logwareSELECTED COMPUTER PROGRAMS FOR PLANNING.docx

logwareSELECTED COMPUTER PROGRAMS FOR PLANNING.docx

《logwareSELECTED COMPUTER PROGRAMS FOR PLANNING.docx》由会员分享，可在线阅读，更多相关《logwareSELECTED COMPUTER PROGRAMS FOR PLANNING.docx（72页珍藏版）》请在冰豆网上搜索。

logwareSELECTEDCOMPUTERPROGRAMSFORPLANNING

SELECTEDCOMPUTERPROGRAMS

FORLOGISTICS/SUPPLYCHAINPLANNING

Version5.0

RonaldH.Ballou

WeatherheadSchoolofManagement

CaseWesternReserveUniversity

（C）Copyright1992-2004RonaldH.BallouAllrightsreserved

SELECTEDCOMPUTERPROGRAMS

FORLOGISTICS/SUPPYCHAINPLANNING

LOGWAREisacollectionofselectedsoftwareprogramsthatisusefulforanalyzingavarietyoflogistics/supplychainproblemsandcasestudies.Itcontainsthefollowingmodules.

Module

Page

FORECASTForecaststimeseriesdatabymeansofexponentialsmoothingandtimeseriesdecompositionmethods

5

ROUTEDeterminestheshortestpaththroughanetworkofroutes

9

ROUTESEQDeterminesthebestsequencetovisitstopsonaroute

13

ROUTERDevelopsroutesandschedulesformultipletrucksservingmultiplestops

15

INPOLFindsoptimalinventoryorderingpoliciesbasedoneconomicorderquantityprinciples

24

COGFindsthelocationofasinglefacilitybytheexactcenter-of-gravitymethod

33

MULTICOGLocatesaselectednumberoffacilitiesbytheexactcenter-of-gravitymethod

37

PMEDLocatesaselectednumberoffacilitiesbytheP-medianmethod

41

WARELOCAAwarehouselocationprogramforspecificallyanalyzingtheUsemoreSoapCompanycasestudy

45

LAYOUTPositionsproductsinwarehousesandotherfacilities

47

MILESComputesapproximatedistancebetweentwopointsusinglatitude-longitudeorlinear-gridcoordinatepoints

49

TRANLPSolvesthetransportationmethodoflinearprogramming

51

LNPROGSolvesgenerallinearprogrammingproblemsbymeansofthesimplexmethod

53

MIPROGSolvesthemixedintegerlinearprogrammingproblembymeansofbranchandbound

55

MULREGFindslinearregressionequationsbymeansofthestepwiseprocedureofregression/correlationanalysis

57

SCSIMSimulatestheflowofaproductthroughfiveechelonsofasupplychannel

62

Eachmoduleisselectedfromthefollowingmasterscreenbyclickingontheappropriatebutton.

HARDWAREREQUIREMENTS

LOGWAREisdesignedformicrocomputersoperatingunderWINDOWS98,NT,2000,orXP.Atleast16MBofRAMshouldbeinstalled.Harddrivespaceofatleast10MBshouldbeavailable.Acolormonitorcapableofproducingatleast640x480pixelsresolutionisneeded,although800x600isbetterand1024x768ispreferred.Resolutionsgreaterthan1024x768pixelsarenotsupported.Alaserprinterispreferred.Amouseisneeded.A3½floppydriveand/oracompactdiskreaderareneeded.

INSTALLINGTHESOFTWAREONAHARDDRIVE

Placetheprogramcompactdisksintheappropriatedrives.InWINDOWS,clickontheStartbuttonandthenselecttheRunoptionfrompop-upmenu.Type“X:

Setup.exe”（“X”beingtheletterdesignatedforyourCDdrive）.TheprogrammayalsobeinstalledwithWindows’Start,Settings,ControlPanel,Add/RemovePrograms,Installoption.Changethesubdirectoryunderwhichtheprogramwillbeinstalledifthedefaultsubdirectoryisnotdesired.

RUNNINGTHEPROGRAMS

Aftertheprogramisinstalled,clickontheStartbuttonandselectPrograms.ChoosetheLogwareicontoactivatetheprogram.Clickonthedesiredprogrammodule.AshortcuticonontheDesktopmayalsobecreated.

EDITINGTHEDATA

Inthosemoduleswhereascreendataeditorispresent,thefirstactionistoopenadatafilebyclickingonthemodule’sStartbutton.Ifafileisnamedthatisnotinthecurrentlistoffiles,adatashellwillbecreatedintowhichanewproblemmaybeentered.Theuseoftheeditorissimpleandtransparentwithalittlepractice;however,afewcommentsaboutitsusemayhelptogetstarted.

PresstheInskeytostartanewlineofdatainamatrix.Thenormalactionistoinsertatextrowattheendofthematrix.TheAddbuttonmayalsobepressed.Thiswillallowarowtobeaddedattheendofthematrixaswellaswithinthematrix.Positionthecursorinthematrixrowwheretherowistobeadded.

PressingtheEsckeyclearsamatrixcell.

∙PressingtheDeletebuttondeletestherowinamatrixhighlightedbythecurrentcursorposition.

∙IfColumnarithmeticistobeused,highlightthematrixcolumnonwhichtheactionistoapply.

Alternatively,thedataforeachmoduleexceptSCSIMmaybecreatedandeditedwiththeuseofExcel.ItisexpectedthattheuserhasabasicknowledgeofExceluse.

COPYINGTHEINSTRUCTIONSANDTHESOFTWARE

Thissoftwareandtheassociatedinstructionsmaybecopiedaslongastheyareusedforeducationalpurposes.Allcopiedmaterialsmustdisplaythefollowingcopyrightnotices.

Copyright1992-2004RonaldH.BallouAllrightsreserved.

RonaldH.Ballouoffersthissoftwareforeducationalpurposesonlyanddoesnotwarrantthesoftwaretobefitforanyparticularapplication.TheuseragreestoreleaseRonaldH.Balloufromallliabilities,expenses,claims,actions,and/ordamagesofanykindarisingdirectlyorindirectlyoutoftheuseofthesecomputerprograms,theperformanceornonperformanceofsuchcomputerprograms,andthebreachofanyexpressedorimpliedwarrantiesarisinginconnectionwiththeiruse.Iftheseconditionsarenotacceptable,thesoftwareshouldbereturnedtoRonaldH.Ballou.

ProfessorRonaldH.Ballou

WeatherheadSchoolofManagement

CaseWesternReserveUniversity

Cleveland,OH44106USA

Tel:

（216）368-3808

Fax:

（216）368-6250

E-mail:

Ronald.Ballou@Weatherhead.CWRU.edu

Uptodateinformationaboutthesoftwaremaybefoundatwww.PrenH

INSTRUCTIONSFOREXPONENTIALSMOOTHING

ANDTIMESERIESDECOMPOSITIONFORECASTING

FORECAST

FORECASTiscomputersoftwarethatforecastsfromtimeseriesdatabymeansofexponentialsmoothingand/ortimeseriesdecompositionmethods.Inlogistics/SC,suchtimeseriesmaybeproductsales,leadtimes,pricespaidforgoods,orshipments.Thephilosophyoftimeseriesforecastingistoprojectanhistoricalpatternofthedataovertime,and,ifpresent,accountfortrendandseasonality.Exponentialsmoothingisamovingaverageapproachthatprojectstheaverageofthemostrecentdataandadaptstheforecasttochangingdataastheyoccur.Ontheotherhand,thetimeseriesdecompositionapproachrecognizesthatmajorreasonsforvariationindataovertimeareduetotrendandseasonalcomponents.Eachoftheseisestimatedandcombinedtoproduceaforecast.ForbackgroundinformationontheforecastmodelsusedinFORECAST,seeChapter8oftheBusinessLogistics/SupplyChainManagement5etextbook.

TorunFORECAST,selecttheappropriatemodulefromtheLOGWAREmastermenu.Openanexistingfileorselectanewone.Prepareorchangethedatabase.Selecttheappropriatemodeltype,whichmaybeeithersomeformofanexponentialmodel（Levelonly,Level-Trend,etc.）orthetimeseriesdecompositionmodel.ClickonSolvetogenerateaforecast.

INPUT

Theinputtobothforecastingmodulesconsistsofthetime-orderedseriesofdata,rankedfromthemosthistorictothemostrecentobservations,andvariousparametervaluesthatguidetheexecutionofthemodels.Thedimensionsofthemodelsallowobservationsforupto200periodsandaforecastofupto50periods.Bothmodeltypesrunfromthesamedatabasealthoughsomeoftheparametersarenotusedinthetimeseriesdecompositionmodel.

ParametersandLabels

Thisportionofthescreensetstheparametersforboththeexponentialsmoothingandtimeseriesdecompositionmodels.Theseguidetheoverallactionofthemodels.Considereachelementonthisscreen.

Problemlabel.Thisisalabelgiventotheproblemyouaresolving.WARNING:

Donotusecommas（,）ordoublequotationmarks（"）inthelabelsincethiswillcauseanerrorinreadingthedatafile.

Numberofdatapoints.Specifythenumberofdataperiodsinthetimeseries.Upto200pointsareallowed.Besurethatthenumberofpointsspecifiedherematchesthenumberofdatapointsactuallyenteredinthetimeseries.

Initializationperiod.Theinitializationperiodisthenumberoftheoldestdatapointsusedtodeterminestartingvaluesfortheexponentialsmoothingmodel.Aminimumof3periodsofdatashouldbedeclaredforthispurpose.Ifaseasonalmodelistobeused,atleastthenumberofperiodsinonefullseasonalcyclemustbespecified.

Errorstatistics.Thenumberofdataperiodsneededtocomputeforecasterrorstatisticsisreferredtoasthevalidationperiod.Theseerrorstatisticsarethemeanabsolutedeviation（MAD）,thebias（BIAS）,andtherootmeansquareerror（RMSE）.ThevalidationperiodisthelastNperiodsofdata.Enoughdatapointsshouldbeusedfromthisvalidationperiodtostrikeareasonableaverageforthesestatistics.

MADisdefinedastheaverageoftheabsolutedifferencesbetweentheactualvaluesandtheforecastvaluesforthevalidationperiod.BIASistheaverageofthedifferencesbetweenactualandforecastvaluesforthevalidationperiod.RMSEissquarerootoftheaverageofthesquareddifferencesbetweentheactualandtheforecastvaluesforthevalidationperiod.

Modeltype.Selectingthemodeltypereferstotheexponentialsmoothingmodelorthetimeseriesdecompositionmodel.Therearefourvariationsoftheexponentialsmoothingmodelforbestrepresentingthecharacterofthetimeseries.ThesearetheLevelonly,Level-Trend,Level-Seasonal,andLevel-Trend-Seasonal.Selectthetypethatbestrepresentsthedata.Alternately,selectthetimeseriesdecompositionmodel.Smoothingconstantsearch.Whenoneoftheformsoftheexponentialmodelisselected,indicatewhetherasearchforthesmoothingconstantsistobeperformedusingFORECAST.Ifnot,thesmoothingconstantsforthemodeltypeselectedmustbespecified.

IfFORECASTistosearch