基于广义马尔科夫链模型的动柔度预测.pdf

《基于广义马尔科夫链模型的动柔度预测.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Dec．2014HydromechatronicsEngineeringVo1．42No．24DynDOI：10．3969／j．issn．1001—3881．2014．24．013amiccompolianceporedictionbasedong~eneralizedMarkqovchainmodelZhi—gangHUANG，Feng—yunXIESchoolofMechanicalandElectronicalEngineering，EastChinaJiaotongUniversity，Nanchang330013，ChinaAbstract：Dyna

2、miccomplianceplaysanimportantroleinmachiningstabilityandaccuracy．Dy-namiccompliancepredictioncanprovidepracticalguidanceforprecisioncompensationandstabil—ityanalysisofmachinetoo1．Itiswelknownthatuncertaintywillcausetheloweraccuracyandre—liability．ThetraditionaIMarkovchainpredictionmethod

3、cannotconsideruncertaintyproblem．Inthispaper，ageneralizedMarkovchainmodelisproposedtoimprovetheaccuracyandreliabilityofprediction．Dynamiccomplianceinmachiningprocessispredictedbytheproposedmode1．Theresultsshowthattheproposedpredictionmethodhasagoodpredictionperformance．Keywords：Dynamicco

4、mpliance，GeneralizedMarkovchain，GeneralizedintervaI．Predictiononthegeneralizedintervalprobabilitytheory[5-6]，1．Introductionisproposedtoimprovetheaccuracyandreliabilityofprediction．IngeneralizedMarkovchainmodel，Dynamiccomplianceistheabilityofmachineprobabilityandintervalareusedtocaptureth

5、euneer—toolcomponentsdeformationtoresistexternalforce．tainty．ApredictionmethodofgeneralizedMarkovItreflectsthestabilityofNCmachinetoolmachiningchainmodelisdeveloped．Inordertopredictdynam—processandhasadirecteffectonmachiningaccura—iccompliance，anexperimentofrelativeexcitationiscy．Allkind

6、sofuncertaintywillbeencounteredindeveloped．Thedynamiccompliancecalculationmod—machiningprocess，suchasrandomnessofmachiningelisproposedforcalculatingpredictionhistoricalprocess，measurementmxor，numericalerror，anddate．Thedynamiccomplianceinmachiningprocessthelackofknowledge[1—2]．Uncertainty

7、willcauseispredictedbythegeneralizedMarkovchainmode1．theloweraccuracyandreliability．ThetraditionalMarkovchainmodelisarandomprocesswiththeca—2．Dynamiccompliancecalculationmodelpabilityofprediction，andhasbeenusedforengi—neeringprediction[3]．TheprobabilisticparameterInordert