恒力铣削智能控制算法研究与仿真.pdf

《恒力铣削智能控制算法研究与仿真.pdf》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、机床与液压Jun．2012HydromechatronicsEngineeringVe1．40No．12DOI：10．3969／j．issn．1001—3881．2012．12．004TheResearchandSimulationofIntelligentControlStrategiesforConstantForceMillingHUPo，HANDedong，LIMaoyue，FUHongyaSchoolofElectro—mechanicalEngineering，HarbinInstituteofTechnology，Harbin150001，

2、ChinaAbstract：Thispaperinvestigatestheapplicationofintelligentcontrolstrategiesinmillingprocesstogetconstantcuringforce．Sincethecuringprocessisunderagoodcondition，thesurfacequalityisimprovedandthehigheraccuracyisobtained．ThemilingprocesswasmodeledbyadevelopedBPneuralnetwork．Intel

3、igentcentreIstrategiesbasedonneuraInetworkandfuzzylogicwereproposedtoguaranteetheconstantcuringforceandimprovethecuttingeficiency．Simulationswereconducted'toverifytheefectivenessoftheproposedstrategiesandtheresultsshowedthatthetimesavingoftwostrategiesare13．9％and13．7％．respectivel

4、y．Keywords：constantforcemilling，intelligentcontrolstrategy，neuralnetwork，fuzzylogicCFRPcompositesof-line『2]．Artifieialneuralnet．1．IntroductionwerkswasusedtopredictcuttingforcesinthecaseofmachiningthehardturningofAISI52100bearingTheefficiencyofmanufacturinghasbeengreatlysteelbyusi

5、ngCBNcuttingtool[3]．NeuralnetworksimprovedbythewideapplicationofCNCmachinewereusedforoff-linefeedrateoptimizationinendtools．However，themajordrawbackisthattheoper—milling『4]．Andinmostneuralnetworkcontrolsys．atingparameterssuchasspindlespeed，feedrate，terns．theneuralnetworkistrained

6、inadvanceandaxialcuttingdepthandradialcuttingdepthcanonlytheweightvaluesbetweendifferentlayersofneuronbechosen，eithermanuallyorautomaticallybeforearefixedwhentheneuralnetworkisworking[1]．thepracticalmachiningprocess，whichcannotbeAlso，inmostfuzzycontrolsystems，thefuzzycontroladjus

7、tedaccordingtoinstantcuttingconditions．Thestrategiesarefixedtoatabletoffuzzyrules[5—6]．cuttingparametersareoftenconservativelyselectedtoThelackofgeneralizationcapabilitymakesthiskindprotectthemachinetoolsandcuttingtools．Asare—ofoptimizationalgorithmsonlyadapttosomecertainsuh，many

8、CNCsystemsrunundertheinefficient0p．condi