基于扩展卡尔曼神经网络算法估计电池SOC
|
摘要
针对汽车锂电池的荷电状态(SOC)的问题,基于Thevenin电路为等效电路并且应用扩展卡尔曼算法(EKF)结合神经网络算法进行估计。在进行卡尔曼滤波算法估算过程中,需要用到实时的估算模型参数值(最新值),即在不同的SOC下模型的参数不同。传统做法是把SOC与各个参数的关系进行普通的拟合,这种方法在拟合过程中存在较大误差。为了解决这个问题,利用神经网络拟合各个电路模型参数与SOC关系曲线。试验结果表明,与单纯的扩展卡尔曼算法相比,该方法能够准确估计电池剩余电量,误差小于3%。
An extended Kalman filter algorithm(EKF) with neural network is used to estimate the state of lithium battery(SOC),which is based on Thevenin equivalent circuit.In the process of extended Kalman filter estimation,the real-time model parameters should be updated with the different SOC regard to the different SOC the different model parameters.The traditional approach which has a big error is that the fitting curve between SOC and the various separate parameters is common.To solve this problem neural net-work is applied to fit curve between the parameters of circuit model and the SOC separately.Finally,the results with the error less than 3 % show that compared with the pure extended Kalman algorithm,the method can realize the more accurate estimation of the remaining battery power.
An extended Kalman filter algorithm(EKF) with neural network is used to estimate the state of lithium battery(SOC),which is based on Thevenin equivalent circuit.In the process of extended Kalman filter estimation,the real-time model parameters should be updated with the different SOC regard to the different SOC the different model parameters.The traditional approach which has a big error is that the fitting curve between SOC and the various separate parameters is common.To solve this problem neural net-work is applied to fit curve between the parameters of circuit model and the SOC separately.Finally,the results with the error less than 3 % show that compared with the pure extended Kalman algorithm,the method can realize the more accurate estimation of the remaining battery power.
引文
[1]SCROSATI B,GARCHE J.Lithium batteries:Status,prospects and future[J].Power Sources,2010,195(9):2419-2430.
[2]孟良荣.电动车电池现状与发展趋势[J].电池工业,2006,11(3):202-206.
[3]谭晓军.电动汽车动力电池管理系统设计[M].广州:中山大学出版社,2011.
[4]麻友良,陈全世,齐占宁.电动汽车用电池SOC定义与检测方法[J].清华大学学报(自然科学版),2001,41(11):95-97.
[5]时玮,姜久春,李索宇,等.磷酸铁锂电池SOC估算方法研究[J].电子测量与仪器学报,2010,24(8):769-774.
[6]尹安东,张万兴.基于神经网络的磷酸铁锂电池SOC预测研究[J].电子测量与仪器学报,2011,25(5):433-437.
[7]凌国维.基于人工神经网络理论的电动汽车用锂离子动力电池组智能管理系统的研究[D].天津:天津大学,2006.
[8]何耀.动力锂电池组状态估计策略及管理系统技术研究[D].合肥:中国科学技术大学,2012.
[9]姜久春,文锋,温家鹏,等.纯电动汽车用锂离子电池的建模和模型参数识别[J].电力科学与技术学报,2010,25(1):67-74.
[10]林成涛,仇斌,陈世全.电流输入电动汽车电池等效电路模型的比较[J].机械工程学报,2005,41(12):76-81.
[11]孙东,陈息坤,毛华夫.基于离散滑模观测器的锂电池荷电状态估计[J].中国电机工程学报,2015(1):185-191.
[12]盛仲飙,同晓荣.BP神经网络在曲线拟合中的应用[J].科学技术与工程,2011,1(28):55-57.
[13]陈光,任志良,孙海柱.最小二乘曲线拟合及Matlab实现[J].兵工自动化,2005,24(3):107-108.
[14]SAEED S,REZA G,BOR Y L.A novel on-board stateof-charge estimation method for aged Li-ion batteries based on model adaptive extended Kalman filter[J].Journal of Power Sources,2014,245(1):337-344.
[15]CHARKHGARD M,FARROKHI M.State-of-charge estimation for lithium-ion batteries using neural networks and EKF[J].IEEE Transactions on Industrial Electronics,2010,57(12):4178-4187.
[2]孟良荣.电动车电池现状与发展趋势[J].电池工业,2006,11(3):202-206.
[3]谭晓军.电动汽车动力电池管理系统设计[M].广州:中山大学出版社,2011.
[4]麻友良,陈全世,齐占宁.电动汽车用电池SOC定义与检测方法[J].清华大学学报(自然科学版),2001,41(11):95-97.
[5]时玮,姜久春,李索宇,等.磷酸铁锂电池SOC估算方法研究[J].电子测量与仪器学报,2010,24(8):769-774.
[6]尹安东,张万兴.基于神经网络的磷酸铁锂电池SOC预测研究[J].电子测量与仪器学报,2011,25(5):433-437.
[7]凌国维.基于人工神经网络理论的电动汽车用锂离子动力电池组智能管理系统的研究[D].天津:天津大学,2006.
[8]何耀.动力锂电池组状态估计策略及管理系统技术研究[D].合肥:中国科学技术大学,2012.
[9]姜久春,文锋,温家鹏,等.纯电动汽车用锂离子电池的建模和模型参数识别[J].电力科学与技术学报,2010,25(1):67-74.
[10]林成涛,仇斌,陈世全.电流输入电动汽车电池等效电路模型的比较[J].机械工程学报,2005,41(12):76-81.
[11]孙东,陈息坤,毛华夫.基于离散滑模观测器的锂电池荷电状态估计[J].中国电机工程学报,2015(1):185-191.
[12]盛仲飙,同晓荣.BP神经网络在曲线拟合中的应用[J].科学技术与工程,2011,1(28):55-57.
[13]陈光,任志良,孙海柱.最小二乘曲线拟合及Matlab实现[J].兵工自动化,2005,24(3):107-108.
[14]SAEED S,REZA G,BOR Y L.A novel on-board stateof-charge estimation method for aged Li-ion batteries based on model adaptive extended Kalman filter[J].Journal of Power Sources,2014,245(1):337-344.
[15]CHARKHGARD M,FARROKHI M.State-of-charge estimation for lithium-ion batteries using neural networks and EKF[J].IEEE Transactions on Industrial Electronics,2010,57(12):4178-4187.