基于神經(jīng)網(wǎng)絡(luò)的凝汽設(shè)備故障診斷專家系統(tǒng)研究.pdf

1、Y1211493分類號(hào)：——UDC：——密級(jí)：編號(hào)：工學(xué)碩士學(xué)位論文基于神經(jīng)網(wǎng)絡(luò)的凝汽設(shè)備故障診斷專家系統(tǒng)研究碩士研究生：指導(dǎo)教師：學(xué)位級(jí)別：學(xué)科、專業(yè)：所在單位：論文提交日期：論文答辯日期：學(xué)位授予單位：張秀春夏虹教授工學(xué)碩士核能科學(xué)與工程核科學(xué)與技術(shù)學(xué)院2006年12月2007年1月哈爾濱工程大學(xué)哈爾濱工程大學(xué)碩士學(xué)位論文ABSTRACTItissignificativelythattheCondensationEquipmentw

2、orkssafelyandsteadilyintheNuclearPowerPlantortheRegularPowerPlantHoweverthefaultsoftheCondensationEqmpmenthavepilositynature，uncertaintyanddisguisewhichmakedi伍cultiesforfaultdiagnosisoftheCondensationEquipmentThefaultsig

3、nandthefaultreasonoftheCondensationEquipmenthavecomplexnonlinearmappingrelationsthereforeitcannotnsesimplemodeltodescribeThishasbroughtthepuzzlefortheclassicfaultdiagnosistheoryTheintelligentdiagnosismethodbasedonArtific

4、ialNeuralNetwork(ANN)andExpertSystem(ES)cansolveaboveproblemsgreatlyandthediagnosisresultsaremoreaccurateandreliableTheproblemwhichhowtOrealizefaultdiagnosisoftheCondensationEquipmentwithintelligentdiagnosishasbeenresear

5、chedinthispaperAllkingsofdiagnosistechnologyandmethodshavebeenanalyzedinthispaperThemodeloffaultdiagnosisESbasedonANNwhichiscalledANNESforshorthasbeenestablishedThissystemismadeupbymemorysystemofknowledge，learningsystems

6、，reasoningmachine，explainingdeviceandhumaneompmerinteractioninterfaceDetailedstructureanddesignmethodofeveryparthavebeanprovidedinthispaperANNEShasmergedtheadvantagethatformidableadaptivelearningabilityofANNandexplicitkn

7、owledgeexpressionofES，simplifiedtheprogressoflearningandgainingdataofANNandbuildinginferenceruleANNisimportantpartofANNESanditisthecriterionthatANNESisgoodorbadThispaperhasapprovedthatadaptivestudyratehasmoreadvantagesth