【摘要】：隨著我國大規(guī)模交直流互聯(lián)電力系統(tǒng)的逐步形成,電網(wǎng)結(jié)構(gòu)發(fā)生了深刻變化,整個(gè)系統(tǒng)運(yùn)行狀況呈現(xiàn)出新特性。能源中心向負(fù)荷中心輸電距離長(zhǎng),線路負(fù)荷重,區(qū)域間的低頻振蕩問題時(shí)有發(fā)生,嚴(yán)重制約了線路傳輸容量上限,威脅著整個(gè)電力系統(tǒng)的安全穩(wěn)定運(yùn)行。低頻振蕩問題的特征值分析方法依賴于根據(jù)機(jī)理所建立的系統(tǒng)詳細(xì)模型,隨后在其工作點(diǎn)附近對(duì)其進(jìn)行線性化處理,最后根據(jù)其特征值以及特征向量等進(jìn)行分析。目前,大部分商用系統(tǒng)分析軟件均采用此種低頻振蕩分析手段。但是,由于系統(tǒng)元件的地區(qū)差異以及新元件的不斷投入運(yùn)行,商業(yè)分析軟件提供的動(dòng)態(tài)元件模型庫并不能完全滿足實(shí)際工程需求。因此,完善軟件的動(dòng)態(tài)模型庫是實(shí)現(xiàn)低頻振蕩問題分析的基礎(chǔ),具有較高的工程實(shí)際應(yīng)用價(jià)值。伴隨著廣域量測(cè)技術(shù)的發(fā)展,依據(jù)量測(cè)信息實(shí)現(xiàn)低頻振蕩問題的監(jiān)測(cè)、分析和控制的技術(shù)手段發(fā)展迅猛。量測(cè)信息反映著系統(tǒng)的實(shí)際運(yùn)行工況,所以根據(jù)量測(cè)信息利用辨識(shí)技術(shù)建立的系統(tǒng)模型更直接的反映了系統(tǒng)實(shí)際運(yùn)行特性,基于辨識(shí)模型的分析與控制器設(shè)計(jì)方法具有重要實(shí)際意義。但是,傳統(tǒng)的辨識(shí)方法在精度上有待進(jìn)一步提高,并且對(duì)于大量的可控元件,如何從中選取以實(shí)現(xiàn)系統(tǒng)阻尼最大化有待研究。本文針對(duì)PSS/E動(dòng)態(tài)元件庫無法滿足實(shí)際工程需求的問題,對(duì)比了 BPA與PSS/E動(dòng)態(tài)元件庫的差異,在PSS/E中建立了無法對(duì)應(yīng)上BPA的勵(lì)磁系統(tǒng)、原動(dòng)機(jī)調(diào)速器以及電力系統(tǒng)穩(wěn)定器等模型,并通過單機(jī)無窮大系統(tǒng)的仿真驗(yàn)證了建立模型的正確性。針對(duì)區(qū)域間低頻振蕩問題,提出了一種基于多元線性回歸的電力系統(tǒng)狀態(tài)空間模型辨識(shí)方法,通過對(duì)數(shù)學(xué)模型以及電力系統(tǒng)的辨識(shí)對(duì)比,證明所提方法較子空間方法有更高的精度;利用所提辨識(shí)方法,建立了算例系統(tǒng)和某實(shí)際電網(wǎng)的降階狀態(tài)空間模型,基于辨識(shí)所得模型,利用幾何指標(biāo)選取最佳的控制器裝設(shè)位置,并通過相位補(bǔ)償方法設(shè)計(jì)了直流阻尼控制器,算例系統(tǒng)以及實(shí)際系統(tǒng)的仿真結(jié)果表明,本文所提直流阻尼控制器方法能夠選出最佳裝設(shè)位置,顯著提高系統(tǒng)阻尼。
[Abstract]:With the gradual formation of large-scale AC / DC interconnected power systems in China, the power network structure has undergone profound changes, and the operation of the whole system presents new characteristics. The transmission distance from the energy center to the load center is long, the line load is heavy, and the problem of low frequency oscillation between regions occurs from time to time, which seriously restricts the upper limit of transmission capacity of the transmission line and threatens the safe and stable operation of the whole power system. The eigenvalue analysis method of low frequency oscillation problem depends on the detailed model of the system established according to the mechanism, then linearized it near its working point, and finally it is analyzed according to its eigenvalue and eigenvector. At present, most commercial system analysis software uses this low frequency oscillation analysis method. However, because of the regional differences of the system components and the continuous operation of the new components, the dynamic component model base provided by the commercial analysis software can not fully meet the actual engineering needs. Therefore, improving the dynamic model base of software is the foundation of low frequency oscillation problem analysis, and has high practical application value. With the development of wide-area measurement technology, the monitoring, analysis and control of low-frequency oscillation problem based on measurement information develop rapidly. The measurement information reflects the actual operating conditions of the system, so the system model based on the identification technology based on the measurement information reflects the actual operating characteristics of the system more directly. The analysis and controller design method based on identification model has important practical significance. However, the accuracy of the traditional identification method needs to be further improved, and for a large number of controllable elements, how to select them to maximize the damping of the system remains to be studied. In this paper, aiming at the problem that PSS/E dynamic component library can not meet the actual engineering requirements, the difference between BPA and PSS/E dynamic component library is compared, and the excitation system which can not correspond to BPA is established in PSS/E. The model of prime mover governor and power system stabilizer is verified by simulation of single machine infinite bus system. In this paper, an identification method of power system state space model based on multiple linear regression is proposed to solve the problem of inter-region low frequency oscillation. The mathematical model and power system identification are compared. It is proved that the proposed method has higher accuracy than the subspace method. By using the proposed identification method, the reduced-order state space model of an example system and a practical power network is established. Based on the identified model, the optimal location of the controller is selected by using geometric indexes. The DC damping controller is designed by the phase compensation method. The simulation results of the example system and the actual system show that the DC damping controller proposed in this paper can select the best installation position and increase the system damping significantly.
【學(xué)位授予單位】：西南交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TM712

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李杰;王艷;紀(jì)志成;;基于改進(jìn)教與學(xué)優(yōu)化算法的永磁同步電機(jī)參數(shù)辨識(shí)[J];系統(tǒng)仿真學(xué)報(bào);2017年02期

2 陳宇川;蔡澤祥;朱林;蘇海林;金小明;周保榮;張東輝;;PSS/E中雙端直流系統(tǒng)自定義建模與仿真[J];電氣自動(dòng)化;2016年04期

3 張東輝;金小明;周保榮;陳宇川;楊歡歡;蔡澤祥;朱林;;PSS/E的自定義建模及其仿真研究[J];電力系統(tǒng)保護(hù)與控制;2016年05期

4 褚曉杰;高磊;印永華;唐曉博;周成;易俊;張健;卜廣全;;基于頻域子空間辨識(shí)和集結(jié)理論的廣域阻尼控制安裝地點(diǎn)與控制信號(hào)選取[J];中國電機(jī)工程學(xué)報(bào);2015年18期

5 索江鐳;胡志堅(jiān);劉宇凱;張子泳;王尉;;基于多新息耦合最小二乘算法的電力系統(tǒng)狀態(tài)空間辨識(shí)[J];電力自動(dòng)化設(shè)備;2015年07期

6 孫琳;李天然;吳華仁;;基于模糊自適應(yīng)PID的廣域阻尼控制[J];電力系統(tǒng)及其自動(dòng)化學(xué)報(bào);2015年05期

7 楊貴軍;孟杰;王雙喜;;基于赤池信息準(zhǔn)則的分類回歸決策樹剪枝算法[J];計(jì)算機(jī)應(yīng)用;2014年S2期

8 翁華;徐政;劉f;趙兵;董桓鋒;許烽;;基于多工況綜合留數(shù)及H_2/H_∞的多回直流系統(tǒng)阻尼控制魯棒設(shè)計(jì)方法[J];電力自動(dòng)化設(shè)備;2014年07期

9 李志強(qiáng);湯涌;何鳳軍;羅煒;;基于時(shí)頻變換的同步發(fā)電機(jī)參數(shù)辨識(shí)方法[J];中國電機(jī)工程學(xué)報(bào);2014年19期

10 劉志雄;孫元章;黎雄;宋斌;劉振盛;方鳳美;;廣域電力系統(tǒng)穩(wěn)定器阻尼控制系統(tǒng)綜述及工程應(yīng)用展望[J];電力系統(tǒng)自動(dòng)化;2014年09期

相關(guān)博士學(xué)位論文 前3條

1 白菲菲;基于廣域量測(cè)的電力系統(tǒng)動(dòng)態(tài)響應(yīng)建模及區(qū)域間振蕩阻尼控制[D];西南交通大學(xué);2016年

2 張鵬;基于廣域量測(cè)的電力系統(tǒng)區(qū)域間低頻振蕩分析[D];西南交通大學(xué);2015年

3 鄭偉;光伏并網(wǎng)逆變器辨識(shí)建模方法及應(yīng)用[D];重慶大學(xué);2014年

相關(guān)碩士學(xué)位論文 前1條

1 周密;南方電網(wǎng)低頻振蕩仿真研究[D];華北電力大學(xué)（北京）;2008年

，

本文編號(hào)：2306815