资源描述：
《基于声发射信号的+小波分析+支撑向量机的刀具磨损检测方法》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。
1、Proceedingsofthe8thWorldCongressonIntelligentControlandAutomationJuly6-92010,Jinan,ChinaCuttingToolWearIdentificationbasedonWaveletPackageandSVMXuTaoandWangTaoDepartmentofAutomationShenyangInstituteofAeronauticalEngineeringShenyang,110136,P.R.Chinawyhxt200
2、0@163.comAbstract-Bycontrastwithconventionalmethods,Acousticachievegoodperformance.So,waveletpackagewasstudiedEmission(AE)sensorpossessesbetterperformancefortooltoextractfeatureofcuttingtool[7-9].AndBPneuralwearidentifying.So,AEsensorisemployedintocuttingt
3、oolnetworkwasdesignedtoidentifycuttingtool’sfailure[9].wearidentificationinthispaper.BecauseofthediversityandBecauseSVMpossessesbetterclassificationperformancetimevaryingofAE,waveletpackagedecompositionandforsmallsample,SVMisemployedtoidentifycuttingtoolSu
4、pportVectorMachine(SVM)areemployedtoprocessAEwearafterfeatureextractionbywaveletpackage,whichsignal.Waveletpackageissuitableforanalyzingnon-stationaryextractsfeaturefromreconstructiondataofeachnodeaftersignal,andSVMpossessesexcellentclassificationcapacityf
5、orsmallsample.Accordingtothesefeatures,signalprocessingdecomposition.methodforidentifyingfaultofcuttingtoolwearbasedonThispaperwillbestructuredasfollows;InSectionIIwaveletpackageandSVMwaspresented.Thecharacteristicswewillintroducethetheoremofwaveletpackage
6、approach;ofthecuttingtoolwearunderdifferentconditionswereInSectionIIIwewillintroducethetheoremofSVMextractedbywaveletpackage,andcuttingtoolwearwasapproach;InSection4wewilldescribeourapproach,identifiedbySVMclassifier.Experimentresultsshowthattheexplaininge
7、achstep;Insection5wewillshowhowmethodbasedonwaveletpackageandSVMissuitableforapplyingthismethodforcuttingtoolwearidentificationinidentifyingcuttingtoolwear,andtherateofsuccessfullyidentifyingis93.3%.theexperiment.Finally,wewillpresentmeaningconclusion.Inde
8、xTerms–CuttingToolWear;AESsensor;WaveletPackage;SupportVectorMachineII.WAVELETPACKAGETHEOREMShort-TimeFourierTransform(STFT)possessesI.INTRODUCTIONlinearitycharacteristicwhendecomposingthefrequencyWiththeincr
