基于遺傳算法改進(jìn)BP神經(jīng)網(wǎng)絡(luò)的風(fēng)電功率預(yù)測研究
王冰冰,趙天樂
(南京理工大學(xué) 自動化學(xué)院,江蘇 南京 210094)
摘 要:風(fēng)電功率預(yù)測對于風(fēng)電場和電網(wǎng)的安全可靠運(yùn)行具有重要意義。以某風(fēng)力發(fā)電機(jī)為研究對象,根據(jù)該風(fēng)機(jī)歷史天氣信息和風(fēng)電功率數(shù)據(jù),使用遺傳算法改進(jìn)BP神經(jīng)網(wǎng)絡(luò),構(gòu)建復(fù)合型神經(jīng)網(wǎng)絡(luò)的風(fēng)電功率預(yù)測系統(tǒng)。運(yùn)用MATLAB軟件對算法進(jìn)行編程與仿真,仿真結(jié)果表明,單一的BP神經(jīng)網(wǎng)絡(luò)預(yù)測系統(tǒng)波動性較高,精度不足,而復(fù)合型的神經(jīng)網(wǎng)絡(luò)算法有效地解決了這一問題,改進(jìn)后的預(yù)測系統(tǒng)精度較高、穩(wěn)定性較強(qiáng),滿足工業(yè)生產(chǎn)需求。
關(guān)鍵詞:風(fēng)電;功率預(yù)測;BP 神經(jīng)網(wǎng)絡(luò);遺傳算法
中圖分類號:TM614 文獻(xiàn)標(biāo)識碼:A 文章編號:1007-3175(2019)12-0016-06
Research on Wind Power Prediction Based on Improved BP Neural Network of Genetic Algorithm
WANG Bing-bing, ZHAO Tian-le
(School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China)
Abstract: Wind power prediction is of great significance for the safe and reliable operation of wind farms and power system. Taking a wind turbine as the research object, according to the historical weather information and power generation data of the turbine, the BP neural network was improved by genetic algorithm, and a composite neural network wind power prediction system was constructed. The arithmetic was programmed and simulated by MATLAB. The simulation results show that the single BP neural network prediction system has high fluctuation and insufficient precision, however, the compound neural network algorithm effectively solves this problem. The improved prediction system has high accuracy and stability, and meets the requirements of industrial production.
Key words: wind power; power prediction; BP neural network; genetic algorithm
參考文獻(xiàn)
[1] CHEN Daojun,GONG Qingwu,ZOU Bichang,et al. A low-carbon dispatch model in a wind power integrated system considering wind speed forecasting and energy-environmental efficiency[J].Energies,2012,5(4):1245-1270.
[2] 黎靜華,桑川川,甘一夫,等. 風(fēng)電功率預(yù)測技術(shù)研究綜述[J]. 現(xiàn)代電力,2017,34(3):1-11.
[3] 陳道君,龔慶武,金朝意,等. 基于自適應(yīng)擾動量子粒子群算法參數(shù)優(yōu)化的支持向量回歸機(jī)短期風(fēng)電功率預(yù)測[J]. 電網(wǎng)技術(shù),2013,37(4):974-980.
[4] 何東,劉瑞葉. 基于主成分分析的神經(jīng)網(wǎng)絡(luò)動態(tài)集成風(fēng)功率超短期預(yù)測[J]. 電力系統(tǒng)保護(hù)與控制,2013,41(4):50-54.
[5] 袁鐵江,晁勤,李義巖,等. 大規(guī)模風(fēng)電并網(wǎng)電力系統(tǒng)經(jīng)濟(jì)調(diào)度中風(fēng)電場出力的短期預(yù)測模型[J].中國電機(jī)工程學(xué)報(bào),2010,30(13):23-27.
[6] 黃辰,吳峻青. 基于人工神經(jīng)網(wǎng)絡(luò)的短期風(fēng)電功率預(yù)測[J]. 華東電力,2014,42(7):1408-1410.
[7] 范高峰,王偉勝,劉純. 基于人工神經(jīng)網(wǎng)絡(luò)的風(fēng)電功率短期預(yù)測系統(tǒng)[J]. 電網(wǎng)技術(shù),2008,32(22):72-76.
[8] 李剛,吳潮,趙建平. 基于改進(jìn)果蠅神經(jīng)網(wǎng)絡(luò)的短期風(fēng)電功率預(yù)測[J]. 測控技術(shù),2018,37(7):23-31.
[9] 楊志凌,劉永前. 應(yīng)用粒子群優(yōu)化算法的短期風(fēng)電功率預(yù)測[J]. 電網(wǎng)技術(shù),2011,35(5):159-164.
[10] 楊琦,張建華,王向峰,等. 基于小波- 神經(jīng)網(wǎng)絡(luò)的風(fēng)速及風(fēng)力發(fā)電量預(yù)測[J]. 電網(wǎng)技術(shù),2009,33(17):44-48.
[11] 肖遷,李文華,李志剛,等. 基于改進(jìn)的小波-BP神經(jīng)網(wǎng)絡(luò)的風(fēng)速和風(fēng)電功率預(yù)測[J]. 電力系統(tǒng)保護(hù)與控制,2014,42(15):80-86.
[12] 劉純,范高峰,王偉勝,等. 風(fēng)電場輸出功率的組合預(yù)測模型[J]. 電網(wǎng)技術(shù),2009,33(13):74-79.
[13] 彭小圣,熊磊,文勁宇,等. 風(fēng)電集群短期及超短期功率預(yù)測精度改進(jìn)方法綜述[J]. 中國電機(jī)工程學(xué)報(bào),2016,36(23):14-25.
[14] BARBOUNIS T G, THEOCHARIS J B, ALEXIADIS M C, et al. Long-term wind speed and power forecasting using local recurrent neural network models[J].IEEE Transactions on Energy Conversion,2006,21(1):273-284.