基于粗糙集理論與證據(jù)理論相結(jié)合的電網(wǎng)故障診斷.pdf

1、華北電力大學(xué)（保定）碩士學(xué)位論文基于粗糙集理論與證據(jù)理論相結(jié)合的電網(wǎng)故障診斷姓名：祝亞靜申請學(xué)位級別：碩士專業(yè)：電力系統(tǒng)及其自動化指導(dǎo)教師：顧雪平20081028華北電力人學(xué)碩十學(xué)位論文摘要摘要基于粗糙集理論和證據(jù)理論處理不確定和不精確問題方面的優(yōu)點，本文提出了一種粗糙集理論和證據(jù)理論相結(jié)合的電網(wǎng)故障診斷方法。首先建立考慮各種故障情況的故障診斷決策表，采用基于可辨識矩陣屬性頻率的算法對決策表進(jìn)行約簡，得到有效數(shù)據(jù)的最簡組合，然后利用粗糙

2、集理論與證據(jù)理論的關(guān)系把有效數(shù)據(jù)轉(zhuǎn)化為證據(jù)理論的基本概率分配，最后把粗糙集理論中屬性重要度的思想引入到電網(wǎng)故障診斷中，根據(jù)證據(jù)理論組合規(guī)則分兩步對證據(jù)進(jìn)行合成，實現(xiàn)對故障的診斷。通過將粗糙集理論與證據(jù)理論相結(jié)合，有效地提高了故障診斷算法的容錯能力，以實際運行的變電所為例驗證了此算法。關(guān)鍵詞：粗糙集理論，證據(jù)理論，故障診斷，電網(wǎng)ABSTRACTBasedontheadVantageofroughsetmeo叮andeVidencethe0

3、巧toprocessuIlceftain鋤dinaccurateproblem，thisthesispresentsanewideaofdistnbutionnetworkfaultdiagnosisbasedonroughsettheo巧鋤deVidencetheoⅨFirst，afaultdia印osisdecisiontableconsideringmanyfaultconditonsisbuilt，andanalgorithmb

4、asedonthe行equencyofdiscemingmatrixattributeisusedtoreducethedecisiontabletoobtainthesimplestef!IbctivedatecombinationThen，usingtherclationbetweenrouglIsettlle0巧觚deVidencetheory，tlleeffectiVedataistransIatcdtobasicprobabi

5、lityassigmnentLastly，theideaofattributesignificanceinrougllsetm∞D(zhuǎn)risin仃oducodt0faultdiagnosisInresponsetotheRuleofCombination，nleCombinationofeVidenceispmcessedbytwostepst0realizethefaultdiagnosisT11rou曲thecombinationoft

6、hese鉚ometllods，me鋤rtol鋤ceofmeproposedfaultdia鯽10sisalgorithmiseffectiVelyimpmVedThisalgorithmhasbeenverifiedbytheapplicationona66kVsubstationZhuYajing(ElectricPowerSyst鋤孤dits—雌omation)DirectedbyPl0￡GuXuepingKEYWoRDS：roug