基于模糊控制的扩展卡尔曼滤波SOC估计研究  

SOC Estimation with Extended Kalman Filter Based on Fuzzy Control

作　　者：方磊 陈勇 赵理 殷康胜 郑阳 

机构地区：[1]北京信息科技大学北京电动车辆协同创新中心,北京100192

出　　处：《系统仿真学报》2018年第1期325-331,共7页Journal of System Simulation

基　　金：清华大学汽车安全与节能国家重点实验室开放基金(KF16032);科技创新服务能力建设-科研基地-新能源汽车北京实验室(PXM2016_014224_000004)

摘　　要：电池荷电状态(state-of-charge,SOC)受到温度、充放电倍率、循环寿命等因素的影响,扩展卡尔曼滤波(EKF)估计电池SOC常用方法之一。针对传统扩展卡尔曼滤波电池SOC估计方法存在的观测方程误差影响SOC估算精度的问题,考虑温度、充放电倍率等因素对观测方程误差的影响,结合模糊控制理论,提出了基于模糊控制的扩展卡尔曼滤波SOC估计法。该方法建立Mamdani型模糊控制器,以温度和充放电电流作为模糊控制器的输入,观测矩阵修正系数作为模糊控制器的输出实时改进滤波过程。仿真结果表明基于模糊修正的扩展卡尔曼滤波法可以提高SOC估计精度,减小由观测方程误差造成的SOC估算误差,在实际工况中,具有较强的适应性。Many factors affect battery＇s state of charge （SOC）, such as temperature, charge/discharge rate, cycle life and so on.Extended Kalman filter （EKF） iscommonly used to estimate battery＇s SOC.The observation equation＇s error affects the accuracy of battery＇s SOC estimationusing traditional EKE. Considering the effects of temperature and charge/discharge rateon the observation equation＇s error, an SOC estimation method using EKF based on fuzzy controlis presented.Mamdani type fuzzy controller is establishedwiththe proposed method, in which temperature and charge^discharge rateare selected as thecontroller ＇sinput, and the observation matrix＇s correction coefficient isused as thecontroller＇s output to improvethefiltering process in real time.The simulationresults show that the proposed methodcan reduce the errorscoming fromthe observation equation and improve the accuracy of SOC estimation.It also has strong adaptability in practical workingcondition.

关 键 词：模糊控制 扩展卡尔曼滤波 SOC估计 锂离子电池 

分 类 号：TM912[电气工程—电力电子与电力传动]