评估以前的交通需求模型预测的准确性.pptx

1、Assessing the Accuracy of Previous Travel Demand Model Forecasts,Robert G.Schiffer,AICPPRESIDENTFuturePlan Consulting,LLCNational Practice Leader,Travel Demand ForecastingMETRO ANALYTICS,Presented to 2023 TRB Innovations in Travel Analysis and Planning Conference,Overview:Assessing the Accuracy of P

2、revious Travel Demand Model Forecasts,BackdropModeling for Long-Range Transportation PlansStatewide Model ExampleSubarea Model ExampleCorridor StudiesSummary&ConclusionsAcknowledgments,Backdrop,Unlike other studies,I focused on 15-to 25-year model forecasts for horizon years that have already passed

3、Primary comparisons were highway assignment forecasts from 1980s and 1990s vs.more recent traffic countsPersonal curiosity of how some of my forecasts panned outFreedom of my own consultancy freed me up to research thisDue to the age of forecasts,had to rely on printed documents,Modeling for Long-Ra

4、nge Transportation Plans,2010 Brevard MPO LRTPBrevard Area Transportation Study 2010 Plan Update,August 19882010 UTPS screenline forecasts vs.2010 FDOT AADT traffic counts2010 E+C model forecast2010 model forecast with prior planAADTs had to be interpolated on 3 out of 36 links comprising screenline

5、sSum of 8 screenlinesE+C forecast+34%over count2005 Plan forecast+30%over count,Modeling for Long-Range Transportation Plans(Contd),2015 Pensacola MPO LRTPPensacola Urbanized Area Transportation Study 2015 Plan Update,May 19892015 UTPS screenline/bridge forecasts vs.2015 FDOT AADT traffic countsScre

6、enline 1=Intracoastal WaterwayScreenline 2=Pensacola BayScreenline 1=13%over countsScreenline 2=-31%to+58%(averaged over 3 build bridge alternatives)depending on which bridges are includedIf volumes on new bridge(never built)are excluded,assignment is-10%vs.count,Statewide Model example,Official 202