供应链下的多级存货管理.docx

供应链下的多级存货管理.docx

《供应链下的多级存货管理.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.docx（4页珍藏版）》请在冰豆网上搜索。

供应链下的多级存货管理

本科毕业论文设计外文翻译原文Multi-echeloninventorymanagementinsupplychainsHistoricallytheechelonsofthesupplychainwarehousedistributorsretailersetc.havebeenmanagedindependentlybufferedbylargeinventories.Increasingcompetitivepressuresandmarketglobalizationareforcingfirmstodevelopsupplychainsthatcanquicklyrespondtocustomerneeds.Toremaincompetitiveanddecreaseinventorythesefirmsmustusemulti-echeloninventorymanagementinteractivelywhilereducingoperatingcostsandimprovingcustomerservice.SupplychainmanagementSCMisanintegrativeapproachforplanningandcontrolofmaterialsandinformationflowswithsuppliersandcustomersaswellasbetweendifferentfunctionswithinacompany.Thisareahasdrawnconsiderableattentioninrecentyearsandisseenasatoolthatprovidescompetitivepower.SCMisasetofapproachestointegratesuppliersmanufacturerswarehousesandstoresefficientlysothatmerchandiseisproducedanddistributedatrightquantitiestotherightlocationsandattherighttimeinordertominimizesystem-widecostswhilesatisfyingservice-levelrequirements.Sothesupplychainconsistsofvariousmembersorstages.Asupplychainisadynamicstochasticandcomplexsystemthatmightinvolvehundredsofparticipants.Inventoryusuallyrepresentsfrom20to60percentofthetotalassetsofmanufacturingfirms.Thereforeinventorymanagementpoliciesprovecriticalindeterminingtheprofitofsuchfirms.InventorymanagementistoagreaterextentrelevantwhenawholesupplychainSCnamelyanetworkofprocurementtransformationanddeliveringfirmsisconsidered.InventorymanagementisindeedamajorissueinSCMi.e.anapproachthataddressesSCissuesunderanintegratedperspective.InventoriesexistthroughouttheSCinvariousformsforvariousreasons.ThelackofacoordinatedinventorymanagementthroughouttheSCoftencausesthebullwhipeffectnamelyanamplificationofdemandvariabilitymovingtowardstheupstreamstages.Thiscausesexcessiveinventoryinvestmentslostrevenuesmisguidedcapacityplansineffectivetransportationmissedproductionschedulesandpoorcustomerservice.ManyscholarshavestudiedtheseproblemsaswellasemphasizedtheneedofintegrationamongSCstagestomakethechaineffectivelyandefficientlysatisfycustomerrequestse.g.reference.BesidetheintegrationissueuncertaintyhastobedealtwithinordertodefineaneffectiveSCinventorypolicy.Inadditiontotheuncertaintyonsupplye.g.leadtimesanddemandinformationdelaysassociatedwiththemanufacturinganddistributionprocessescharacterizeSCs.Inventorymanagementinmulti-echelonSCsisanimportantissuebecausetherearemanyelementsthathavetocoordinatewitheachother.Theymustalsoarrangetheirinventoriestocoordinate.Therearemanyfactorsthatcomplicatesuccessfulinventorymanagemente.g.uncertaindemandsleadtimesproductiontimesproductpricescostsetc.especiallytheuncertaintyindemandandleadtimeswheretheinventorycannotbemanagedbetweenechelonsoptimally.Mostmanufacturingenterprisesareorganizedintonetworksofmanufacturinganddistributionsitesthatprocurerawmaterialprocessthemintofinishedgoodsanddistributethefinishgoodstocustomers.Theterms�6�7multi-echelon�6�8or�6�7multilevel�6�7production/distributionnetworksarealsosynonymouswithsuchnetworksorSCwhenanitemmovesthroughmorethanonestepbeforereachingthefinalcustomer.InventoriesexistthroughouttheSCinvariousformsforvariousreasons.Atanymanufacturingpointtheymayexistasrawmaterialsworkinprogressorfinishedgoods.Theyexistatthedistributionwarehousesandtheyexistin-transitor�6�7inthepipeline�6�8oneachpathlinkingthesefacilities.Manufacturersprocurerawmaterialfromsuppliersandprocessthemintofinishedgoodssellthefinishedgoodstodistributorsandthentoretailand/orcustomers.Whenanitemmovesthroughmorethanonestagebeforereachingthefinalcustomeritformsa�6�7multi-echelon�6�8inventorysystem.Theechelonstockofastockpointequalsallstockatthisstockpointplusin-transittooron-handatanyofitsdownstreamstockpointsminusthebackordersatitsdownstreamstockpoints.Theanalysisofmulti-echeloninventorysystemsthatpervadesthebusinessworldhasalonghistory.Multi-echeloninventorysystemsarewidelyemployedtodistributeproductstocustomersoverextensivegeographicalareas.Giventheimportanceofthesesystemsmanyresearchershavestudiedtheiroperatingcharacteristicsunderavarietyofconditionsandassumptions.SincethedevelopmentoftheeconomicorderquantityEOQformulabyHarris1913researchersandpractitionershavebeenactivelyconcernedwiththeanalysisandmodelingofinventorysystemsunderdifferentoperatingparametersandmodelingassumptions.Researchonmulti-echeloninventorymodelshasgainedimportanceoverthelastdecademainlybecauseintegratedcontrolofSCsconsistingofseveralprocessinganddistributionstageshasbecomefeasiblethroughmoderninformationtechnology.ClarkandScarfwerethefirsttostudythetwo-echeloninventorymodel.Theyprovedtheoptimalityofabase-stockpolicyforthepure-serialinventorysystemanddevelopedanefficientdecomposingmethodtocomputetheoptimalbase-stockorderingpolicy.BesslerandVeinottextendedtheClarkandScarfmodeltoincludegeneralarborescentstructures.Thedepot-warehouseproblemdescribedabovewasaddressedbyEppenandSchragewhoanalyzedamodelwithastocklesscentraldepot.Theyderivedaclosed-formexpressionfortheorder-up-to-levelundertheequalfractileallocationassumption.Severalauthorshavealsoconsideredthisprobleminvariousforms.Owingtothecomplexityandintractabilityofthemulti-echelonproblemHadleyandWhitinrecommendtheadoptionofsingle-locationsingle-echelonmodelsfortheinventorysystems.Sherbrookeconsideredanorderingpolicyofatwo-echelonmodelforwarehouseandretailer.Itisassumedthatstockoutsattheretailersarecompletelybacklogged.AlsoSherbrookeconstructedtheMETRICmulti-echelontechniqueforcoverableitemcontrolmodelwhichidentifiesthestocklevelsthatminimizetheexpectednumberofbackordersatthelower-echelonsubjecttoabudgetconstraint.Thismodelisthefirstmulti-echeloninventorymodelformanagingtheinventoryofserviceparts.Thereafteralargesetofmodelswhichgenerallyseektoidentifyoptimallotsizesandsafetystocksinamulti-echelonframeworkwereproducedbymanyresearchers.Inadditiontoanalyticalmodelssimulationmodelshavealsobeendevelopedtocapturethecomplexinteractionofthemulti-echeloninventoryproblems.Sofarliteraturehasdevotedmajorattentiontotheforecastingoflumpydemandandtothedevelopmentofstockpoliciesformulti-echelonSCsInventorycontrolpolicyformulti-echelonsystemwithstochasticdemandhasbeenawidelyresearchedarea.MorerecentpapershavebeencoveredbySilverandPyke.Theadvantageofcentralizedplanningavailableinperiodicreviewpoliciescanbeobtainedincontinuousreviewpoliciesbydefiningthereorderlevelsofdifferentstagesintermsofechelonstockratherthaninstallationstock.Rauetal.DiksanddeKokDongandLeeMitraandChatterjeeHarigaChenAxsaterandZhangNozickandTurnquistandSoandZhenguseamathematicmodelingtechniqueintheirstudiestomanagemulti-echeloninventoryinSCs.DiksanddeKok�6�8sstudyconsidersadivergentmulti-echeloninventorysystemsuchasadistributionsystemoraproductionsystemandassumesthattheorderarrivesafterafixedleadtime.Harigapresentsastochasticmodelforasingle-periodproductionsystemcomposedofseveralassembly/processingandstoragefacilitiesinseries.ChenAxsaterandZhangandNozickandTurnquistconsideratwo-stageinventorysystemintheirpapers.AxsaterandZhangandNozickandTurnquistassumethattheretailersfacestationaryandindependentPoissondemand.MitraandChatterjeeexamineDeBodtandGraves�6�8model1985whichtheydevelopedintheirpaper�6�8Continuous-reviewpoliciesforamulti-echeloninventoryproblemwithstochasticdemand�6�8forfast-movingitemsfromtheimplementationpointofview.Theproposedmodificationofthemodelcanbeextendedtomulti-stageserialandtwo-echelonassemblysystems.InRauetal.�6�8smodelshortageisnotallowedleadtimeisassumedtobenegligibleanddemandrateandproductionrateisdeterministicandconstant.SoandZhengusedananalyticalmodeltoanalyzetwoimportantfactorsthatcancontributetothehighdegreeoforder-quantityvariabilityexperiencedbysemiconductormanufacturers:

supplier�6�8sleadtimeandforecastdemandupdating.Theyassumethattheexternaldemandsfacedbytheretailorarecorrelatedbetweentwosuccessivetimeperiodsandthattheretailerusesthelatestdemandinformationtoupdateitsfuturedemandforecasts.Furthermoretheyassumethatthesupplier�6�8sdeliveryleadtimesarevariableandareaffectedbytheretailer�6�8sorderquantities.DongandLee�6�8spaperrevisitstheserialmulti-echeloninventorysystemofClarkandScarfanddevelopsthreekeyresults.Firsttheyprovideasimplelower-boundapproximationtotheoptimalecheloninventorylevelsandanupperboundtothetotalsystemcostforthebasicmodelofClarkandScarf.SecondtheyshowthatthestructureoftheoptimalstockingpolicyofClarkandScarfholdsundertime-correlateddemandprocessingusingaMartingalemodelofforecastevolution.Thirdtheyextendtheapproximationtothetime-correlateddemandprocessandstudyinparticularforanautoregressivedemandmodeltheimpactofleadtimesandautocorrelationonthep