汽轮发电机组振动故障诊断系统的开发

上海交通大学硕士学位论文汽轮发电机组振动故障诊断系统的开发姓名：卫振华申请学位级别：硕士专业：动力机械及工程指导教师：忻建华20050101iii125MW300MW133002iv34AbstractvTHEDEVELOPMENTOFTHEVIBRATIONFAULTDIAGNOSISSYSTEMFORTURBOSETABSTRACTVibrationisanimportantcriteriarelatedtotheoperationoftheturboset.Thetechnologyofthetreatment,analysisanddiagnosisforthevibrationdatacameintobeingwiththeapplicationoftheturboset.Withtheriseoftheprocessparametersandcapacity,themorehighreliabilityofthedevicesispursuedandtheeconomiclossesoftheabnormalshutdowncausedbyhighshaftingvibrationarealsoincreased.Theabovereasonspropelledthedevelopmentofthediagnosistechnologymatchingthepractice.Thepaperpresentsthestudyforthefaultanalysisanddiagnosissystemofthetubosetvibrationonthebasisofthesystematicreferenceandsummaryofthedomesticandaboardturbosetfaultdiagnosisdevelopment,relatedon-linediagnosissystemandthecharacteristicsofdatacollectionsystem.Thesystemhasbeenappliedtotheturbosetrated125MWinYangshupuPowerPlantandreated300MWinWujingPowerPlant.Intheoneyear’soperation,thesystemfoundouttheturbosetdiagnosissuchassteamgapvibrationandthepull-outofrotatablepartswhichwashelpfultothesafeoperationofpowerplantThispaperpaysemphasisonthefollowingworks:AbstractviFirstly,thisthesisintroducesthegeneralstructureofthesystemcombinedtheturbosetvibrationdiagnosticwiththermalpartlifemanagement.ThetasksnamelytheconnectionbetweendatacollectorandBently3300,thecommunicationbetweencomputerandthedatacollectorandthedevelopmentofvibrationdatabaseareachievedinthispaper.Secondlyintegratedwiththeformeron-linemodelaccordingtotheoff-linecalculationforcasingandrotorofthesameturbine,thevibrationdiagnosissystemisabletocalculatethethermalstressandlifeconsumptiongiventheconditionssuchasstartup,shutdownandloadvariation.Atthemeantime,thesedataissavedinthevibrationdatabase,usedbyvibrationdiagnosismodelandviewedbytheoperator.Thirdly,thehierarchicalclassificationdiagnosismodelbasedonmembershipgradeandruleisputforwardinthispaper.Itisthefirsttimethatarulecollectionforturbineshaftingvibrationdiagnosisisbuiltupwithintegrationofthermalparameters,operationparametersandvibrationparameters.Thepaperpresentstenmaintypefaultsthroughthesummaryfortheformerinformantaswellasthepracticalexperience.Thehierarchicalclassificationdiagnosismodelbasedonmembershipgradeandruleanditsruleknowledgecollectionarecomposedwiththemembershipfunctionswhichcombinedvibrationsymptomandthermalparametersaswellasfaulttypecorrespondingtoeachvibrationtype.AbstractviiThenthecollectionisimplementedwiththeprogram.Withthereasoningandofthesystem,thefaultsymptomsandfeaturesarefoundoutandthecorrespondingfaultsareidentified.Atlast,thepracticaloperationofthisvibrationdiagnosissystemshowsthatthesystemhasgoodhuman-machinecommunicationabilityanddiagnosisability.ThesystemcaughttwovibrationfaultsduringtheoperationinYashupuPowerPlantandWujingPowerPlant,whichconciliatedthetimefortheeliminationofthefaults.KEYWORDS:VibrationFaultdiagnosissystem,membershipgrade,hierarchicalclassification,dynamicbalancingi05111iiviiimMWHzr/min111[1]1270EPRIbentlyEnterkirdCSIBentlyDM2000MCM2000AID[1]-[8]80,,,1121-11MMMD32MFD23RB204ZJZ15HZ16MMDS90711131125MW992000#51.6mm200010#5#520013#5#54#20034#1#210200313234512125MW1214112330043SQLSERVER7.0/200094FFT5125MWEXCEL541025300MW62121VDA-2000DSPbently3300VDA-2000223DSP200V~240VVD-20110/100M2-1DCSMIS1nTCP/IPFig.2-1Blockdiagramofsystemhardware781024NMM*N=1024N=163264128814A/DA/D()A/D0~20V23P0220VLNPE22VDA-2000Fig.22StructureofVDA-2000vibrationdatacollectionP9P8P7P6P5P4P3P2P1P023Fig.23Connectionport8P18P2~P91622125MWDASBently3300SQLSERVER7.0/20002-49121252433074231520650r/min7()532182.5MW982124Fig.24Structureofdatabase102-1IDintTimeDatetimeSpeedSmallintPowerrealMaxXsmallintXMaxYsmallintYCNumsmallintPos_1varbinaryXPos_2varbinaryYPos_3varbinary222-2IDintTimeDatetimeSpeedSmallintA000RealA001RealA269RealVarbinarybinary+4512204810244binary20484Varbinary21SQLSERVER7.0/2000SQLSERVER2000SQLSERVER7.0Varbinarysql_variantsql_variant80161123VDA-2000MicrosoftSQLServer12FFT125MW1331[12]121234514612345632[9]-[11]1580%130m80%23320%1020200MW2600r/min163179m1441m693220m6047m1186040m6m1236171233[29][30]11823411922012/——3421313222,1,2122340~100m,600~700m1abcd235~451.2~1.6MPa2412342590901234561226341/21/31/21/310123427567891034.12281234/29[12]11nq30qnnq3-1[12]S0S01S02S0nS011S01PS0n1S0nq…………31Fig31TheModelBasedOnHierarchical31321111234513-1433Fig.3-2HierarchicalClassificationTreeofTheflutterfailureofSteamTurbine1,2,1.52,31/2281,,,,(,)0.53-2343-2(1)(2)3-333Fig.3-15OrganizationChartOftheKnowledgeForFaultNode3533123456789√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√√,√√√√√√√√√√√√√*1.0.010.39f12.0.40.49f13.0.5f14.0.51f10.99f15.f16.2f17.35f18.f19.f1f1112234365617110248210249310241041024115102412610242FFT3145a0.5bc123456190%260%380%490%5n…41Weight3_4.dat0.50.837351238123456414142434445464741Real-timeTrackofAxesDataofVibrationReal-timeTrendFig.41Keymonitoringdiagram3942YWaveformXWaveformYSpectrumXSpectrumshaftcenterLocus43Fig.42DiagramofWavefo