本文選題：電力系統(tǒng) + 優(yōu)化規(guī)劃�。� 參考：《湖南大學(xué)》2014年碩士論文

【摘要】：科學(xué)合理的電力系統(tǒng)規(guī)劃是電力系統(tǒng)安全、可靠、經(jīng)濟(jì)運(yùn)行的前提。分布式發(fā)電能給用戶帶來便捷、環(huán)保的能源，將分布式電源與主網(wǎng)供電結(jié)合是智能電網(wǎng)的發(fā)展方向。分布式電源的接入將給傳統(tǒng)的配電網(wǎng)規(guī)劃和運(yùn)行帶來深刻的變革。在這樣的背景下，本文從含分布式電源系統(tǒng)負(fù)荷預(yù)測(cè)方法及配電網(wǎng)擴(kuò)展規(guī)劃兩個(gè)方面進(jìn)行了研究。 電力系統(tǒng)負(fù)荷預(yù)測(cè)是電網(wǎng)規(guī)劃的前提和基礎(chǔ)，由于分布式電源的安裝容量和位置具有隨機(jī)性和不確定性，其出力情況也將受到氣候等諸多因素的影響，這些都將加大含分布式電源系統(tǒng)負(fù)荷預(yù)測(cè)的難度。本文在分析傳統(tǒng)負(fù)荷預(yù)測(cè)方法的基礎(chǔ)上，總結(jié)了含分布式電源系統(tǒng)負(fù)荷預(yù)測(cè)研究思路。支持向量機(jī)技術(shù)具有學(xué)習(xí)能力強(qiáng)，，能處理小樣本，并且具有良好的精度，但選擇合適的參數(shù)具有一定的難度，為此引入自適應(yīng)粒子群優(yōu)化算法對(duì)支持向量機(jī)的參數(shù)進(jìn)行選擇，并在算法中加入了極值擾動(dòng)策略，防止其陷入局部最優(yōu)。建立了基于自適應(yīng)粒子群算法和支持向量機(jī)的含分布式電源系統(tǒng)負(fù)荷預(yù)測(cè)模型。算例分析結(jié)果表明，本文提出的方法與傳統(tǒng)的SVM法相比具有更好的預(yù)測(cè)精度。 從分布式電源并網(wǎng)對(duì)配電網(wǎng)運(yùn)行和規(guī)劃的影響出發(fā)，基于雙層規(guī)劃的方法建立了含分布式電源配電網(wǎng)擴(kuò)展規(guī)劃的模型，上層規(guī)劃為電網(wǎng)電源規(guī)劃，下層規(guī)劃為配電網(wǎng)網(wǎng)架優(yōu)化。針對(duì)傳統(tǒng)規(guī)劃中沒有考慮對(duì)分布式電源的種類和運(yùn)行時(shí)間進(jìn)行選擇的問題，提出了一種基于年持續(xù)負(fù)荷曲線和電源成本特性的方法來選擇分布式電源的種類和投入工作的時(shí)間。在綜合考慮分布式電源接入對(duì)網(wǎng)絡(luò)損耗和電壓質(zhì)量影響的基礎(chǔ)上對(duì)分布式電源的候選位置進(jìn)行選擇。本文采用改進(jìn)的遺傳算法對(duì)配電網(wǎng)規(guī)劃模型進(jìn)行優(yōu)化求解，運(yùn)用簡(jiǎn)化的二進(jìn)制編碼方式對(duì)染色體進(jìn)行編碼，并對(duì)傳統(tǒng)的前推回代潮流計(jì)算方法做了基于層次關(guān)聯(lián)矩陣的改進(jìn)。利用圖論的知識(shí)對(duì)配電網(wǎng)規(guī)劃中可能出現(xiàn)的不可行解問題進(jìn)行了修復(fù)。最后，通過對(duì)修改后的IEEE33節(jié)點(diǎn)系統(tǒng)進(jìn)行仿真分析，證明了規(guī)劃模型及其算法改進(jìn)的合理性和有效性。
[Abstract]:Scientific and reasonable power system planning is the premise of safe, reliable and economical operation of power system. Distributed generation can bring users convenient and environmentally friendly energy. It is the development direction of smart grid to combine distributed generation with main network power supply. The access of distributed generation will bring profound changes to the traditional distribution network planning and operation. In this context, the load forecasting method and the distribution network expansion planning are studied in this paper. Load forecasting of power system is the premise and foundation of power network planning. Due to the randomness and uncertainty of installation capacity and location of distributed power generation, its output will also be affected by climate and many other factors. All these will increase the difficulty of load forecasting with distributed power system. Based on the analysis of traditional load forecasting methods, this paper summarizes the research ideas of load forecasting in distributed power systems. Support vector machine (SVM) technology has a strong learning ability, can process small samples, and has good precision, but it is difficult to select suitable parameters. Therefore, an adaptive particle swarm optimization algorithm is introduced to select the parameters of support vector machine. The extremum perturbation strategy is added to the algorithm to prevent it from falling into local optimum. A load forecasting model based on adaptive particle swarm optimization (APSO) and support vector machine (SVM) is proposed. The numerical results show that the proposed method has better prediction accuracy than the traditional SVM method. Starting from the influence of distributed power grid connection on distribution network operation and planning, a model of distribution network expansion planning with distributed power generation is established based on bilevel programming method. The upper planning is power planning and the lower planning is the optimization of distribution network frame. In order to solve the problem of not considering the choice of the type and operation time of distributed generation in traditional planning, a method based on the annual continuous load curve and the cost characteristics of power supply is proposed to select the type of distributed power and the time when it is put into operation. On the basis of considering the influence of distributed power access on network loss and voltage quality, the candidate positions of distributed power supply are selected. In this paper, an improved genetic algorithm is used to optimize the distribution network planning model, a simplified binary coding method is used to encode chromosomes, and an improvement based on hierarchical correlation matrix is made for the traditional forward generation power flow calculation method. The infeasible solution problem in distribution network planning is repaired by using the knowledge of graph theory. Finally, through the simulation analysis of the modified IEEE33 node system, the rationality and effectiveness of the improved programming model and its algorithm are proved.
【學(xué)位授予單位】：湖南大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TM715

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 劉琳;陶順;肖湘寧;李英毓;;分布式發(fā)電及其對(duì)配電網(wǎng)網(wǎng)損的影響分析[J];電工電能新技術(shù);2012年03期

2 章文俊;程浩忠;王一;歐陽武;;基于樹形結(jié)構(gòu)編碼單親遺傳算法的配電網(wǎng)優(yōu)化規(guī)劃[J];電工技術(shù)學(xué)報(bào);2009年05期

3 王小波;劉德強(qiáng);;基于人工神經(jīng)網(wǎng)絡(luò)的短期負(fù)荷預(yù)測(cè)的研究[J];電力學(xué)報(bào);2011年04期

4 王守相;王慧;蔡聲霞;;分布式發(fā)電優(yōu)化配置研究綜述[J];電力系統(tǒng)自動(dòng)化;2009年18期

5 王敏,丁明;含分布式電源的配電系統(tǒng)規(guī)劃[J];電力系統(tǒng)及其自動(dòng)化學(xué)報(bào);2004年06期

6 章文俊;程浩忠;程正敏;姚茵;谷慶利;;配電網(wǎng)優(yōu)化規(guī)劃研究綜述[J];電力系統(tǒng)及其自動(dòng)化學(xué)報(bào);2008年05期

7 蔣燕;王少楊;封蕓;;基于遞歸等權(quán)組合模型的中長(zhǎng)期電力負(fù)荷預(yù)測(cè)[J];電力系統(tǒng)及其自動(dòng)化學(xué)報(bào);2012年01期

8 陳民鈾;朱博;徐瑞林;徐鑫;;基于混合智能技術(shù)的微電網(wǎng)剩余負(fù)荷超短期預(yù)測(cè)[J];電力自動(dòng)化設(shè)備;2012年05期

9 劉曉飛,彭建春,高效,陳景懷,卜永紅;基于單親遺傳算法的配電網(wǎng)絡(luò)規(guī)劃[J];電網(wǎng)技術(shù);2002年03期

10 劉楊華;吳政球;涂有慶;黃慶云;羅華偉;;分布式發(fā)電及其并網(wǎng)技術(shù)綜述[J];電網(wǎng)技術(shù);2008年15期

相關(guān)博士學(xué)位論文 前2條

1 陸寧;基于群集智能與算法融合的電力負(fù)荷組合預(yù)測(cè)[D];華中科技大學(xué);2010年

2 麻秀范;含分布式電源的配電網(wǎng)規(guī)劃與優(yōu)化運(yùn)行研究[D];華北電力大學(xué);2013年

本文編號(hào)：1924431