evolution of the greater vancouver transportation modelWord格式.docx
《evolution of the greater vancouver transportation modelWord格式.docx》由会员分享,可在线阅读,更多相关《evolution of the greater vancouver transportation modelWord格式.docx(17页珍藏版)》请在冰豆网上搜索。
13thAnnualInternationalEMME/2User’sConference
Houston,Texas
October28-30,1998
by
KarolyKrajczar
GreaterVancouverRegionalDistrict
Burnaby,B.C.
1.Introduction
ThispaperprovidesanoverviewoftheevolutionoftheGreaterVancouverRegionalDistrict(GVRD)TransportationModelandfocusesontherecentenhancementsthatwerecompletedin1998aspartoftheGVRD’sTransportationModelEnhancementProgram.Thispaperdescribeshowtheimprovementswereidentifiedandprioritizedandprovidesasummaryofthemodelenhancements.Theotherpapersinthissessionwillelaborateonthreeofthemajorenhancements(Transitcaptivity,HOV,ParkandRide).
2.EvolutionoftheGreaterVancouverTransportationModel
The1970’s
TheGVRDhasalonghistoryinthedevelopmentandapplicationoftransportationmodelsdatingbacktothemid-1970’s.In1975,theGVRD’sfirstcomputer-basedtransportationmodelwasdevelopedinconjunctionwiththeUniversityofBritishColumbia(UBC)ComputerScienceDepartment.Themodelconsistedof125trafficzones,aroadandtransitnetworkandalgorithmstoimplementatraditionalfour-steptransportationplanningmodelforthemorningpeakperiod.SpecializedsoftwarewaswrittentoimplementthemodelontheUBCmainframecomputer.Themodelhadlimitedgraphicalcapabilities,withoutputconsistingprimarilyoftablesoflinkvolumesandotherattributedata.
In1979,themodelunderwentasignificantre-calibrationinpreparationfortheGVRDRapidTransitProjectwhichwasacomprehensivesetofstudiesoffuturerapidtransitlinesandtechnologyoptionsfortheGVRD.Atthistime,thetrafficzonesystemwasexpandedto152zones,thetransitassignmentalgorithmwasenhancedandthefour-stepmodelwasre-calibrated.Themodelwascalibratedusingtraveldatafromthe1978NorthVancouverOrigin-DestinationSurveyandaseriesofstudiesbytheCityofVancouverincludingthe1975DowntownVancouverODSurvey,the1976TransitStudyandthe1977CordonSurvey.
The1980’s
In1983,theGVRDconductedacomparativereviewoftransportationmodellingsoftware.TheGVRD’sexistingsoftware(developedin-house)wascomparedwiththeFederalHighwayAdministration’sUrbanTransportationPlanningSystem(UTPS)andINROConsultantsEMME/2.ThereviewcommitteeselectedEMME/2andtheGVRDbecamethesecondNorthAmericansitetoobtainEMME/2(Portland,Oregonwasthefirstsite).
In1984themodelwasre-calibratedinconjunctionwiththeimplementationofEMME/2.Thetrafficzonesystemwasexpandedto370zones,roadandtransitnetworkswereupdatedandthefour-stepmodelwascalibratedto1981Censusdatawhichcontainedtheplaceofwork-placeofresidencesurvey.Significantchangestothemodelstructureincludedtheintroductionoftime-spaceequationsforautoworktripsandcombinedimpedanceforworktripdistribution.Atthistime,theGVRDpurchasedthePixelcomputer(a32-bitmicrocomputer)toreducethehighcomputingcostsassociatedwithrunningthemodelatUBC.InthosedaysthemacrolanguagewasnotavailableinEMME/2andthemodelequationswereentereddirectlyintothematrixcalculator(eachmatrixcalculationtookabout30minutesanda15iterationassignmentrequiredapproximately14hoursofcomputingtime).
In1985,acomprehensivesetof24hourtravelsurveyswasconductedtoprovideadescriptionoftravelpatternsintheGVRDpriortoSkyTrain.Theprojectincludeda25,000householdtelephoneinterviewsurvey(afivepercentsampleofthe522,000householdsintheVancouverCMA),a1,700householdtripdiarysurvey(a0.3percentsample),trafficscreenlinecounts,anexternalcordonsurveyandataxiandbusinesstravelsurvey.Datafromthehouseholdsurveyiscontainedinaseriesofthreereports.
Between1986and1987,themodelwasre-calibratedtothe1985surveydata.Thetrafficzonesystemwasincreasedslightlyto380zones,networkswereupdatedandthefourstepmodelwasre-calibrated.Themodelwasthenvalidatedtopost-Skytrainconditionsin1987.Althoughseveralmajornewfacilitieshadbeenintroducedtotheregionbetween1985and1987(e.g.Skytrain,AlexFraserBridge,theEast-WestFreewayinRichmond),themodelcomparedcloselywith1987screenlinecounts.
Oneofthefirstapplicationsofthenewregionalmodelwastoexaminetheimpactsoflong-termgrowthonmajortransportationfacilitiesbytheCityofRichmond.Thefindingsofthisstudyshowedthatwithcurrenttrendslandusemanyofthemajorrivercrossingswouldbeatcapacityby1996.ThisstudyresultedinaninitiativebyRichmondcounciltorequestthattheprovinceleadacomprehensiveinter-agencystudyonlanduseandtransportationfortheGVRD.
Atthesametime,manyoftheGVRDmunicipalitiesbegantodeveloptheirownsub-areatransportationmodelsthatcontainedmoredetailedzonesystemsandnetworkstodealwithmunicipaltransportationplanning.Burnaby,Surrey,Coquitlam,RichmondandNorthVancouverweresomeofthefirstmunicipalitieswiththeirownsub-areamodels.ManyofthesemodelswereafternoonpeakperiodmodelsasopposedtotheGVRDmorningpeakperiodmodel.Thesub-areamodelswererunbydialing-intotheGVRDPixelcomputer.
In1988,theCityofVancouverandGVRDconductedaregionaltruckingsurvey.Informationfromthissurveywasusedtocalibratea24hourtruckmodel.Thismodelprovidedinformationontwocategoriesoftruckmovements-lighttruckswithaGVWof4,500-20,000kgandheavytruckswithaGVWover20,000kg.
DuetoincreaseddemandsonEMME/2,adecisionwasmadein1989totransferfromthePixelcomputertotheGVRDVAXcomputer.AmultipleEMME/2VAXlicensewaspurchasedfromINROConsultantsInc.whichprovidedmunicipalitiesandgovernmentagencieswithbetteraccesstotheirmodels.Theresultwasasignificantincreaseinspeedandexpandedusebyvariousagencies.
Amajor1989studytomakeextensiveuseoftheregionalmodelwastheFreedomtoMove-GreaterVancouverTransportationTaskForcestudy.Thisstudy,whichwassponsoredbytheProvincialGovernmentandtheGVRD,producedanintegratedtransportationplanfortheGVRDtotheyear2001.
The1990s
In1991,theregionalmodelandtruckmodelwerevalidatedtocurrentconditions.Thevalidationprocessincludedexpandingthetrafficzonesystemto445zones,updatingthelanduseinputsandnetworksandsomemodificationstothefour-stepmodelandsignificantchangestothetruckmodelstructure.Duringthevalidationexercise,theGVRDdecidedtotransferEMME/2fromtheVAXtothePC.Again,thereasonbeinglowercosts,betteraccesstothemodelsandfastercomputingtimeforusers.
ThisworkwasdoneinpreparationforTransport2021whichwastheregion’sfirstlong-rangetransportationplanwithassociatedpoliciesonlanduse,transportationdemandmanagement(TDM)andinvestmentprioritiesfortransitandroad.Transport2021wasatwoyearprojectthatmadeextensiveuseoftheregionalmodelfortestinglandusescenarios,TDMstrategiesandroadandtransitnetworks.
In1992theGVRD,BCTransitandtheMinistryofTransportationandHighwaysjointlyfundedamajorhouseholdtravelsurveyforthemorningpeakperiod(6:
00-9:
00a.m.).Thistelephonesurveycollecteddatafrom15,000households(atwopercentsampleofthe690,000householdsintheLowerMainland-LionsBaytoChilliwack)withinformationonmorethan25,000morningtrips.Oneoftheprimaryobjectivesofthissurveywastocollectdatarequiredtore-calibratetheregionalmodeltocurrentmorningconditions.Thiswasfollowedbya24hourtripdiarysurveyin1994.Thiswasamail-backsurveythatcollectedinformationfrom1,600households(0.2percentsampleofthe740,000householdsintheLowerMainland)withinformationonmorethan11,000tripsthroughouttheday.Thetripdiarywasdesignedtocollectbaselineinformationforregionalplanningpurposes,aswellasforcalibratinganafternoonpeakperiodtransportationmodel.
Inlate1994,theProvince,GVRDandBCTransitembarkedonafollow-upstudytoTransport2021calledtheSouthCoastTransportationSystemPlan.ThisstudywasintendedtoprovideamoredetaileddescriptionoftheinfrastructurerequirementsidentifiedinTransport2021.Ataboutthesametime,BCTransitwasinvolvedinthedevelopmentoftheir10YearPlanandinaMultipleAccountEvaluationofRapidTransitOptionsfortheGVRD.BothofthesestudiesmadeextensiveuseoftheregionalmodelandtheGVRDtruckmodel.
In1995,theGVRDconductedare-calibrationofthemodeltothe1992TravelSurvey.Insummary,thetrafficzonesystemwasexpandedfrom445zonesto652zones,theroadnetworkwasexpandedfrom7,400to13,000links,thetransitnetworkwasupdatedandnewintersection-basedvolume-delayfunctionswereintroduced,replacingtheoldBureauofPublicRoads(BPR)functions.
In1996,theGVRDestablishedtheTransportationModelEnhancementProgramtoaddresstheevolvingneedsofthemodel-usersintheregion.Thefirststepinvolvedupdatingthemodelto1996conditions,developingapreliminaryPMTransportationModel,collectingadditionaldataandidentifyingfuturemodelenhancements.Theidentificationoffuturemodelenhancementsinvolvedinputfromconsultants,municipalandprovincialstaff.Thisfeedbackwascategorizedandprioritized(dependinguponneed,dataavailability,difficulty,cost,inter-dependencies,etc.).Theresultwasashort,mediumandlong-termmodellingresearchprogramthatfocusesonmodelenhancement,modelexpansion,usersupportanddatacollection.
Theshort-termworkprogramwascompletedin1998andinvol