基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx

上传人:b****3 文档编号:3800494 上传时间:2022-11-25 格式:DOCX 页数:16 大小:302.40KB
下载 相关 举报
基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx_第1页
第1页 / 共16页
基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx_第2页
第2页 / 共16页
基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx_第3页
第3页 / 共16页
基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx_第4页
第4页 / 共16页
基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx_第5页
第5页 / 共16页
点击查看更多>>
下载资源
资源描述

基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx

《基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx》由会员分享,可在线阅读,更多相关《基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx(16页珍藏版)》请在冰豆网上搜索。

基于事例推理的夹具设计研究与应用Application机械类外文翻译.docx

基于事例推理的夹具设计研究与应用Application机械类外文翻译

Applicationanddevelopment

Ofcasebasedreasoninginfixturedesign

Abstract:

Basedonthecasebaseddesigning(CBD)methodology,thefixturesimilarityisintworespects:

thefunctionandthestructureinformation.Then,thecomputeraidedfixturedesignsystemiscreatedoncasebasedreasoning(CBR),inwhichtheattributesofthemainfeaturesofworkpieceandstructureoffixtureascaseindexcodearedesignedfortheretrieveofthesimilarcases,andthestructureandhierarchicalrelationofcaselibraryaresetupforstore.Meanwhile,thealgorithmbasedontheknowledgeguidedintheretrieveofthesimilarcases,thestrategyofcaseadaptationandcasestorageinwhichthecaseidentifcationnumberisusedtodistinguishfromsimilarcasesarepresented.Theapplicationofthesysteminsomeprojectsimprovesthedesignefficiencyandgetsagoodresult.

Keywords:

casebasedreasoning。

fixturedesign。

computeraideddesign(CAD)

Fixturesaredevicesthatserveasthepurposeofholdingtheworkpiecesecurelyandaccurately,andmaintainingaconsistentrelationshipwithrespecttothetoolswhilemachining.Becausethefixturestructuredependsonthefeatureoftheproductandthestatusoftheprocessplanningintheenterprise,itsdesignisthebottleneckduringmanufacturing,whichrestrainstoimprovetheefficiencyandleadtime.Andfixturedesignisacomplicatedprocess,basedonexperiencethatneedscomprehensivequalitativeknowledgeaboutanumberofdesignissuesincludingworkpiececonfiguration,manufacturingprocessesinvolved,andmachiningenvironment.ThisisalsoaverytimeconsumingworkwhenusingtraditionalCADtools(suchasUnigraphics,CATIAorPro/E),whicharegoodatperformingdetaileddesigntasks,butprovidefewbenefitsfortakingadvantageofthepreviousdesignexperienceandresources,whicharepreciselythekeyfactorsinimprovingtheefficiency.Themethodologyofcasebasedreasoning(CBR)adaptsthesolutionofapreviouslysolvedcasetobuildasolutionforanewproblemwiththefollowingfoursteps:

retrieve,reuse,revise,andretain[1].Thisisamoreusefulmethodthantheuseofanexpertsystemtosimulatehumanthoughtbecauseproposingasimilarcaseandapplyingafewmodificationsseemstobeselfexplanatoryandmoreintuitivetohumans.Sovariouscasebaseddesignsupporttoolshavebeendevelopedfornumerousareas[2-4],suchasininjectionmoldinganddesign,architecturaldesign,diecastingdiedesign,processplanning,andalsoinfixturedesign.Sunusedsixdigitalstocomposetheindexcodethatincludedworkpieceshape,machineportion,bushing,the1stlocatingdevice,the2ndlocatingdeviceandclampingdevice[5].Butthesystemcannotbeusedforotherfixturetypesexceptfordrillfixtures,andcannotsolvetheproblemofstorageofthesameindexcodethatneedstoberetained,whichisveryimportantinCBR[6].

1ConstructionofaCaseIndexandCaseLibrary

1.1Caseindex

Thecaseindexshouldbecomposedofallfeaturesoftheworkpiece,whicharedistinguishedfromdifferentfixtures.Usingallofthemwouldmaketheoperationinconvenient.Becausetheformsofthepartsarediverse,andthetechnologyrequirementsofmanufactureintheenterprisealsodevelopcontinuously,lotsoffeaturesusedasthecaseindexwillmakethesearchrateslow,andthemainfeatureunimportant,forthereasonthattherelativeweightwhichisallottedtoeveryfeaturemustdiminish.Andontheotherhand,itishardtoincludeallthefeaturesinthecaseindex.

Therefore,consideringthepracticalityandthedemandofrapiddesign,thecaseindexincludesboththemajorfeatureoftheworkpieceandthestructureoffixture.Thecaseindexcodeismadeupof16digits:

13digitsforcasefeaturesand3digitsforcaseidentificationnumber.

Thefirst13digitsrepresent13features.Eachdigitiscorrespondingtoanattributeofthefeature,whichmaybeoneof“*”,“?

”,“1”,“2”,…,“A”,“B”,…,“Z”,…,etc.Inwhich,“*”meansanyone,“?

”uncertain,“0”nothing.

Thesystemrules:

fixturetype,workpieceshape,locatingmodelcannotbe“*”or“?

”.Whenthesystemisdesigned,theattributeinformationofthethreeitemsdoesnothavetheseoptions,whichmeansthecertainattributemustbeselected.

Thelastthreedigitsarethecaseidentificationnumber,whichmeansthe13digitsofthecasefeaturearethesame,andthenumberofthesethreedigitsisusedfordistinguishingthem.

Thesystemalsorules:

“000”isaprototypecase,whichisusedforretrieval,andothercasesare“001”,“002”,…,whichareusedforreferencecasestobesearchedbydesigners.Ifoccasionallyoneofthemneedstobechangedastheprototypecase,firstitmustberequiredtoapplytochangetheoneto“000”,andtheformerischangedtoreferentialcaseautomatically.

TheconstructionofthecaseindexcodeisshowninFig.1.

1.2Caselibrary

Thecaselibraryconsistsoflotsofpredefinedcases.Caserepresentationisoneofthemostimportantissuesincasebasedreasoning.Socompoundingwiththeindexcode,.

1.3HierarchicalformofCase

Thestructuresimilarityofthefixtureisrepresentedasthewholefixturesimilarity,componentssimilarityandcomponentsimilarity.Sothewholefixturecaselibrary,componentscaselibrary,componentcaselibraryoffixtureareformedcorrespondingly.Usuallydesigninformationofthewholefixtureiscomposedofworkpieceinformationandworkpieceprocedureinformation,whichrepresentthefixturesatisfyingthespecificallydesigningfunctiondemand.Thewholefixturecaseismadeupoffunctioncomponents,whicharedescribedbythefunctioncomponents’namesandnumbers.Thecomponentscaserepresentsthemembers.(functioncomponentandotherstructurecomponents,maindrivenparameter,thenumber,andtheirconstrainrelations.)Thecomponentcase(thelowestlayerofthefixture)isthestructureoffunctioncomponentandothercomponents.Inthemodernfixturedesigntherearelotsofparametricstandardpartsandcommonnonstandardparts.Sothecomponentcaselibraryshouldrecordthespecificationparameterandthewayinwhichitkeepsthem.

2StrategyofCaseRetrieval

Inthecasebaseddesignoffixtures,themostimportantthingistheretrievalofthesimilarity,whichcanhelptoobtainthemostsimilarcase,andtocutdownthetimeofadaptation.Accordingtotherequirementoffixturedesign,thestrategyofcaseretrievalcombinesthewayofthenearestneighborandknowledgeguided.Thatis,firstsearchondepth,thenonbreadth。

theknowledgeguidedstrategymeanstosearchontheknowledgerulefromroottotheobject,whichisfirstlysearchedbythefixturetype,thenbytheshapeoftheworkpiece,thirdlybythelocatingmethod.Forexample,ifthecaseindexcodeincludesthemillingfixtureoffixturetype,thesearchisjustforallmillingfixtures,thenforboxofworkpieceshape,thethirdfor1plane+2pineoflocatingmethod.Ifthereisnomatchofit,thenthesearchstopsondepth,andreturnstotheupperlayer,andretrievesalltherelativecasesonbreadth.

Retrievalalgorithms:

1)Accordingtothecaseindexinformationoffixturecaselibrary,searchtherelevantcaselibrary。

2)Matchthecaseindexcodewiththecodeofeachcaseofthecaselibrary,andcalculatethevalueofthesimilaritymeasure。

3)Sorttheorderofsimilaritymeasure,thebiggestvalue,whichisthemostanalogicalcase.

Similaritybetweentwocasesisbasedonthesimilaritybetweenthetwocases.features.Thecalculationofsimilaritymeasuredependsonthetypeofthefeature.Thevalueofsimilaritycanbecalculatedfornumericalvalues,forexample,compareWorkpiecewiththeweightof50kgand20kg.Thevaluecanalsobecalculatedbetweennonnumericalvalues,forexample,nowthefirst13digitsindexcodeisallnonnumericalvalues.Thesimilaritymeasureofafixtureiscalculatedasfollows:

whereSisthesimilaritymeasureofcurrentfixture,nisthenumberoftheindexfeature,

istheweightofeachfeature,

isthesimilaritymeasureoftheattribute

ofthei2thfeaturewiththeattribute

ofrelativefeatureofthej-thcaseinthecaselibrary.Atthesametime,

thevaluecountsasfollows:

.

Where

isthevalueoftheindexattributeofthei-thfeature,and

isthevalueofattributeoftherelativei-thfeatureofthej-thcaseincaselibrary.

Sotherearetwomethodstoselecttheanalogicalfixture.Oneistosetthevalue.Ifthevaluesofsimilaritymeasureofcurrentcaseswerelessthanagivenvalue,thosecaseswouldnotbeselectedasanalogicalcases.Whenthecaselibraryisinitiallysetup,andthereareonlyafewcases,thevaluecanbesetsmaller.Iftherearelotsofanalogicalcases,thevalueshouldgetlarger.Theotherisjusttosetthenumberoftheanalogicalcases(suchas10),whichisthelargestvalueofsimilaritymeasurefromthesortedorder.

3CaseadaptationandCaseStorage

3.1Caseadaptation

Themodificationoftheanalogicalcaseinthefixturedesignincludesthefollowingthreecases:

1)Thesubstitutionofcomponentsandthecomponent。

2)Adjustingthedimensionofcomponentsandthecomponentwhiletheformremains。

3)Theredesignofthemodel.

Ifthecomponentsandcomponentofthefixturearecommonobjects,theycanbeedited,substitutedanddeletedwithtools,whichhavebeendesigned.

3.2Casestorage

Beforesavinganewfixturecaseinthecaselibrary,thedesignermustconsiderwhetherthesavingisvaluable.Ifthecasedoesnotincreasetheknowledgeofthesystem,itisnotnecessarytostoreitinthecaselibrary.Ifitisvaluable,thenthedesignermustanalyzeitbeforesavingittoseewhetherthecas

展开阅读全文
相关资源
猜你喜欢
相关搜索

当前位置:首页 > 职业教育 > 职业技术培训

copyright@ 2008-2022 冰豆网网站版权所有

经营许可证编号:鄂ICP备2022015515号-1