mfg 580制造模拟级测试审查march MFG321.pptx

MFG580ManufacturingSimulation

Class11March21,2017

ReviewofTest2MFG580Winter2017March21,2017

ProjectModelFlowchartDueTodayMFG580Winter2017March21,2017HandinyourprintedFlowchartMakesureitincludesallassumptionsHaveawelldefinedscopeofthemodel

Chapter9ModelingRandomnessMFG580Winter2017March21,2017

ModelingRandomnessSimulationmodelbuildersneedabasicunderstandingofstatisticsTohaveavalidmodelyouneedtomodelthedatacorrectlyStochasticdistributionsaredatamodelsthatdefineinputs,processtimes,downtimesandrepairtimesUsingjustaverageswillvoidamodelbecauseitwillgivethesamewronganswereverytime.MFG580Winter2017March21,2017

Samplingfromthedistributions(normal)MFG580Winter2017March21,20170zmxP(Xx)=P(Zz)0zmxprobabilitydensityfunction,f(x)10cumulativedistributionfunction,F(x)P(Xx)=P(Zz)

StandardNormalTableMFG580Winter2017March21,2017

Samplingfromthenormal(manualmethod)Toobtainarandomsamplefromanormaldistribution,wecouldrandomlyselectavaluefromastandardnormaltable(randomzi).Therandomvalueusedtoenterthenormaltablecouldcomefromarandomnumbertable(ui).0.6730600.450.6736

GeneratingcontinuousrandomvariatesMFG580Winter2017March21,2017Ingeneral,forcontinuousdistributions,0zmxP(Xx)=P(Zz)probabilitydensityfunction,f(x)0zmx10P(Xx)=P(Zz)cumulativedistributionfunction,F(x)

GeneratinguniformlydistributedrandomvariatesMFG580Winter2017March21,2017Forexample,ifservicestimesareequallylikelybetween10and30minutes,i.e.a=10andb=30,andifu1=0.67306,thenx1=10+20(0.67306)=23.46

GeneratingexponentiallydistributedrandomvariatesMFG580Winter2017March21,2017Forexample,ifservicetimesareexponentiallydistributedwithameanof20minutes,i.e.b=20,andifu1=0.67306,thenx1=-20*ln(1-0.67306)=22.36

Algorithmsforgenerat