基于KELM-AdaBoost方法的短期风电功率预测（英文）
|
摘要
针对短期风电功率预测系统,提出基于集成学习理论的具有数据集实例权重更新机制的KELM-AdaBoost方法。Ada Boost方法能够自动学习多个弱回归器并将其提升为预测精度高的强回归器,核极限学习机(Kernel Extreme Learning Machine, KELM)方法作为AdaBoost方法的基学习器,其以核函数表示未知的隐含层非线性特征映射,仅需通过正则化最小二乘算法调节网络的输出权值就能达到最小的训练误差,且KELM中不仅使用了RBF核函数,还使用了可允许的多维张量积小波核函数。将KELM-AdaBoost方法分别应用于不同地区的短期风电功率单步直接预测和多步间接预测中,并与RBF,SVM, ELM, KELM, RBF-AdaBoost, SVM-AdaBoost, ELM-AdaBoost方法在同等条件下相比较,实验结果表明,所提出的KELM-AdaBoost方法在预测精度上优于已有的预测方法,蕴藏着巨大潜力和较好的应用前景。
For short-term wind power forecasting, a KELM-AdaBoost method with weight update mechanism for a data set instance is proposed based on ensemble learning theory. The AdaBoost method can automatically learn multiple weak regressors and boost them into an arbitrarily accurate strong regressor, meanwhile, using kernel extreme learning machine(KELM) as the base learner of the AdaBoost method, which only adjusts the output weights of networks by using the regularization least square algorithm to achieve the minimum training error and the unknown nonlinear feature mapping of the hidden layer is represented with a kernel function, and the KELM method not only uses the RBF kernel function, but also uses the permissible multi-dimension tensor product wavelet kernel function. The proposed KELM-AdaBoost method is applied to the single-step direct forecasting of short-term wind power and the multi-step indirect forecasting in different regions respectively, and the validity of the KELM-AdaBoost method is verified by comparing its accuracy with RBF, SVM, ELM, KELM, RBF-AdaBoost, SVM-AdaBoost, ELM-AdaBoost methods under the same condition, the experiment results show that the proposed KELM-AdaBoost method is superior to the existing forecasting methods on the forecasting accuracy, therefore, it contains a huge potential and good application prospect.
For short-term wind power forecasting, a KELM-AdaBoost method with weight update mechanism for a data set instance is proposed based on ensemble learning theory. The AdaBoost method can automatically learn multiple weak regressors and boost them into an arbitrarily accurate strong regressor, meanwhile, using kernel extreme learning machine(KELM) as the base learner of the AdaBoost method, which only adjusts the output weights of networks by using the regularization least square algorithm to achieve the minimum training error and the unknown nonlinear feature mapping of the hidden layer is represented with a kernel function, and the KELM method not only uses the RBF kernel function, but also uses the permissible multi-dimension tensor product wavelet kernel function. The proposed KELM-AdaBoost method is applied to the single-step direct forecasting of short-term wind power and the multi-step indirect forecasting in different regions respectively, and the validity of the KELM-AdaBoost method is verified by comparing its accuracy with RBF, SVM, ELM, KELM, RBF-AdaBoost, SVM-AdaBoost, ELM-AdaBoost methods under the same condition, the experiment results show that the proposed KELM-AdaBoost method is superior to the existing forecasting methods on the forecasting accuracy, therefore, it contains a huge potential and good application prospect.
引文
[1]Yan J,Liu Y Q,Han S,et al.Reviews on uncertainty analysis of wind power forecasting[J].Renewable and Sustainable Energy Reviews,2015,52:1322-1330.
[2]Aggarwal S K,Gupta M.Wind power forecasting:a review of statistical models-wind power forecasting[J].International Journal of Energy Science,2013,3(1):1-10.
[3]Jaesung Jung,Robert P.Broadwater.Current status and future advances for wind speed and power forecasting[J].Renewable and Sustainable Energy Reviews,2014,31(1):762-777.
[4]刘瑞叶,黄磊.基于动态神经网络的风电场输出功率预测[J].电力系统自动化,2012,36(11):19-22,37.Liu R Y,Huang L.Wind power forecasting based on dynamic neural networks[J].Automation of Electric Power Systems,2012,36(11):19-22,37.
[5]王丽婕,冬雷,缪晓钟,等.基于小波分析的风电场短期发电功率预测[J].中国电机工程学报,2009,29(28):30-33.Wang L J,Dong L,Liao X Z,et al.Short-term power prediction of a wind farm based on wavelet analysis[J].Proceedings of the CSEE,2009,29(28):30-33.
[6]Seo I-Y,Ha B-N,Lee S-W,et al.Short-Term wind power prediction using fuzzy clustering and support vector regression[J].Journal of Energy Power Engineering,2012,6(10):1605-1610.
[7]Freund Y,Schapire R E.A decision-theoretic generation of on-line learning and an application to boosting[J].Journal of Computer and System Sciences.1997,55(1):119-139.
[8]高云龙,潘金艳,吉国力,等.基于Boosting梯度下降理论的时间序列建模方法[J].中国科学:技术科学,2011,41(7):929-943.Gao Y L,Pan J Y,Ji G L,et al.Time series modeling based on the theory of the Boosting gradient descent method[J].SCIENTIASLNICA Technological,2011,41(7):929-943.
[9]寇鹏,高峰.几何转换boosting回归算法及其在高耗能企业负荷预测中的应用[J].系统工程理论与实践,2013,33(7):1880-1888.Kou P,Gao F.Boosting regression method based on geometric conversion and its application to in energy-intensive enterprise[J].Systems engineering-theory and practice,2013,33(7):1880-1888.
[10]Gao Feng.Boosting regression methods based on a geometric conversion approach:Using svms base learners[J].Neurocomputing2013,113:67-87.
[11]Huang G B,Zhu Q Y,Siew C K.Extreme learning machine:theory and applications[j].Neurocomputing,2006,70(1):489-501.
[12]刘士荣,李松峰,宁康红,等.基于极端学习机的光伏发电功率短期预测[J].控制工程,2013,20(2):386-390.Liu S R,Li S F,Ning K H,et al.Short-Term forecasting of PVcapacity based on extreme learning machine in similar days[J].Control Engineering of China,2013,20(2):386-390.
[13]Shabbir A,Verdoolaege G,Vega J,et al.ELM regime classification by conformal prediction on an information manifold[J].IEEETransactions on Plasma Science,2015,43(12):4190-4199.
[14]Huang G B,Zhou H,Ding X,et al.Extreme learning machine for regression and multiclass classification[J].IEEE Transactions on Systems,Man,and Cybernetics,Part B.2012,42(2):513-529.
[15]周杰,矩阵分析及应用[M].四川,四川大学出版社,2009:180-181.Zhou J.Matris analysis and application[M].Xian,Si Chuan university press,2009:180-181.
[16]Nan-Ying Liang,Guang-Bin Huang.A fast and accurate online sequential learning algorithm for feedforward networks[J].IEEETransactions on Neural Networks,2006,17(6):1411-1423.
[17]Duffy N,H.D.Boosting method for regression[J].Machine Learning,2002,47(2-3):153-200.
[18]Shrestha D L,Solomatine D P.Experiments with AdaBoost.RT,an Improved Boosting Scheme for regression[J].Neural Computation,2006,18(7):1678-1710
[19]李军,郭林.基于WKGV-KICA的盲源信号分离算法[J].控制与决策,2013,28(7):972-977Li J,Guo L.Blind source separation algorithm based on WKGV-KICAalgorithm[J].Control and Decision,2013,28(7):972-977.
[20]Hu Q,Zhang S,Yu M,et al.Short-Term Wind Speed or Power Forecasting With Heteroscedastic Support Vector Regression[J].IEEE Transactions on Sustainable Energy,2016,7(1):241-249.
[21]Jung J,Broadwater R P.Current status and future advances for wind speed and power forecasting[J].Renewable&Sustainable Energy Reviews,2014,31(2):762-777.
[22]Guglielmo D'Amicoa;Filippo Petronib;Flavio Pratticoc.Wind speed prediction for wind farm applications by Extreme Value Theory and Copulas[J].Journal of Wind Engineering&Industrial Aerodynamics,2015,22(6):229-236.
[23]Alberta Electric System Operator(AESO),wind power integration.http://www.aeso.ca/gridoperations/13902.html.
[24]Potter C W,Lew D,Mc Caa J,et al.Creating the dataset for the western wind and solar integration study(U.S.A)[J].Wind Engineering,2008,32(4):325-338.
[2]Aggarwal S K,Gupta M.Wind power forecasting:a review of statistical models-wind power forecasting[J].International Journal of Energy Science,2013,3(1):1-10.
[3]Jaesung Jung,Robert P.Broadwater.Current status and future advances for wind speed and power forecasting[J].Renewable and Sustainable Energy Reviews,2014,31(1):762-777.
[4]刘瑞叶,黄磊.基于动态神经网络的风电场输出功率预测[J].电力系统自动化,2012,36(11):19-22,37.Liu R Y,Huang L.Wind power forecasting based on dynamic neural networks[J].Automation of Electric Power Systems,2012,36(11):19-22,37.
[5]王丽婕,冬雷,缪晓钟,等.基于小波分析的风电场短期发电功率预测[J].中国电机工程学报,2009,29(28):30-33.Wang L J,Dong L,Liao X Z,et al.Short-term power prediction of a wind farm based on wavelet analysis[J].Proceedings of the CSEE,2009,29(28):30-33.
[6]Seo I-Y,Ha B-N,Lee S-W,et al.Short-Term wind power prediction using fuzzy clustering and support vector regression[J].Journal of Energy Power Engineering,2012,6(10):1605-1610.
[7]Freund Y,Schapire R E.A decision-theoretic generation of on-line learning and an application to boosting[J].Journal of Computer and System Sciences.1997,55(1):119-139.
[8]高云龙,潘金艳,吉国力,等.基于Boosting梯度下降理论的时间序列建模方法[J].中国科学:技术科学,2011,41(7):929-943.Gao Y L,Pan J Y,Ji G L,et al.Time series modeling based on the theory of the Boosting gradient descent method[J].SCIENTIASLNICA Technological,2011,41(7):929-943.
[9]寇鹏,高峰.几何转换boosting回归算法及其在高耗能企业负荷预测中的应用[J].系统工程理论与实践,2013,33(7):1880-1888.Kou P,Gao F.Boosting regression method based on geometric conversion and its application to in energy-intensive enterprise[J].Systems engineering-theory and practice,2013,33(7):1880-1888.
[10]Gao Feng.Boosting regression methods based on a geometric conversion approach:Using svms base learners[J].Neurocomputing2013,113:67-87.
[11]Huang G B,Zhu Q Y,Siew C K.Extreme learning machine:theory and applications[j].Neurocomputing,2006,70(1):489-501.
[12]刘士荣,李松峰,宁康红,等.基于极端学习机的光伏发电功率短期预测[J].控制工程,2013,20(2):386-390.Liu S R,Li S F,Ning K H,et al.Short-Term forecasting of PVcapacity based on extreme learning machine in similar days[J].Control Engineering of China,2013,20(2):386-390.
[13]Shabbir A,Verdoolaege G,Vega J,et al.ELM regime classification by conformal prediction on an information manifold[J].IEEETransactions on Plasma Science,2015,43(12):4190-4199.
[14]Huang G B,Zhou H,Ding X,et al.Extreme learning machine for regression and multiclass classification[J].IEEE Transactions on Systems,Man,and Cybernetics,Part B.2012,42(2):513-529.
[15]周杰,矩阵分析及应用[M].四川,四川大学出版社,2009:180-181.Zhou J.Matris analysis and application[M].Xian,Si Chuan university press,2009:180-181.
[16]Nan-Ying Liang,Guang-Bin Huang.A fast and accurate online sequential learning algorithm for feedforward networks[J].IEEETransactions on Neural Networks,2006,17(6):1411-1423.
[17]Duffy N,H.D.Boosting method for regression[J].Machine Learning,2002,47(2-3):153-200.
[18]Shrestha D L,Solomatine D P.Experiments with AdaBoost.RT,an Improved Boosting Scheme for regression[J].Neural Computation,2006,18(7):1678-1710
[19]李军,郭林.基于WKGV-KICA的盲源信号分离算法[J].控制与决策,2013,28(7):972-977Li J,Guo L.Blind source separation algorithm based on WKGV-KICAalgorithm[J].Control and Decision,2013,28(7):972-977.
[20]Hu Q,Zhang S,Yu M,et al.Short-Term Wind Speed or Power Forecasting With Heteroscedastic Support Vector Regression[J].IEEE Transactions on Sustainable Energy,2016,7(1):241-249.
[21]Jung J,Broadwater R P.Current status and future advances for wind speed and power forecasting[J].Renewable&Sustainable Energy Reviews,2014,31(2):762-777.
[22]Guglielmo D'Amicoa;Filippo Petronib;Flavio Pratticoc.Wind speed prediction for wind farm applications by Extreme Value Theory and Copulas[J].Journal of Wind Engineering&Industrial Aerodynamics,2015,22(6):229-236.
[23]Alberta Electric System Operator(AESO),wind power integration.http://www.aeso.ca/gridoperations/13902.html.
[24]Potter C W,Lew D,Mc Caa J,et al.Creating the dataset for the western wind and solar integration study(U.S.A)[J].Wind Engineering,2008,32(4):325-338.