供应链下的多级存货管理外文翻译.docx

供应链下的多级存货管理外文翻译.docx

《供应链下的多级存货管理外文翻译.docx》由会员分享，可在线阅读，更多相关《供应链下的多级存货管理外文翻译.docx（11页珍藏版）》请在冰豆网上搜索。

供应链下的多级存货管理外文翻译

本科毕业论文（设计）

外文翻译

原文：

Multi-echeloninventorymanagementinsupplychains

Historically,theechelonsofthesupplychain,warehouse,distributors,retailers,etc.,havebeenmanagedindependently,bufferedbylargeinventories.Increasingcompetitivepressuresandmarketglobalizationareforcingfirmstodevelopsupplychainsthatcanquicklyrespondtocustomerneeds.Toremaincompetitiveanddecreaseinventory,thesefirmsmustusemulti-echeloninventorymanagementinteractively,whilereducingoperatingcostsandimprovingcustomerservice.

Supplychainmanagement（SCM）isanintegrativeapproachforplanningandcontrolofmaterialsandinformationflowswithsuppliersandcustomers,aswellasbetweendifferentfunctionswithinacompany.Thisareahasdrawnconsiderableattentioninrecentyearsandisseenasatoolthatprovidescompetitivepower.SCMisasetofapproachestointegratesuppliers,manufacturers,warehouses,andstoresefficiently,sothatmerchandiseisproducedanddistributedatrightquantities,totherightlocationsandattherighttime,inordertominimizesystem-widecostswhilesatisfyingservice-levelrequirements.Sothesupplychainconsistsofvariousmembersorstages.Asupplychainisadynamic,stochastic,andcomplexsystemthatmightinvolvehundredsofparticipants.

Inventoryusuallyrepresentsfrom20to60percentofthetotalassetsofmanufacturingfirms.Therefore,inventorymanagementpoliciesprovecriticalindeterminingtheprofitofsuchfirms.Inventorymanagementis,toagreaterextent,relevantwhenawholesupplychain（SC）,namelyanetworkofprocurement,transformation,anddeliveringfirms,isconsidered.InventorymanagementisindeedamajorissueinSCM,i.e.anapproachthataddressesSCissuesunderanintegratedperspective.

InventoriesexistthroughouttheSCinvariousformsforvariousreasons.ThelackofacoordinatedinventorymanagementthroughouttheSCoftencausesthebullwhipeffect,namelyanamplificationofdemandvariabilitymovingtowardstheupstreamstages.Thiscausesexcessiveinventoryinvestments,lostrevenues,misguidedcapacityplans,ineffectivetransportation,missedproductionschedules,andpoorcustomerservice.

Manyscholarshavestudiedtheseproblems,aswellasemphasizedtheneedofintegrationamongSCstages,tomakethechaineffectivelyandefficientlysatisfycustomerrequests（e.g.reference）.Besidetheintegrationissue,uncertaintyhastobedealtwithinordertodefineaneffectiveSCinventorypolicy.Inadditiontotheuncertaintyonsupply（e.g.leadtimes）anddemand,informationdelaysassociatedwiththemanufacturinganddistributionprocessescharacterizeSCs.

Inventorymanagementinmulti-echelonSCsisanimportantissue,becausetherearemanyelementsthathavetocoordinatewitheachother.Theymustalsoarrangetheirinventoriestocoordinate.Therearemanyfactorsthatcomplicatesuccessfulinventorymanagement,e.g.uncertaindemands,leadtimes,productiontimes,productprices,costs,etc.,especiallytheuncertaintyindemandandleadtimeswheretheinventorycannotbemanagedbetweenechelonsoptimally.

Mostmanufacturingenterprisesareorganizedintonetworksofmanufacturinganddistributionsitesthatprocurerawmaterial,processthemintofinishedgoods,anddistributethefinishgoodstocustomers.Theterms‘multi-echelon’or‘multilevel‘production/distributionnetworksarealsosynonymouswithsuchnetworks（orSC）,whenanitemmovesthroughmorethanonestepbeforereachingthefinalcustomer.InventoriesexistthroughouttheSCinvariousformsforvariousreasons.Atanymanufacturingpoint,theymayexistasrawmaterials,workinprogress,orfinishedgoods.Theyexistatthedistributionwarehouses,andtheyexistin-transit,or‘inthepipeline’,oneachpathlinkingthesefacilities.

Manufacturersprocurerawmaterialfromsuppliersandprocessthemintofinishedgoods,sellthefinishedgoodstodistributors,andthentoretailand/orcustomers.Whenanitemmovesthroughmorethanonestagebeforereachingthefinalcustomer,itformsa‘multi-echelon’inventorysystem.Theechelonstockofastockpointequalsallstockatthisstockpoint,plusin-transittooron-handatanyofitsdownstreamstockpoints,minusthebackordersatitsdownstreamstockpoints.

Theanalysisofmulti-echeloninventorysystemsthatpervadesthebusinessworldhasalonghistory.Multi-echeloninventorysystemsarewidelyemployedtodistributeproductstocustomersoverextensivegeographicalareas.Giventheimportanceofthesesystems,manyresearchershavestudiedtheiroperatingcharacteristicsunderavarietyofconditionsandassumptions.Sincethedevelopmentoftheeconomicorderquantity（EOQ）formulabyHarris（1913）,researchersandpractitionershavebeenactivelyconcernedwiththeanalysisandmodelingofinventorysystemsunderdifferentoperatingparametersandmodelingassumptions.Researchonmulti-echeloninventorymodelshasgainedimportanceoverthelastdecademainlybecauseintegratedcontrolofSCsconsistingofseveralprocessinganddistributionstageshasbecomefeasiblethroughmoderninformationtechnology.ClarkandScarfwerethefirsttostudythetwo-echeloninventorymodel.Theyprovedtheoptimalityofabase-stockpolicyforthepure-serialinventorysystemanddevelopedanefficientdecomposingmethodtocomputetheoptimalbase-stockorderingpolicy.BesslerandVeinottextendedtheClarkandScarfmodeltoincludegeneralarborescentstructures.Thedepot-warehouseproblemdescribedabovewasaddressedbyEppenandSchragewhoanalyzedamodelwithastocklesscentraldepot.Theyderivedaclosed-formexpressionfortheorder-up-to-levelundertheequalfractileallocationassumption.Severalauthorshavealsoconsideredthisprobleminvariousforms.Owingtothecomplexityandintractabilityofthemulti-echelonproblemHadleyandWhitinrecommendtheadoptionofsingle-location,single-echelonmodelsfortheinventorysystems.

Sherbrookeconsideredanorderingpolicyofatwo-echelonmodelforwarehouseandretailer.Itisassumedthatstockoutsattheretailersarecompletelybacklogged.Also,SherbrookeconstructedtheMETRIC（multi-echelontechniqueforcoverableitemcontrol）model,whichidentifiesthestocklevelsthatminimizetheexpectednumberofbackordersatthelower-echelonsubjecttoabudgetconstraint.Thismodelisthefirstmulti-echeloninventorymodelformanagingtheinventoryofserviceparts.Thereafter,alargesetofmodelswhichgenerallyseektoidentifyoptimallotsizesandsafetystocksinamulti-echelonframework,wereproducedbymanyresearchers.Inadditiontoanalyticalmodels,simulationmodelshavealsobeendevelopedtocapturethecomplexinteractionofthemulti-echeloninventoryproblems.

Sofarliteraturehasdevotedmajorattentiontotheforecastingoflumpydemand,andtothedevelopmentofstockpoliciesformulti-echelonSCsInventorycontrolpolicyformulti-echelonsystemwithstochasticdemandhasbeenawidelyresearchedarea.MorerecentpapershavebeencoveredbySilverandPyke.Theadvantageofcentralizedplanning,availableinperiodicreviewpolicies,canbeobtainedincontinuousreviewpolicies,bydefiningthereorderlevelsofdifferentstages,intermsofechelonstockratherthaninstallationstock.

Rauetal.,DiksanddeKok,DongandLee,MitraandChatterjee,Hariga,Chen,AxsaterandZhang,NozickandTurnquist,andSoandZhenguseamathematicmodelingtechniqueintheirstudiestomanagemulti-echeloninventoryinSCs.DiksanddeKok’sstudyconsidersadivergentmulti-echeloninventorysystem,suchasadistributionsystemoraproductionsystem,andassumesthattheorderarrivesafterafixedleadtime.Hariga,presentsastochasticmodelforasingle-periodproductionsystemcomposedofseveralassembly/processingandstoragefacilitiesinseries.Chen,AxsaterandZhang,andNozickandTurnquistconsideratwo-stageinventorysystemintheirpapers.AxsaterandZhangandNozickandTurnquistassumethattheretailersfacestationaryandindependentPoissondemand.MitraandChatterjeeexamineDeBodtandGraves’model（1985）,whichtheydevelopedintheirpaper’Continuous-reviewpoliciesforamulti-echeloninventoryproblemwithstochasticdemand’,forfast-movingitemsfromtheimplementationpointofview.Theproposedmodificationofthemodelcanbeextendedtomulti-stageserialandtwo-echelonassemblysystems.InRauetal.’smodel,shortageisnotallowed,leadtimeisassumedtobenegligible,anddemandrateandproductionrateisdeterministicandconstant.SoandZhengusedananalyticalmodeltoanalyzetwoimportantfactorsthatcancontributetothehighdegreeoforder-quantityvariabilityexperiencedbysemiconductormanufacturers:

supplier’sleadtimeandforecastdemandupdating.Theyassumethattheexternaldemandsfacedbytheretailorarecorrelatedbetweentwosuccessivetimeperiodsandthattheretailerusesthelatestdemandinformationtoupdateitsfuturedemandforecasts.Furthermore,theyassumethatthesupplier’sdeliveryleadtimesarevariableandareaffectedbytheretailer’sorderquantities.DongandLee’spaperrevisitstheserialmulti-echeloninventorysystemofClarkandScarfanddevelopsthreekeyresults.First,theyprovideasimplelower-boundapproximationtotheoptimalecheloninventorylevelsandanupperboundtothetotalsystemcostforthebasicmodelofClarkandScarf.Second,theyshowthatthestructureoftheoptimalstockingpolicyofClarkandScarfholdsundertime-correlateddemandprocessingusingaMartingalemodelofforecastevolution.Third,theyextendtheapproximationtothetime-correlateddemandprocessandstudy,inpart