通行能力瓶颈影响下的枢纽交叉口信号自适应控制策略及应用
Abstract
Duetotheconstraintsoftheroadenvironment,thelimitedconstructionconditions,the
differentsettingsoftheentrancesandexitsoftheroadsectionandtheinfluenceof
multi-channelconcurrentflowandotherevents,therewillbebottlenecksectionsoftraffic
congestionintherelevantareasofthecity.Thesebottlenecksectionswillhaveabadimpacton
urbantraffic.Theimplementationofeffectivecontrolstrategiescansignificantlyimprovethe
congestionofurbantrafficbottleneckroads.Aimingatthedynamiccharacteristicsofbottleneck
trafficnetwork,thispaperdesignsadynamicoptimizationmethodofregionaltrafficsignal
basedoniterativelearning,whichcansolvetheproblemoftrafficcongestionunderbottleneck
section.ThesimulationtestenvironmentisbuiltbyVISSIMsimulationplatformtoverifythe
effectivenessoftheproposedalgorithm,andthelandingverificationofShengliRoadand
TushanRoadinBengbuCityiscarriedout.Themainworkofthispaperisasfollows:
1.Firstly,accordingtothedifferentconfidenceofdifferentdatasourcesinthedetectionof
directtrafficflowstateindicatorssuchastrafficvolume,queuelengthandoccupancy,the
methodofmissingdataandmulti-sourcedatafusionisstudied,andthecompletelevelindexof
fusiondataisestablished.Secondly,throughthestudyofcomplexnetworktheory,thetraffic
controlsub-regiondivisionmethodsuitablefordifferentcontrolstrategiesisobtained.Finally,
duetotheroadtrafficconditionsaresusceptibletomoreaccidentalfactors,themissing
informationgeneratedbythedatatomakeup.
2.Consideringthetrafficcharacteristicsofthebottlenecksection,thedriversdrivingmode,
theviolationdrivingbehaviorofthebottlenecksectionandthevehiclelanechangeareanalyzed
andstudied,andatrafficsub-areaiterativelearningcontrolmetho