バスプローブデータを用いた豊田市の道路渋滞分析に関する研究
45/48

39 13Conclusions•Summaryofcontents,majorfindingsandfuturetasksarelistedasfollows.SummaryofcontentsThisstudyappliedtheGaussianmixturemodeltofitthetraveltimeofaroadbottleneckinToyotaCity.InordertoconsidertheimpactofTMCworkers`commutingbehaviorandthatofthepeakperiod,sampledatawerecategorizedinto4groupsworkingdays(peak),workingdays(off-peak),non-workingdays(peak)andnon-workingdays(off-peak).WeappliedtheGaussianmixturedistributiontofitthetraveltimedistributionforfourgroups.MajorfindingstheGaussianmixturedistributionfittingfortraveltimeduringtheoff-peakperiodinworkingdaysandthatinnon-workingoneshaveasmallermixtureproportion,whichindicatedalowprobabilityofthecongestedstate.Gaussianmixturedistributionsappliedforpeakandoff-peakperiodinnon-workingdaysofTMChaveabimodalfeature.TheGaussianmixturedistributionoutperformstheGaussiandistributionin4categorizedgroups.FuturetasksWeincreasethesampledatatoinvestigatethereliabilityoftraveltimeinthisroadsection.TheimpactofweatherconditionisconsideredintheproposedGaussianmixturedistribution.14Thank you for your attention!Email: yang@ttri.or.jp

元のページ  ../index.html#45

このブックを見る