本文選題：電力系統(tǒng) + 不良數(shù)據(jù)檢測辨識��； 參考：《東北石油大學》2017年碩士論文

【摘要】：當前電力行業(yè)發(fā)展迅速,電力系統(tǒng)的規(guī)模也在不斷擴大,隨著電力網(wǎng)絡的結構和運行模式更加復雜化,電網(wǎng)數(shù)據(jù)傳輸中產(chǎn)生錯誤數(shù)據(jù)的概率越來越高,由于客觀原因的存在而產(chǎn)生的的錯誤數(shù)據(jù)又被稱為不良數(shù)據(jù),由于不良數(shù)據(jù)無法正確顯示系統(tǒng)的運行狀況,其存在將嚴重影響狀態(tài)估計的準確性與可靠性,進而對電力系統(tǒng)的安全穩(wěn)定運行產(chǎn)生不利的影響。電力系統(tǒng)不良數(shù)據(jù)檢測辨識的目的在于去除量測數(shù)據(jù)中的不良數(shù)據(jù),為電力系統(tǒng)狀態(tài)估計提供準確的數(shù)據(jù)。目前使用的不良數(shù)據(jù)檢測辨識的方法主要是基于量測數(shù)據(jù)殘差的方法,隨著電網(wǎng)結構的復雜化,該方法檢測結果存在漏檢與誤檢的弊端日益凸顯。本文采用基于模糊聚類的方法實現(xiàn)將原始量測中的良性數(shù)據(jù)與不良數(shù)據(jù)分離,算例仿真結果顯示了該方法相比較傳統(tǒng)方法的優(yōu)越性。本文首先介紹了電力系統(tǒng)不良數(shù)據(jù)檢測辨識的知識背景和研究現(xiàn)狀,對比分析了當前不良數(shù)據(jù)檢測辨識方法的優(yōu)缺點。針對不良數(shù)據(jù)間存在相關性而容易出現(xiàn)殘差污染和殘差淹沒的情況,傳統(tǒng)的殘差檢測方法表現(xiàn)不佳,于是本文提出使用EGSA-FCM算法實現(xiàn)不良數(shù)據(jù)的檢測辨識,該方法以模糊聚類算法中的模糊C均值算法為基礎,通過引入本文提出的增強型萬有引力搜索算法實現(xiàn)對SCADA系統(tǒng)上傳的量測數(shù)據(jù)進行前期搜索,該方法提高了計算效率和準確性。最后將用于聚類有效性判斷的COS指標應用于對最佳聚類數(shù)目的判定,獲得最佳聚類結果,通過已知良性數(shù)據(jù)所在聚類,最終得到量測數(shù)據(jù)中良性數(shù)據(jù)與不良數(shù)據(jù)的分類。針對基于EGSA-FCM算法的不良數(shù)據(jù)檢測辨識方法,本文編制了檢測辨識程序。將該方法應用IEEE14節(jié)點電力系統(tǒng)和某地區(qū)電網(wǎng)變區(qū)中,檢測辨識結果表明本文所提出的方法與傳統(tǒng)的檢測辨識方法相比有效避免了誤檢和漏檢的發(fā)生,檢測結果更加準確。
[Abstract]:With the rapid development of power industry and the expansion of power system scale, with the complexity of power network structure and operation mode, the probability of generating wrong data in power network data transmission is higher and higher.The wrong data caused by the existence of objective reasons is also called bad data. Because the bad data can not correctly display the operating status of the system, its existence will seriously affect the accuracy and reliability of the state estimation.Furthermore, it has an adverse effect on the safe and stable operation of power system.The purpose of power system bad data detection and identification is to remove the bad data from the measurement data and to provide accurate data for power system state estimation.At present, the method of detection and identification of bad data is mainly based on the residual of measured data. With the complexity of power network structure, the defects of the detection results of this method are increasingly prominent.In this paper, the method based on fuzzy clustering is used to separate the benign data from the bad data in the original measurement. The simulation results show that this method is superior to the traditional method.This paper first introduces the knowledge background and research status of power system bad data detection and identification, and compares and analyzes the advantages and disadvantages of current bad data detection and identification methods.Because of the correlation between bad data and the possibility of residual pollution and residual inundation, the traditional residual detection method is not good, so this paper proposes to use EGSA-FCM algorithm to detect and identify bad data.This method is based on the fuzzy C-means algorithm of fuzzy clustering algorithm. By introducing the enhanced universal gravity search algorithm proposed in this paper, the pre-search of the measurement data uploaded by SCADA system is realized.This method improves the calculation efficiency and accuracy.Finally, the COS index, which is used to judge the clustering validity, is applied to the determination of the best clustering number, and the best clustering result is obtained. Finally, the classification of the benign data and the bad data in the measured data is obtained through the known clustering of the benign data.Aiming at the bad data detection and identification method based on EGSA-FCM algorithm, a detection and identification program is developed in this paper.The method is applied to the IEEE14 node power system and a region power system. The detection and identification results show that the method proposed in this paper is more effective than the traditional detection and identification methods to avoid the occurrence of false detection and miss detection, and the detection results are more accurate.
【學位授予單位】：東北石油大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TM732

【參考文獻】

相關期刊論文 前10條

1 馬亞南;;電力系統(tǒng)異常數(shù)據(jù)的檢測辨識與修正探究[J];中國電業(yè)(技術版);2014年04期

2 李超順;周建中;肖劍;;基于改進引力搜索算法的勵磁控制PID參數(shù)優(yōu)化[J];華中科技大學學報(自然科學版);2012年10期

3 李沛;段海濱;;基于改進萬有引力搜索算法的無人機航路規(guī)劃[J];中國科學:技術科學;2012年10期

4 劉勇;馬良;;非線性極大極小問題的混沌萬有引力搜索算法求解[J];計算機應用研究;2012年01期

5 谷文祥;郭麗萍;殷明浩;;模糊c-均值算法和萬有引力算法求解模糊聚類問題[J];智能系統(tǒng)學報;2011年06期

6 蔣德瓏;王克文;王祥東;;基于模糊等價矩陣聚類分析的不良數(shù)據(jù)辨識[J];電力系統(tǒng)保護與控制;2011年21期

7 唐俊;;電力系統(tǒng)狀態(tài)估計研究綜述[J];中國新技術新產(chǎn)品;2011年20期

8 李子豐;;用物體與微粒子的動量交換解釋萬有引力定律[J];中國西部科技;2011年28期

9 申元霞;王國胤;曾傳華;;PSO模型種群多樣性與學習參數(shù)的關系研究[J];電子學報;2011年06期

10 金林鵬;李均利;魏平;陳剛;;用于函數(shù)優(yōu)化的最大引力優(yōu)化算法[J];模式識別與人工智能;2010年05期

相關碩士學位論文 前3條

1 周小寶;電力系統(tǒng)狀態(tài)估計不良數(shù)據(jù)檢測與辨識方法研究與應用[D];湖南大學;2013年

2 鄒賢;基于GSA的昆明電網(wǎng)不良數(shù)據(jù)辨識與修正研究[D];華北電力大學;2012年

3 蔣德瓏;采用模糊聚類方法進行電力系統(tǒng)不良數(shù)據(jù)辨識[D];鄭州大學;2011年

，

本文編號：1758483