利用WRF（Weather Research and Forecasting）模式, 对2006年河北省张北地区某风电场区域全年回报的风速和风向， 以及与对应时间段70 m高度的测风塔实测资料进行了对比分析， 发现模式预报效果较好。利用2008年全年风电场每台风机的实际功率与对应时刻轮毂高度风速、 风向、 气温、 相对湿度和气压回报资料， 使用支持向量机（Support Vector Machine， SVM）回归方法建立了每台风机10 min一次的风电场功率预报模型， 并利用该模型进行了2009年为期一年的预报试验， 检验模型的预报性能。结果表明， 集WRF模式和SVM方法建立的风电功率预报方法具有较好的预报效果。各月预报相关系数在0.71~0.82之间， 归一化均方根误差在9.8%~16.5%之间， 归一化平均绝对误差在5\^4%~10.5%之间； 全年预报相关系数为0.79， 归一化均方根误差为13.3%， 归一化平均绝对误差为8\^3%。
Based on WRF(Weather Research and Forecasting Model) and SVM(Support Vector Machine) regression method, short-term wind power forecast system was established. In order to verify the accuracy of WRF, wind speed and wind direction in a wind farm of Zhangbei region in 2006 were hindcasted by WRF model, which were used to compare with the observed data of wind tower at 70 m height. The verification was satisfactory. Wind power forecast model of every 10 min for 30 wind turbines using SVM regression method were developed based on actual wind power recorded data and wind speed, wind direction, atmospheric temperature, relative humidity and atmospheric pressure values which hindcasted by WRF at 70 m height in 2008. For the assessing the forecasting effect of this wind power forecast model, forecasting experiments in 2009 were carried out. The results show that the method to combine WRF model with the SVM method to establish a wind power forecasting produce good prediction results, correlation coefficients ranged from 0.71 to 0.82, normalized root mean square error ranged from 9.8% to 16.5%, and normalized mean absolute error was between 5.4% and 10.5%. The whole year′s correlation coefficient is 0.79, normalized root mean square error is 13.3%, and normalized mean absolute error is 8.3%.
[1]Ma Lei, Luan Shiyan, Jiang Chuanwen, et al. A review on the forecasting of wind speed and generated power[J]. Renewable and Sustainable Energy Reviews, 2009, 13(4): 915-920.
[2]Landberg L, Watson S J. Short-term Prediction of Local Wind Conditions[J]. Bound-Layer Meteor, 1994, 70: 95-171.
[3]Focken U, Lange M, Waldl P. Previen to a wind power prediction system with an innovative upscaling algorithm[C]. 2001 European Wind Energy Association Conference EWEC'01, Copenhagen, Denmark, 2001.
[4]Zack J W, Brower M C, Bailey B H. Validating of the for Ewind model in wind forecasting application[C]. EUWEC Special Topic Conference Wind Power for the 21st Century, Kassel Germany, 2000.
[5]The ANEMOS Project: Next Generation Forecasting Of Wind Power[EB/OL]. Http://anemos.cma.fr, 2004.
[6]Spérez S, Jimenez P A, Navarro J, et al. Using the MM5 model for wind prediction in a complex terrain site[C]. European Wind Energy Conference & Exhibition EWEC 2003, Madrid, Spain, 2003.
[7]Landberg L. Short-term Prediction of Local Wind Conditions[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2001, 89(3/4): 235-245.
[8]穆海振,徐家良,杨永辉.数值模拟在上海海上风能资源评估中的应用[J].高原气象, 2007, 26(1): 196-202.
[9]王颖,张镭,胡菊,等. WRF模式对山谷城市边界层模拟能力的检验及地面气象特征分析[J].高原气象, 2010, 29(9): 1397-1407.
[10]惠小英,高晓清,桂俊祥,等.酒泉风电基地高分辨率风能资源的数值模拟[J].高原气象, 2011, 30(2): 538-544.
[11]程兴宏,陶树旺,魏磊,等.基于WRF模式和自适应偏最小二乘回归算法的风能预报试验研究[J].高原气象, 2012, 31(5): 1461-1470.
[12]Alexiadis M, Dokopoulos P, Sahsamanoglou H, et al. Short term forecasting of wind speed and related electrical power[J]. Solar Energy, 1998, 63(1): 61-68.
[13]Huang Z, Chalabi Z S. Use of time-series analysis to model and forecast wind speed[J]. Wind Engegy, 1995, 56(2-3): 311-322.
[14]Sfetsos A. A comparison of various forecasting techniques applied to mean hourly wind speed time series[J]. Renewable Energy, 2000, 21(1): 23-35.
[15]Milligan M, Schwartz M N, Wan Y. Statistical wind power forecasting for U.S. wind farms[C]. The 17th Conference on Probability and Statistics in the Atmospheric Sciences/2004 American Meteorological Society Annual Meeting Seattle, Washington, 2004, 1: 11-15.
[16]Torres J L, Garcla A, Blas M D, et al. Forecast of hourly average wind speed with ARMA models in Spain[J]. Solar Energy, 2005, 79(1): 65-77.
[17]Mohandes M A, Shafiqur R, Talal O H. A Neural networks approach for wind speed prediction[J]. Renewable Energy, 1998, 13(3): 345-354.
[18]Bilgili M, Sahin B, Yasar A. Application of artificial neural networks for the wind speed prediction of target station using reference stations data[J]. Renewable Energy, 2007, 32(14): 2350-2360.
[19]Li S H. Wind power prediction using recurrent multilayer perceptron neural networks[C]. Power Engineering Society General Meeting, 2003, 4: 2325-2330.
[20]Damousis I G, Alexiadis M C, Therocharis J B, et al. A fuzzy model for wind speed prediction and power generation in wind parks using spatial correlation[J]. IEEE Transactions on Energy Conversion, 2004, 19(2): 352-361.
[21]Andrew K, Haiyang Z, Zhe S. Wind farm power prediction: A data-mining approach[J]. Wind Energy, 2009, 34(3): 583-590.
[22]Mohandes M A, Halawani T O, Rehman S, et al. Support vector machines for wind speed prediction[J]. Renewable Energy, 2004, 29(6): 939-947.
[23]Ji G R, Han P, Zhai Y J. Wind speed forecasting based on support vector machine with forecasting error estimation[C]. Proceedings of the 6th International Conference on Machine Learning and Cybernetics, 2007: 2735-2739.
[24]Sancho S S, Emilio G O, ngel M, et al. Short term wind speed prediction based on evolutionary support vector regression algorithms[J]. Expert Systems with Applications, 2011, 38(4): 4052-4057.
[25]William C S, Joseph B K. A time-split nonhydrostatic atmospheric model for weather research and forecasting applications[J]. J Comput Phys, 2008, 227(7): 3465-3485.
[26]Zavisa J, Black T, Pyle M, et al. High resolution applications of the WRF NMM[C]. 21st Conference on Weather Analysis and Forecasting/17th Conference on Numerical Weather Prediction, Washington: American Meteorological Society, 2005.
[27]王定成,方廷健,唐毅,等.支持向量机回归理论与控制的综述[J].模式识别与人工智能, 2003, 16(2): 192-197.
[28]冯汉中,徐会明,徐琳娜. SVM方法与长江上游降水落区预报[J].高原气象, 2004, 23(1): 45-64.
[29]张学工.关于统计学习理论与支持向量机[J].自动化学报, 2000, 26(1): 32-42.
[30]Trevor H, Robert T, Jerome F, 著. 范明, 柴玉梅, 昝红英, 译. 统计学习基础[M]. 北京: 电子工业出版社, 2003: 263-272.
