简介：Thispaperproposesagraphical-basedmethodologytoevaluatetheperformanceofamanufacturingsystemintermsofnetworkmodel.Wefocusonamanufacturingsystemwhichconsistsofmultipledistinctproductionlines.Atransformationtechniqueisdevelopedtobuildthemanufacturingsystemasamanufacturingnetwork.Insuchamanufacturingnetwork,thecapacityofeachmachineismultistateduetofailure,partialfailure,ormaintenance.Thus,thismanufacturingnetworkisalsoregardedasamultistatenetwork.Weevaluatetheprobabilitythatthemanufacturingnetworkcanmeetagivendemand,wheretheprobabilityisreferredtoasthesystemreliability.Asimplealgorithmintegratingdecompositiontechniqueisproposedtogeneratetheminimalcapacityvectorsthatmachinesshouldprovidetoeventuallysatisfydemand.Thesystemreliabilityisderivedintermsofsuchcapacityvectorsafterwards.ApracticalapplicationinthecontextofICcardmanufacturingsystemisutilizedtodemonstratetheperformanceevaluationprocedure.

简介：Thetraditionaldataenvelopmentanalysis(DEA)modelcanevaluatetherelativeefficienciesofasetofdecisionmakingunits(DMUs)withexactvaluesofinputsandoutputs,butitcannothandleimprecisedata.Imprecisedata,forexample,canbeexpressedintheformoftheintervaldataormixturesofintervaldataandordinaldata.Inthisstudy,across-efficiencymethodisintroducedintotheDEAmodeltocalculatetheintervalofcross-efficiencyvalues,basedonwhichanewTOPSISmethodisproposedtoranktheDMUs.Twoexamplesarepresentedtoillustrateandvalidatetheproposedmethod.