供应链需求预测--rickyblcu

DemandForecastinginaSupplyChain7-12大綱預測在供應鏈的角色預測的特性主要企業預測項目預測的方法與組成時間序列預測預測誤差的衡量指標執行預測的建議CPFR3預測在供應鏈的角色ThebasisforallstrategicandplanningdecisionsinasupplychainExamples:Production:scheduling,inventory,aggregateplanningMarketing:salesforceallocation,promotions,newproductionintroductionFinance:plant/equipmentinvestment,budgetaryplanningPersonnel:workforceplanning,hiring,layoffsAllofthesedecisionsareinterrelated4預測的特性Forecastsarealwayswrong.Shouldincludeexpectedvalueandmeasureoferror.Long-termforecastsarelessaccuratethanshort-termforecasts(forecasthorizonisimportant)Aggregateforecastsaremoreaccuratethandisaggregateforecasts5主要企業預測項目市場需求量母體數預測單位需求量預測驅動變數預測市場佔有率預測企業銷售量預測單價預測生命週期預測6預測的方法主觀法(subjectivemethods)預測人員依個人主觀的判斷進行預測常應用在缺乏歷史資料時透過專家進行主觀預測草根法GrassrootsBottomup市調法MarketresearchLong-rangeNewproductsales歷史類推法Historicalanalogy類似的產品經驗類推DelphiMethod以問卷方式蒐集專家意見以進行預測經由問卷溝通，專家間無直接互動以避免主控性以統計量收斂為停止指標7預測的方法客觀法(objectivemethods)以歷史資料為基礎進行預測TimeSeries(外插法)假設過去之需求資料是未來需求良好指標下，使用歷史資料進行預測，適合當需求環境穩定、無劇烈變動時進行Causal(因果關係法)假設需求與環境中某些因素是高度相關，藉由發現需求與環境因素的相關性去估計未來的需求TransferFunctionModel(轉換函數模式)結合TimeSeries與Causal兩者，經由解釋變數與應變數之歷史資料產生轉換函數，再將解釋變數之預測值代入轉換函數產生應變數之預測值ARIMAT、SARIMAT8需求資料的組成Observeddemand(O)=Systematiccomponent(S)+Randomcomponent(R)Level(currentdeseasonalizeddemand)Trend(growthordeclineindemand)Seasonality(predictableseasonalfluctuation)•Systematiccomponent:Expectedvalueofdemand•Randomcomponent:Thepartoftheforecastthatdeviatesfromthesystematiccomponent•Forecasterror:differencebetweenforecastandactualdemand9需求資料組成的關係類型相乘系統部分＝水準×趨勢×季節性因素相加系統部分＝水準＋趨勢＋季節性因素混合系統部分＝（水準＋趨勢）×季節性因素10時間序列預測QuarterDemandDtII,20038000III,200313000IV,200323000I,200434000II,200410000III,200418000IV,200423000I,200538000II,200512000III,200513000IV,200532000I,200641000Forecastdemandforthenextfourquarters.11時間序列預測010,00020,00030,00040,00050,00003,203,303,404,104,204,304,405,105,205,305,406,112預測的方法StaticAdaptiveMovingaverageSimpleexponentialsmoothingHolt’smodel(withtrend)Winter’smodel(withtrendandseasonality)13預測的流程UnderstandtheobjectivesofforecastingIntegratedemandplanningandforecastingIdentifymajorfactorsthatinfluencethedemandforecastUnderstandandidentifycustomersegmentsDeterminetheappropriateforecastingtechniqueEstablishperformanceanderrormeasuresfortheforecast14時間序列預測GoalistopredictsystematiccomponentofdemandMultiplicative:(level)(trend)(seasonalfactor)Additive:level+trend+seasonalfactorMixed:(level+trend)(seasonalfactor)StaticmethodsAdaptiveforecasting15靜態法Assumeamixedmodel:Systematiccomponent=(level+trend)(seasonalfactor)Ft+l=[L+(t+l)T]St+l=forecastinperiodtfordemandinperiodt+lL=estimateoflevelforperiod0T=estimateoftrendSt=estimateofseasonalfactorforperiodtDt=actualdemandinperiodtFt=forecastofdemandinperiodt16靜態法EstimatinglevelandtrendEstimatingseasonalfactors17範例資料分析產品之需求有季節性的現象每年度之第二季為全年度需求最低之時需求皆是從每年度之第二季遞增至下年度之第一季此需求變化呈現週期現象，每個週期為一年三個週期的需求水準有逐漸上升的趨勢18LevelandTrend因子的估計Beforeestimatinglevelandtrend,demanddatamustbedeseasonalizedDeseasonalizeddemand=demandthatwouldhavebeenobservedintheabsenceofseasonalfluctuationsPeriodicity(p)thenumberofperiodsafterwhichtheseasonalcyclerepeatsitselffordemandatTahoeSalt(Table7.1,Figure7.1)p=419去季節因子的需求資料[Dt-(p/2)+Dt+(p/2)+S2Di]/2pforpevenDt=(sumisfromi=t+1-(p/2)tot+1+(p/2))SDi/pforpodd(sumisfromi=t-(p/2)tot+(p/2)),p/2truncatedtolowerinteger20去季節因子的需求資料Fortheexample,p=4isevenFort=3:D3={D1+D5+Sum(i=2to4)[2Di]}/8={8000+10000+[(2)(13000)+(2)(23000)+(2)(34000)]}/8=19750D4={D2+D6+Sum(i=3to5)[2Di]}/8={13000+18000+[(2)(23000)+(2)(34000)+(2)(10000)]/8=2062521去季節因子的需求資料ThenincludetrendDt=L+tTwhereDt=deseasonalizeddemandinperiodtL=level(deseasonalizeddemandatperiod0)T=trend(rateofgrowthofdeseasonalizeddemand)Trendisdeterminedbylinearregressionusingdeseasonalizeddemandasthedependentvariableandperiodastheindependentvariable(canbedoneinExcel)Intheexample,L=18,439andT=52422需求的時間序列(Figure7.3)01000020000300004000050000123456789101112PeriodDemandDtDt-bar23估計季節因子UsethepreviousequationtocalculatedeseasonalizeddemandforeachperiodSt=Dt/Dt=seasonalfactorforperiodtIntheexample,D2=18439+(524)(2)=19487D2=13000S2=13000/19487=0.67Theseasonalfactorsfortheotherperiodsarecalculatedinthesamemanner24估計季節因子(Fig.7.4)tDtDt-barS-bar18000189630.42=8000/18963213000194870.67=13000/19487323000200111.15=23000/20011434000205351.66=34000/20535510000210590.47=10000/21059618000215830.83=18000/21583723000221071.04=23000/22107838000226311.68=38000/22631912000231550.52=12000/231551013000236790.55=13000/236791132000242031.32=32000/242031241000247271.66=41000/2472725估計季節因子Theoverallseasonalfactorfora“season”isthenobtainedbyaveragingallofthefactorsfora“season”Iftherearerseasonalcycles,forallperiodsoftheformpt+i,1ip,theseasonalfactorforseasoniisSi=[Sum(j=0tor-1)Sjp+i]/rIntheexample,thereare3seasonalcyclesinthedataandp=4,soS1=(0.42+0.47+0.52)/3=0.47S2=(0.67+0.83+0.55)/3=0.68S3=(1.15+1.04+1.32)/3=1.17S4=(1.66+1.68+1.66)/3=1.6726預測未來需求Usingtheoriginalequation,wecanforecastthenextfourp