1、東南大學(xué)碩士學(xué)位論文汽輪機(jī)數(shù)字電液控制系統(tǒng)故障診斷的研究姓名：許燕申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：動(dòng)力機(jī)械及工程指導(dǎo)教師：楊建明20030901————一壅噩盔堂亟主堂焦堡_文AbstractWiththerapiddevelopmentofcomputetechnology，thepowerandfrequencyadjustingofmodernsteamturbineunitarefulfilledbydigitelectrohydrau

2、lic(DEH)controlsystemIntheDEHsystem，thefunctionsofadjusting，protectioninspectionandsequentialcontrolarebuiitin，whichimprovethequalityofadjustingrealizethecomplexcontrolalgorithmandstrengthenthesystemsecurity，egtheover—sp

3、eedprotectionHoweverontheonehand，thesystemfaults，suchasthefaultofelectrohydranlicservo—valveinducedbyoilsystem，cannotbeavoidedOntheotherhandwhenfanltsoccurs。operatorsandmaintainmencannotestimatethefaultandsolveitintimesi

4、ncetheyhave1ittleknowledgeoftechnologyandareshortofessentialequipmentsThereforeitisnecessarytoinvestigatethefaultdiagnosisofDEHcontrolsystemandthisworkalsohaveveryhighpracticalvalueofengineeringThecontentsoffaultdiagnosi

5、sincludebothofthefollowingtwosides：staticanalysisanddynamicdiagnosisThefirstonegivestheideaonhowtodealwiththefaultsbydoingcompleteanalysisofpotentialsystemfaultsinadvanceTheotherismainlywiththeviewofon一1inediagnosisandsh

6、owshowtodetectthefaultsBothoftwosidessupplementeachotherandmakesurethesystemfaultscanbefoundandsolvedintimeInthisthesis，afullstaticanalysisofDEHcontrolsystemisgiven。usingFaultTreeAnalysis(FTA)andFailure^lodeEffectsandCri

7、ticalityAnalysis(F臟CA)Basedoiltheresultsofanalysis，thesisalsogivesthesuggestionfromthepointofoperationandmaintenanceonoilsystemArtificialneuralnetwork(ANN)isverysuitableforapplicationofnonlineardynamicsystemSothethesiswi

8、lllookforaconvenientandreliablemethodforthedynamicfaultdiagnosis，InfactthisisadynamicsystemidentificationproblembasedontheneuralnetworkThethesisalsowantstogiveadirectionondiagnosismethodforthesubsequentsystemextensionInt

9、hethesistherearedetailexplanationsforthefaultdiagnosisofDEHThethesisconsistsofthreepartsThefirstpartmakessummarystatementaboutbackgroundandinvestigationstatusonrelatedresearchfieldThesecondpartismadeoftheChapterTwoandCha

10、pterThreewhichgivesthestaticfaultsanalysisofDEHsystemincludingthediscussionofoilsystemInthethirdpart，thereareelaborateillustrationsofhowtouseneuralnetworkindynamicfaultsdiagnosisThemathematicalmodelandneuralnetworkidenti

11、ficationmodelofpositionservesystemarebuiItAtlast，takingtheelectricalhydraulicservovalveasexample，thepapergivesdiagnosisresultsofthreetypicalfaultsKeywords：steamTurbine，DEH。FaultDiagnosis，F(xiàn)rA，F(xiàn)MECA，NeuralNetworkServo—valv