基于改进在线核极限学习机的蓄电池SOC预测
|
摘要
为对蓄电池荷电状态(SOC)进行准确、快速的在线预测,提出一种改进的在线核极限学习机方法(IO-KELM),以电池工作电压、电流和表面温度为输入量,电池SOC为输出量建立预测模型。IO-KELM采用Cholesky分解将核极限学习机(KELM)从离线模式扩展到在线模式,使网络输出权值随新样本的逐次加入递推求解更新,以简单的四则运算替代复杂的矩阵求逆,提高了网络的泛化能力和在线学习效率。仿真实验表明,相比于KELM及直接在线建模的KELM算法(DO-KELM),IO-KELM具有更高的预测精度、更强的鲁棒性及更快的计算速度。
In order to conduct an accurate and fast online prediction for the state of charge(SOC) of battery, an improved online kernel extreme learning machine(IO-KELM) algorithm is proposed. In this work, a prediction model is presented with charge voltage, current and surface temperature as inputs and SOC of battery as output. The IO-KELM adopts Cholesky factorization to extend the kernel extreme learning machine(KELM) from offline mode to online mode. Meanwhile, the output weights of the network are updated by successive join of the new samples, and the matrix inverse operation is replaced with arithmetic. Hence, the generalization ability and the computational efficiency of the model are improved. Compared with KELM and direct online-KELM(DO-KELM) algorithm, simulation results indicate that the IO-KELM has higher prediction accuracy, stronger robustness and faster calculation speed.
In order to conduct an accurate and fast online prediction for the state of charge(SOC) of battery, an improved online kernel extreme learning machine(IO-KELM) algorithm is proposed. In this work, a prediction model is presented with charge voltage, current and surface temperature as inputs and SOC of battery as output. The IO-KELM adopts Cholesky factorization to extend the kernel extreme learning machine(KELM) from offline mode to online mode. Meanwhile, the output weights of the network are updated by successive join of the new samples, and the matrix inverse operation is replaced with arithmetic. Hence, the generalization ability and the computational efficiency of the model are improved. Compared with KELM and direct online-KELM(DO-KELM) algorithm, simulation results indicate that the IO-KELM has higher prediction accuracy, stronger robustness and faster calculation speed.
引文
[1]侯江娜,孙玉坤,王鹏飞.基于RVM算法的蓄电池SOC预测[J].电源技术,2015,3:523-526.HOU Jiangna,SUN Yukun,WANG Pengfei.SOC prediction for battery based on relevance vector machine algorithm[J].Chinese journal of power sources,2015,3:523-526.
[2]Ranjbar A H,Banaei A,Khoobroo A.Online estimation of state of charge in Li-ion batteries using impulse response concept[J].IEEE Transactions on Smart Grid(S1949-3053),2012,3(1):360-367.
[3]Alvarez Anton J C,Garcia Nieto P J,Blanco Viejo C,et al.Support vector machines used to estimate the battery state of charge[J].IEEE Transactions on Power Electronics(S0885-8993),2013,28(12):5919-5926.
[4]Van Bree P J,Veltman A,Hendrix W H A,et al.Prediction of battery behavior subject to high rate partial state of charge[J].IEEE Transactions on Vehicular Technology(S0018-9545),2009,58(2):588-595.
[5]Lee D T,Shiah S J,Lee C M,et al.State of charge estimation for electric scooters by using learning mechanisms[J].IEEE Transactions on Vehicular Technology(S0018-9545),2007,56(2):544-556.
[6]朱志莹,孙玉坤.群优化支持向量机的磁轴承转子位移预测模型[J].中国电机工程报,2012,32(33):118-123.ZHU Zhiying,SUN Yukun.Predictive modeling of rotor displacement for magnetic bearing using particle swarm optimized-least squares support vector machines[J].Proceedings of the CSEE,2012,32(33):118-123.
[7]Huang G B,Zhu H M,Ding X J,et al.Extreme leaning machine for regression and Multiclass classification[J].IEEE Transactions on Systems Man and Cybernetics(S1083-4419),Part B:Cybernetics,2012,42(2):513-529.
[8]Feng G,Huang G B,Lin Q,et al.Error minimized extreme learning machine with growth of hidden nodes and incremental learning[J].IEEE Trans on Neural Netw(S1045-9227),2009,20(8):1352-1359.
[9]张英堂,马超,李志宁,等.基于快速留一交叉验证的核极限学习机在线建模[J].上海交通大学学报,2014,48(5):641-646.ZHANG Y T,MA C,LI Z N,et al.Online modeling of kernel extreme learning machine based on fast leave-one-out cross-validation[J].Journal of Shanghai Jiao Tong University,2014,48(5):641-646.
[10]张贤达.矩阵分析与应用[M].北京:清华大学出版社,2005.Zhang Xianda.Matrix analysis and applications[M].Beijing:Tsinghua university press,2005.
[2]Ranjbar A H,Banaei A,Khoobroo A.Online estimation of state of charge in Li-ion batteries using impulse response concept[J].IEEE Transactions on Smart Grid(S1949-3053),2012,3(1):360-367.
[3]Alvarez Anton J C,Garcia Nieto P J,Blanco Viejo C,et al.Support vector machines used to estimate the battery state of charge[J].IEEE Transactions on Power Electronics(S0885-8993),2013,28(12):5919-5926.
[4]Van Bree P J,Veltman A,Hendrix W H A,et al.Prediction of battery behavior subject to high rate partial state of charge[J].IEEE Transactions on Vehicular Technology(S0018-9545),2009,58(2):588-595.
[5]Lee D T,Shiah S J,Lee C M,et al.State of charge estimation for electric scooters by using learning mechanisms[J].IEEE Transactions on Vehicular Technology(S0018-9545),2007,56(2):544-556.
[6]朱志莹,孙玉坤.群优化支持向量机的磁轴承转子位移预测模型[J].中国电机工程报,2012,32(33):118-123.ZHU Zhiying,SUN Yukun.Predictive modeling of rotor displacement for magnetic bearing using particle swarm optimized-least squares support vector machines[J].Proceedings of the CSEE,2012,32(33):118-123.
[7]Huang G B,Zhu H M,Ding X J,et al.Extreme leaning machine for regression and Multiclass classification[J].IEEE Transactions on Systems Man and Cybernetics(S1083-4419),Part B:Cybernetics,2012,42(2):513-529.
[8]Feng G,Huang G B,Lin Q,et al.Error minimized extreme learning machine with growth of hidden nodes and incremental learning[J].IEEE Trans on Neural Netw(S1045-9227),2009,20(8):1352-1359.
[9]张英堂,马超,李志宁,等.基于快速留一交叉验证的核极限学习机在线建模[J].上海交通大学学报,2014,48(5):641-646.ZHANG Y T,MA C,LI Z N,et al.Online modeling of kernel extreme learning machine based on fast leave-one-out cross-validation[J].Journal of Shanghai Jiao Tong University,2014,48(5):641-646.
[10]张贤达.矩阵分析与应用[M].北京:清华大学出版社,2005.Zhang Xianda.Matrix analysis and applications[M].Beijing:Tsinghua university press,2005.