k W/30 k Wh Vanadium redox battery as the object, the SOC estimation of the Vanadium redox battery is realized by using the proposed algorithm. The results show that the algorithm can estimate the SOC of the Vanadium redox battery accurately and save the cost of the SOC measure when an additional single testing cell is added.

关 键 词：全钒液流电池 SOC 扩展卡尔曼滤波 系统辨识 实时仿真 

分 类 号：TM911[电气工程—电力电子与电力传动]