王娜等
摘要:備選風(fēng)電場在壽命周期內(nèi)的平均年發(fā)電量是風(fēng)電場宏觀選址的一個(gè)重要參考判據(jù).為了提高風(fēng)電場平均年發(fā)電量的預(yù)測精度,提出了一種基于風(fēng)電場附近多個(gè)氣象站長期測風(fēng)數(shù)據(jù)的區(qū)域信息融合的平均年發(fā)電量預(yù)測方法.首先分別建立各氣象站與風(fēng)電場同期小時(shí)風(fēng)速之間的相關(guān)模型,應(yīng)用相關(guān)模型得到多個(gè)長期小時(shí)風(fēng)速預(yù)測值,再用神經(jīng)網(wǎng)絡(luò)對長期小時(shí)風(fēng)速預(yù)測值進(jìn)行融合處理得出最終的小時(shí)風(fēng)速預(yù)測值,在此基礎(chǔ)上進(jìn)行風(fēng)電場平均年發(fā)電量的估算.仿真結(jié)果表明:本文提出的區(qū)域信息融合方法對年平均發(fā)電量的預(yù)測誤差比采用單一氣象站數(shù)據(jù)的預(yù)測誤差最高可降低11.32%.
關(guān)鍵詞:平均年發(fā)電量;測量相關(guān)預(yù)測;信息融合;神經(jīng)網(wǎng)絡(luò)
中圖分類號:TM315 文獻(xiàn)標(biāo)識碼:A
Average Annual Energy Output Prediction Based
on Regional Information Fusion
WANG Na1, SHAO Xia1,GAO Yunpeng1,WAN Quan2
(1.College of Electrical and Information Engineering, Hunan Univ, Changsha,Hunan410082 China;
2.State Grid Hunan Electric Power Company Research Institute, Changsha,Hunan410007,China)
Abstract:Annual energy output of a candidate site in its life span is an important reference criterion of wind farm macro siting. A regional information fusion method, which allows the use of multiple reference wheather stations with a long history of wind speed and wind direction measurements, was proposed to improve the annual energy output prediction accuracy. Firstly, the correlation model was established between the shortterm wind data of a single reference wheather station and the candidate wind farm, and the multiple longterm wind speeds of candidate site based on different reference stations were predicted by using the model. Then, the multiple prediction results were integrated by neural network to obtain the final longterm hourly wind speed data, and the annual energy output was subsequently determined on the basis of the knowledge of these wind speeds. The simulation results show that, by using the proposed method, the error reduction up to 11.32% has been achieved in the relative error of the average annual power output, with respect to the case of using a single reference wheather station method.
Key words: average annual energy output; measurementcorrelatepredict(MCP); information fusion; neural network
風(fēng)能資源評估是風(fēng)電場選址的關(guān)鍵,其中備選風(fēng)電場在整個(gè)壽命周期內(nèi)的平均年發(fā)電量是一個(gè)重要的參考判據(jù).風(fēng)電場的壽命周期通常為20~25年,在此期間平均年發(fā)電量的估算受風(fēng)速變化(日變化、季節(jié)變化、年際變化)影響較大,要準(zhǔn)確地進(jìn)行評估至少需要數(shù)年甚至數(shù)十年的風(fēng)速觀測數(shù)據(jù),這樣才能減少由于風(fēng)速變化帶來的不確定性[1].但是,在實(shí)際工程的規(guī)劃階段不可能用如此長的時(shí)間來收集現(xiàn)場數(shù)據(jù).在缺乏備選風(fēng)電場長期可靠風(fēng)速記錄的情況下,廣泛采用測量-相關(guān)-預(yù)測算法(MCP,MeasureCorrelatePredict)來進(jìn)行風(fēng)資源評估,即在備選風(fēng)電場址處設(shè)立測風(fēng)塔進(jìn)行1~2年觀測,利用這個(gè)短期觀測數(shù)據(jù)和風(fēng)電場附近氣象站20~30年的歷史觀測數(shù)據(jù)進(jìn)行評估.
目前MCP算法主要有線性回歸法[2-4]、方差比法[2]、Weibull尺度法[3]、概率函數(shù)法[5]、神經(jīng)網(wǎng)絡(luò)法[1,6]和Bayesian網(wǎng)絡(luò)法[7]等.其中線性回歸法、方差比法、Weibull尺度法、概率函數(shù)法均是利用風(fēng)電場附近單一氣象站的信息進(jìn)行預(yù)測.實(shí)際上風(fēng)電場附近可能存在多個(gè)與風(fēng)電場距離較近且風(fēng)速相關(guān)性較強(qiáng)的氣象站,在這種情況下只采用一個(gè)氣象站的信息勢必會影響預(yù)測的精度.為了進(jìn)一步提高預(yù)測精度,本文提出了基于多氣象站信息即區(qū)域信息融合的平均年發(fā)電量預(yù)測算法.該算法由3部分構(gòu)成,首先分別建立各參考?xì)庀笳九c風(fēng)電場同期小時(shí)風(fēng)速之間的相關(guān)模型,應(yīng)用相關(guān)模型得到多個(gè)長期小時(shí)風(fēng)速預(yù)測值;然后用神經(jīng)網(wǎng)絡(luò)對長期小時(shí)風(fēng)速預(yù)測值進(jìn)行融合處理得出最終的預(yù)測值;最后,在長期小時(shí)風(fēng)速的基礎(chǔ)上進(jìn)行風(fēng)電場平均年發(fā)電量的估算.與文獻(xiàn)[1]和[7]相比,算法更加靈活,并且可以運(yùn)用目前成熟的MCP算法,對年平均發(fā)電量的預(yù)測誤差相比采用單一氣象站數(shù)據(jù)預(yù)測方法最高可降低11.32%.