發(fā)布時間：2018-01-15 20:19

  本文關(guān)鍵詞：基于負荷特性分析的中長期負荷預(yù)測研究　出處：《湖南大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 中長期負荷特性分析 主導(dǎo)因素辨識 偏最小二乘法 相關(guān)性分析 系統(tǒng)開發(fā)

【摘要】：中長期負荷預(yù)測的結(jié)果是電力系統(tǒng)規(guī)劃的基礎(chǔ)和依據(jù)，是開展電量綜合平衡、電力結(jié)構(gòu)優(yōu)化調(diào)整等工作的前提。開展中長期負荷預(yù)測有利于資源的優(yōu)化配置、燃料計劃的制定和電網(wǎng)運行方式的安排，是保證電網(wǎng)運行可靠性和合理性的重要前提，對電網(wǎng)規(guī)劃部門有著重要的理論意義和實用價值。 負荷特性分析作為電力系統(tǒng)中的一項重要工作，有助于預(yù)測人員準確地把握負荷曲線的變化規(guī)律，從而獲得較為理想的預(yù)測結(jié)果。本文根據(jù)某省電網(wǎng)歷史負荷數(shù)據(jù)和相關(guān)資料，在對地區(qū)電網(wǎng)中長期負荷特性指標統(tǒng)計和分析的基礎(chǔ)上，深入探究了各特性指標與影響因素間的關(guān)系。之后，對負荷特性指標的特點和規(guī)律加以總結(jié)，為開展中長期負荷預(yù)測的研究工作做準備。 相比于短期負荷，中長期負荷的影響因素繁雜、時間跨度長、不確定因素多。本文基于中長期負荷特性的分析結(jié)果，提出了電力負荷的主導(dǎo)因素辨識法，，并結(jié)合偏最小二乘回歸法應(yīng)用于年電量預(yù)測。該方法從負荷特性指標的角度出發(fā)，根據(jù)負荷分析的結(jié)果選出基準負荷指標，并結(jié)合Pearson相關(guān)分析法判別影響因素與基準負荷指標的關(guān)聯(lián)程度，進而選出電力負荷的主導(dǎo)因素。通過算例分析，證實了此方法可大大減小預(yù)測模型中由于工作人員主觀經(jīng)驗而帶來的偏差，弱化了噪聲因素的干擾，模型的實用性得到加強。同時，由于無關(guān)因素的剔除，模型對主成分的提取能力得到加強，預(yù)測結(jié)果更為精確。 本文從電力企業(yè)長遠發(fā)展和工作需要角度出發(fā)，開發(fā)了電力需求預(yù)測及負荷特性分析平臺。系統(tǒng)基于Java平臺和Oracle數(shù)據(jù)庫，采用了多層體系的B/S（Browser/Server）結(jié)構(gòu)，以電力需求預(yù)測和數(shù)據(jù)挖掘分析為核心，緊密結(jié)合計算機網(wǎng)絡(luò)技術(shù)、通信技術(shù)、信息安全技術(shù)與智能技術(shù)。系統(tǒng)充分考慮了電力負荷的主要影響因素，密切聯(lián)系了電網(wǎng)工作的實際需求，有效地提高了預(yù)測結(jié)果的準確性，為電網(wǎng)規(guī)劃提供了重大的技術(shù)支撐。
[Abstract]:The result of medium and long term load forecasting is the foundation and basis of power system planning, and the premise of carrying out the work of comprehensive balance of electricity quantity and optimization and adjustment of power structure, etc. The development of medium and long term load forecasting is beneficial to the optimal allocation of resources. The formulation of fuel plan and the arrangement of power grid operation mode are the important premises to ensure the reliability and rationality of power grid operation, and have important theoretical significance and practical value to the power network planning department. As an important work in power system, load characteristic analysis is helpful for forecasters to accurately grasp the changing law of load curve. Based on the historical load data and related data of a province, the paper makes statistics and analysis of the medium and long-term load characteristics of regional power network. This paper probes into the relationship between the characteristic indexes and the influencing factors, and then summarizes the characteristics and rules of the load characteristic indexes in order to prepare for the research work of the medium- and long-term load forecasting. Compared with the short-term load, the medium and long term load has a complex influence factor, long time span and many uncertain factors. Based on the analysis results of the medium and long term load characteristics, the paper puts forward the main factor identification method of the power load. Combined with the partial least square regression method, the method is applied to the annual electricity forecasting. This method selects the benchmark load index according to the results of the load analysis from the point of view of the load characteristic index. Combined with the Pearson correlation analysis, the correlation degree between the influencing factors and the benchmark load index is determined, and then the dominant factors of the power load are selected. It is proved that this method can greatly reduce the deviation caused by the staff's subjective experience in the prediction model, weaken the interference of noise factors, and enhance the practicability of the model. At the same time, the irrelevant factors are eliminated. The ability of the model to extract principal components is enhanced, and the prediction results are more accurate. In this paper, the power demand forecasting and load characteristic analysis platform based on Java platform and Oracle database is developed from the perspective of the long-term development and work needs of electric power enterprises. A multilayer B / S browser / Server structure is adopted, which is based on power demand prediction and data mining analysis, and combines closely with computer network technology and communication technology. Information security technology and intelligent technology. The system fully considered the main factors of power load, closely linked to the actual needs of the grid work, and effectively improved the accuracy of the forecast results. It provides important technical support for power network planning.
【學(xué)位授予單位】：湖南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TM715

【參考文獻】

相關(guān)期刊論文 前10條

1 陳偉;周峰;韓新陽;單葆國;;國家電網(wǎng)負荷特性分析研究[J];電力技術(shù)經(jīng)濟;2008年04期

2 崔凱;李敬如;趙彪;劉海波;;城市飽和負荷及其預(yù)測方法研究[J];電力技術(shù)經(jīng)濟;2008年06期

3 季澤宇;袁越;鄒文仲;;改進偏最小二乘回歸在電力負荷預(yù)測中的應(yīng)用[J];電力需求側(cè)管理;2011年01期

4 程義明;羅滇生;蔡劍彪;許甜田;郭精人;;湘潭地區(qū)負荷特性分析[J];電力需求側(cè)管理;2011年06期

5 陳章潮,熊崗;應(yīng)用改進的灰色GM(1,1)模型進行長期電力需求預(yù)測[J];電力系統(tǒng)自動化;1993年07期

6 韋凌云,吳捷,劉永強;基于系統(tǒng)動力學(xué)的電力系統(tǒng)中長期負荷預(yù)測[J];電力系統(tǒng)自動化;2000年16期

7 賀蓉,曾剛,姚建剛,青志文,沈新祥,劉明清;天氣敏感型神經(jīng)網(wǎng)絡(luò)在地區(qū)電網(wǎng)短期負荷預(yù)測中的應(yīng)用[J];電力系統(tǒng)自動化;2001年17期

8 朱六璋;;短期負荷預(yù)測的組合數(shù)據(jù)挖掘算法[J];電力系統(tǒng)自動化;2006年14期

9 肖峻;張晶;朱濤;史常凱;張海平;;基于關(guān)聯(lián)分析的城市用電負荷研究[J];電力系統(tǒng)自動化;2007年17期

10 李智勇;陳志剛;徐政;麻敏華;;中國全社會用電量增長主導(dǎo)因素辨識[J];電力系統(tǒng)自動化;2010年23期

本文編號：1429903