風(fēng)電場發(fā)電功率組合預(yù)測方法研究
胡婷1，劉觀起1，邵龍1，劉哲2，孫勃3
1 華北電力大學(xué)，河北 保定 071000；
2 河北省電力科學(xué)研究院，河北 石家莊 050000；
3 圖們市供電分公司，吉林 延邊 133100
摘 要：針對風(fēng)電場發(fā)電功率的短期預(yù)測，闡述了組合預(yù)測的方法原理。分別建立基于相空間重構(gòu)的RBF 神經(jīng)網(wǎng)絡(luò)模型、時間序列模型、支持向量機(jī)模型三種單項(xiàng)預(yù)測模型，并在此基礎(chǔ)上確立加權(quán)系數(shù)，得到了兩個組合預(yù)測模型。預(yù)測結(jié)果顯示組合預(yù)測較單項(xiàng)預(yù)測的效果有了很大的改善，具有實(shí)際意義和應(yīng)用價值。
關(guān)鍵詞： 神經(jīng)網(wǎng)絡(luò)；時間序列；支持向量機(jī)；組合預(yù)測
中圖分類號：TM614 文獻(xiàn)標(biāo)識碼：A 文章編號：1007-3175(2013)05-0023-05
Study on Combination Model of Wind Power Generation Prediction
HU Ting1, LIU Guan-qi1, SHAO Long1, LIU Zhe2, SUN Bo3
1 North China Electrical Power University, Baoding 071000, China;
2 Electric Power Research Institute of Hebei Province, Shijiazhuang 050000, China;
3 Tumen Power Supply Subsidiary, Yanbian 133100, China
Abstract: Aiming at short-term predication of wind generation power, this paper described the method and principle of combined prediction. This paper constructed three kinds of single predicting models, including radial basis function (RBF) neural network model based on phase space reconstruction, time series model and support vector machine model, and on this basis, weighing coefficients were determined to get two groups of combined prediction models. The predicting result shows that the effect of combined prediction is improved more than that of single prediction, with practical significance and applicable value.
Key words: neural network; time series; support vector machine; combined prediction
參考文獻(xiàn)
[1] 雷亞洲.與風(fēng)電并網(wǎng)相關(guān)的研究課題[J].電力系統(tǒng)自動化，2003，27(8)：84-88.
[2] 顧為東.大規(guī)模非并網(wǎng)風(fēng)電系統(tǒng)開發(fā)與應(yīng)用[J].電力系統(tǒng)自動化，2008，32(19)：1-4.
[3] 舒進(jìn)，張保會，李鵬，等.變速恒頻風(fēng)電機(jī)組運(yùn)行控制[J].電力系統(tǒng)自動化，2008，32(16)：89-93.
[4] Billinton R, Guang Bai. Generating capacity adequacy associated with wind energy[J]. IEEE Transactions on Energy Conversion, 2004,19(3)：641-646.
[5] Karki R, Hu PO. Wind power simulation model for reliability evaluation[C]//Proceedings of Canadian Conference on Electrical and Computer Engineering, 2005：541-544.
[6] 楊秀媛，肖洋，陳樹勇.風(fēng)電場風(fēng)速和發(fā)電功率預(yù)測研究[J].中國電機(jī)工程學(xué)報，2005，25(11)：1-5.
[7] 丁明，張立軍，吳義純. 基于時間序列分析的風(fēng)電場風(fēng)速預(yù)測模型[J]. 電力自動化設(shè)備，2005，25(8)：32-34.
[8] Fonte P M, Quadrado J C. ANN approach to WECS power forecast[C]//Proceedings of IEEE Conference on Emerging Technologies and Factory Automation, 2005, 1：1069-1072.
[9] Rohrig K, Lange B. Application of wind power prediction tools for power system operations[C]//Proceedings of IEEE Power Engineering Society General Meeting, 2006：5.
[10] Pinson P, Kariniotakis G N. Wind power forecasting using fuzzy neural networks enhanced with online prediction risk assessment[C]//IEEE Power Technology Conference Proceedings, 2003, 2：8.
[11] Karki R, Hu PO, Billinton R. A simplified wind power generation model for reliability evaluation[J].IEEE Transactions on Energy Conversion, 2006, 21(2)：533-540.
[12] Mohanders M A, Halawani T O, Rehman S, Hussain Ahmed A. Support vector machines for wind speed prediction[J]. Renewable Energy, 2004,29(16)：939-947.
[13] Salcedo-Sanz Sancho, Ortiz-Garcia Emilio G, Perez-Bellido Angel M, Portilla-Figueras Antonio, Prieto Luis. Short term wind speed prediction based on evolutionary support vector regression algorithms[J]. Expert Systems with Applications, 2011, 38(4)：4052-4057.
[14] 牛東曉，曹樹華.電力負(fù)荷預(yù)測技術(shù)及其應(yīng)用[M].北京：中國電力出版社，2006.
[15] 張國強(qiáng)， 張伯明. 基于組合預(yù)測的風(fēng)電場風(fēng)速及風(fēng)電機(jī)功率預(yù)測[J]. 電力系統(tǒng)自動化，2009，33(18)：92-95.
