【摘要】：隨著我國能源結(jié)構(gòu)的調(diào)整,風(fēng)電裝機(jī)容量快速增長,及時(shí)準(zhǔn)確地預(yù)測風(fēng)電功率可為電網(wǎng)合理調(diào)度提供重要依據(jù),減少棄風(fēng),有效地提高風(fēng)電利用率。同時(shí),隨著風(fēng)電場智能化水平的提高,風(fēng)電監(jiān)測數(shù)據(jù)規(guī)模不斷增長,對(duì)傳統(tǒng)風(fēng)電功率預(yù)測模型的計(jì)算性能提出了新的挑戰(zhàn)。近年來,以機(jī)器學(xué)習(xí)理論為基礎(chǔ)的人工神經(jīng)網(wǎng)絡(luò)法和支持向量機(jī)法及其改進(jìn)算法在短期風(fēng)電功率預(yù)測中得到廣泛應(yīng)用,機(jī)器學(xué)習(xí)算法中存在較多迭代計(jì)算場景,云計(jì)算技術(shù)中的Spark分布式內(nèi)存計(jì)算框架,可高效進(jìn)行迭代式數(shù)據(jù)處理,有效提高算法的執(zhí)行性能。針對(duì)現(xiàn)有短期風(fēng)電功率預(yù)測模型存在泛化性較弱、模型結(jié)構(gòu)和參數(shù)確定困難、可解釋性差等問題,本文綜合隨機(jī)森林回歸算法、M5P模型樹、差分進(jìn)化算法、選擇性集成方法,提出了一種基于改進(jìn)隨機(jī)森林回歸算法的短期風(fēng)電功率預(yù)測方法,并采用Spark云計(jì)算平臺(tái)實(shí)現(xiàn)算法并行化,主要開展了以下幾個(gè)方面的研究工作:(1)傳統(tǒng)隨機(jī)森林回歸算法以分類回歸樹為元決策樹,針對(duì)分類回歸樹預(yù)測精度較低、不能給出一個(gè)連續(xù)的輸出且預(yù)測值無法超出訓(xùn)練集數(shù)據(jù)范圍等問題,本文采用M5P模型樹作為元決策樹,在葉節(jié)點(diǎn)上構(gòu)造多元線性回歸模型,有效提高了元決策樹的預(yù)測精度。(2)針對(duì)隨機(jī)森林中存在部分預(yù)測性能較差且多樣性較低的元決策樹,本文提出了一種改進(jìn)的差分進(jìn)化算法,并將其應(yīng)用到隨機(jī)森林元決策樹的選擇性集成中,在所有元決策樹中選擇部分最優(yōu)的元決策樹子集構(gòu)成新的隨機(jī)森林,進(jìn)行加權(quán)計(jì)算得到最終預(yù)測結(jié)果。(3)針對(duì)隨機(jī)森林算法計(jì)算復(fù)雜度較高的問題,分析了隨機(jī)森林算法和差分進(jìn)化算法的并行性,研究了云計(jì)算體系架構(gòu),采用云計(jì)算技術(shù)中的Spark分布式內(nèi)存計(jì)算框架,對(duì)上述預(yù)測算法進(jìn)行并行化改進(jìn),有效提高了算法的執(zhí)行性能。(4)以內(nèi)蒙古某地區(qū)風(fēng)電監(jiān)測數(shù)據(jù)作為實(shí)際算例,將本文方法與現(xiàn)有短期風(fēng)電功率預(yù)測算法和傳統(tǒng)的隨機(jī)森林回歸算法進(jìn)行對(duì)比;同時(shí)在實(shí)驗(yàn)室服務(wù)器上采用Cloudera公司的發(fā)行版CDH5版本搭建云計(jì)算平臺(tái),對(duì)提出的算法進(jìn)行并行化性能測試。實(shí)驗(yàn)結(jié)果表明本文提出的方法具有較高的預(yù)測精度、泛化性能、可解釋性,且具有較好的并行性能。
[Abstract]:With the adjustment of energy structure in China, the installed capacity of wind power is increasing rapidly. Forecasting wind power accurately and timely can provide an important basis for the reasonable dispatch of power grid, reduce the abandonment of wind, and effectively improve the utilization rate of wind power. At the same time, with the improvement of the intelligent level of wind farm, the scale of wind power monitoring data is increasing, which poses a new challenge to the computational performance of traditional wind power prediction model. In recent years, artificial neural network (Ann), support vector machine (SVM) and its improved algorithms based on machine learning theory have been widely used in short-term wind power prediction, and there are many iterative computing scenarios in machine learning algorithms. The Spark distributed memory computing framework in cloud computing technology can efficiently perform iterative data processing and improve the performance of the algorithm. In view of the existing short-term wind power prediction model has some problems such as weak generalization, difficulty in determining the model structure and parameters, poor interpretability, etc., this paper synthesizes stochastic forest regression algorithm, M5P model tree, differential evolution algorithm, selective integration method, etc. A short-term wind power prediction method based on improved stochastic forest regression algorithm is proposed, and the algorithm is parallelized using Spark cloud computing platform. The main research work is as follows: (1) the traditional stochastic forest regression algorithm takes the classification regression tree as the meta-decision tree, aiming at the low prediction accuracy of the classification regression tree. In this paper, we use M5P model tree as meta-decision tree to construct multivariate linear regression model on leaf node. The prediction accuracy of meta-decision tree is improved effectively. (2) an improved differential evolutionary algorithm is proposed to solve the problem of partial poor prediction performance and low diversity of meta-decision trees in random forests. It is applied to the selective ensemble of stochastic forest meta-decision tree, and the partial optimal subset of meta-decision tree is selected among all meta-decision trees to form a new random forest. The final prediction results are obtained by weighted computation. (3) aiming at the high computational complexity of stochastic forest algorithm, the parallelism of stochastic forest algorithm and differential evolution algorithm is analyzed, and the cloud computing architecture is studied. The Spark distributed memory computing framework in cloud computing technology is adopted to improve the performance of the algorithm effectively. (4) the wind power monitoring data in Inner Mongolia is taken as an example. The proposed method is compared with the existing short-term wind power prediction algorithm and the traditional stochastic forest regression algorithm. At the same time, we use the CDH5 version of Cloudera company to build the cloud computing platform on the laboratory server, and test the parallelization performance of the proposed algorithm. The experimental results show that the proposed method has high prediction accuracy, generalization, interpretability and good parallelism.
【學(xué)位授予單位】：華北電力大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TM614;TP3;TP181

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 錢政;裴巖;曹利宵;王婧怡;荊博;;風(fēng)電功率預(yù)測方法綜述[J];高電壓技術(shù);2016年04期

2 張穎超;郭曉杰;葉小嶺;鄧華;;一種短期風(fēng)電功率集成預(yù)測方法[J];電力系統(tǒng)保護(hù)與控制;2016年07期

3 ;2015年中國風(fēng)電裝機(jī)容量統(tǒng)計(jì)[J];風(fēng)能;2016年02期

4 ;2015年全球風(fēng)電裝機(jī)統(tǒng)計(jì)[J];風(fēng)能;2016年02期

5 王保義;王冬陽;張少敏;;基于Spark和IPPSO_LSSVM的短期分布式電力負(fù)荷預(yù)測算法[J];電力自動(dòng)化設(shè)備;2016年01期

6 楊錫運(yùn);關(guān)文淵;劉玉奇;肖運(yùn)啟;;基于粒子群優(yōu)化的核極限學(xué)習(xí)機(jī)模型的風(fēng)電功率區(qū)間預(yù)測方法[J];中國電機(jī)工程學(xué)報(bào);2015年S1期

7 馬斌;張麗艷;;一種基于徑向基神經(jīng)網(wǎng)絡(luò)的短期風(fēng)電功率直接預(yù)測方法[J];電力系統(tǒng)保護(hù)與控制;2015年19期

8 吳瀟雨;和敬涵;張沛;胡駿;;基于灰色投影改進(jìn)隨機(jī)森林算法的電力系統(tǒng)短期負(fù)荷預(yù)測[J];電力系統(tǒng)自動(dòng)化;2015年12期

9 薛禹勝;郁琛;趙俊華;Kang LI;Xueqin LIU;Qiuwei WU;Guangya YANG;;關(guān)于短期及超短期風(fēng)電功率預(yù)測的評(píng)述[J];電力系統(tǒng)自動(dòng)化;2015年06期

10 王德文;孫志偉;;電力用戶側(cè)大數(shù)據(jù)分析與并行負(fù)荷預(yù)測[J];中國電機(jī)工程學(xué)報(bào);2015年03期

相關(guān)碩士學(xué)位論文 前6條

1 楊劍南;基于人工神經(jīng)網(wǎng)絡(luò)的短期風(fēng)功率預(yù)測研究[D];華北電力大學(xué);2015年

2 孫科;基于Spark的機(jī)器學(xué)習(xí)應(yīng)用框架研究與實(shí)現(xiàn)[D];上海交通大學(xué);2015年

3 李霄;基于提升小波和最小二乘支持向量機(jī)的風(fēng)電功率預(yù)測[D];上海交通大學(xué);2015年

4 唐振坤;基于Spark的機(jī)器學(xué)習(xí)平臺(tái)設(shè)計(jì)與實(shí)現(xiàn)[D];廈門大學(xué);2014年

5 李貞貴;隨機(jī)森林改進(jìn)的若干研究[D];廈門大學(xué);2013年

6 蔡林霖;隨機(jī)森林的模型選擇及其并行化方法[D];哈爾濱工業(yè)大學(xué);2013年

，

本文編號(hào)：2296195