风险管理软件Crystal Ball使用指导.docx

编号：时间：2021年x月x日书山有路勤为径学海无涯苦作舟页码：Monte-CarloSimulationwithCrystalBall®TorunasimulationusingCrystalBall®:1.SetupSpreadsheetBuildaspreadsheetthatwillcalculatetheperformancemeasure(e.g.profit)intermsoftheinputs(randomornot).Forrandominputsjustenteranynumber.2.DefineAssumptions—i.e.randomvariablesDefinewhichcellsarerandomandwhatdistributiontheyshouldfollow.3.DefineForecast—i.e.outputorperformancemeasureDefinewhichcell(s)youareinterestedinforecasting(typicallytheperformancemeasuree.g.profit).4.ChooseNumberofTrialsSelectthenumberoftrials.IfyouwouldlaterliketogeneratetheSensitivityAnalysischartchoose“SensitivityAnalysis”underOptionsinRunPreferences.5.RunSimulationRunthesimulation.Ifyouwouldliketochangeparametersandre-runthesimulationyoushould“reset”thesimulation(clickonthe“ResetSimulation”buttononthetoolbarorintheRunmenu)first.6.ViewResultsTheforecastwindowshowingtheresultsofthesimulationappearsautomaticallyafter(orduring)thesimulation.Manydifferentresultsareavailable(frequencychartcumulativechartstatisticspercentilessensitivityanalysisandtrendchart).Theresultscanbecopiedintotheworksheet.CrystalBallToolbar:DefineDefineRunStartResetForecastTrendAssumptionsForecastPreferencesSimulationSimulationWindowChartWaltonBookstoreSimulationwithCrystalBall®RecalltheWaltonBookstoreexample:ItisAugustandtheymustdecidehowmanyofnextyear’snaturecalendarstoorder.Eachcalendarcoststhebookstore$7.50andissoldfor$10.AfterFebruaryallunsoldcalendarsarereturnedtothepublisherforarefundof$2.50percalendar.SupposeWaltonpredictsdemandwillbesomewherebetween100and300(discreteuniform).Demand=d~Uniform[100300]OrderQuantity=Q(decisionvariable)Revenue=$10*Min(Qd)Cost=$7.50*QRefund=$2.50*Max(Q–d0)Profit=Revenue–Cost+RefundStep#1