资源描述：
《外文翻译---旋转机械轴心轨迹识别方法研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、附录ResearchofidentificationofshaftorbitforrotatingmachineXIAOSheng-guang（TestCenterofChongqingUniversity，Chongqing，China）Abstract：Anovelapproachfortheidentificationofshaftorbitispresented.Thevibrationdisplacementsignalsacquiredintwomutuallyverticaldirections
2、weretreatedthroughnoisesuppressionandfittedtoformashaftorbit.Thenthedirectionchangingcharacterwasextractedandallshaftorbitswereclassifiedandidentifiedwiththefunctiondiscriminatedmethodaccordingtothepatternrecognitiontheory.Eachtypeofshaftorbitwasdescribedin
3、detailwithonecharacter，whichcanhelptojudgetheoperationstatusofthemechanicalandtheextentofthefailure.Theanalysisandsimulationgotgoodresults.Keywords：Shaftorbits；Faultdiagnosis；Geometricfeatures；Patternrecognition；Thinningclassification1TheintroductionWiththe
4、developmentofscienceandtechnologyandmodernindustry,torotatingmachinesthelarge-scale,high-speedandautomationdirection,theshapeofrotatingmachinerystatemonitoringandfaultdiagnosisisputforwardhigherrequest,theaxistrajectoryforrotatingmachineryisanimportantstate
5、characteristicparameters,canbesimpleandstraightview,vividlyreflecttherunningstatusofequipment.Throughtotheaxisoftrackobservation,candeterminesomeofthecommonfaults,suchasoilfilmvortexStill,oilfilmoscillation,shaftnotmedium.Thetraditionalaxislocusandshapethed
6、ynamiccharacteristicsidentificationisbasedontheman-machinedialoguemode,seriousaffectthelevelofintelligentfaultdiagnosis.Inordertoimprovethedegreeofintelligentfaultdiagnosis,itisnecessarytoin-depthstudythetrajectoryoftheaxisofrotatingmachineryautomaticidenti
7、ficationtechnology.Axispathatpresent,alreadyhaveseveralidentificationmethods,including[1-2]invariantmomentmethod,atwo-dimensionalimagegraylevelmatrix[3].literature[1-2]axispathwithsevenmomentinvariantsasfeaturevectors,recognitionbythedistancebetweenthechara
8、cteristicsofaxialtrajectoryshape,literature[3]theaxistrajectoryimagecoding,usingneuralnetworkforidentification.Bothmethodscanbetteridentifyaxispath,butthemethodiscomplex,relativelylargeamountofcalculat
