简介：Theintegralpartoftheoptimalvelocitycar-followingmodelsistheoptimalvelocityfunction(OVF),whichcanbederivedfrommeasuredvelocity-spacingdata.ThispaperdiscussesseveralcharacteristicsoftheOVFandpresentsregressionanalysisontwoclassicaldatasets,theLincolnandHollandtunnels,withdifferentpossibleOVFs.ThenumericalsimulationoftheformationoftrafficcongestionisconductedwiththreedifferentheuristicOVFs,demonstratingthatthesefunctionsgiveresultssimilartothoseofthefamousBandoOVF(Bandoetal.,1995).Alsoanalternativemethodispresentfordeterminingthesensitivityandmodelparametersbasedonasinglecardrivingtoafixedbarrier.

简介：Weproposeanewcellularautomation(CA)trafficmodelthatisbasedonthecar-followingmodel.Aclassofdrivingstrategiesisusedinthecar-followingmodelinsteadoftheaccelerationintheNaSchtrafficmodel.Inourmodel,somerealisticdriverbehaviouranddetailedvehiclecharacteristicshavebeentakenintoaccount,suchasdistance-headwayandsafedistance,etc.Thesimulationresultsshowthatourmodelcanexhibitsometrafficflowstatesthathavebeenobservedintherealtraffic,andbothofthemaximumfluxandthecriticaldensityareveryclosetotherealmeasurement.Moreover,itiseasytoextendourmethodtomulti-lanetraffic.