基于數(shù)據(jù)挖掘的交通流狀態(tài)辨識算法研究.pdf

1、山東理工大學(xué)碩士學(xué)位論文基于數(shù)據(jù)挖掘的交通流狀態(tài)辨識算法研究姓名：張敬磊申請學(xué)位級別：碩士專業(yè)：載運工具運用工程指導(dǎo)教師：王曉原20060518山東理工大學(xué)碩士學(xué)位論文AbstractAbstractTrafficflowconditionrecognitionisoneoftheimportantissuesofIntelligentTransportationSystemsOTS)especiallyforitsAdvancedTr

2、affIcManagementSystems(ATMS)andAdvancedTravelerInformationSystems(ATIS)researchTheobjectiveoftrafficflowconditionrecognitionistecognizetheconditionoftrafficflowtimelyandaccuratelytakeeffectivetrafficincidentmanagementtra

3、fficcontrolandtmmcflowguidancemeasurestocleanouttrafficincidentstimelyInthiswaythelossofcasualtieswillbeloweredandthesecondincidentswillbeavoidedInaddition，thecontrolmeasuresoftransportationdemandcarlbetakeneffectivelyan

4、dtherouterecommendationscanbeavailabletotravelersSOastolessentrafficpressure，economizeenergyandreducepollutionAccordingtothedemandsofdynamictrafficmanagementtrafficincidentmanagementandurbantrafficflowguidance，thetheoryo

5、ftrafficflowconditionrecognition，includingtrafficflowshort—termforecastingandrealtimedetectionisstudieddeeplyinthispaperAnarchitectureoftrafficconditionrecognitionsystemisputforwardbasedonthebasicthoughtofdynamictrafficm

6、anagementandthearchitectureoftrafficincidentmanagementsystem，Twotrafficflowconditionrecognitionalgorithms，namelythetrafficflowshort—termforecastingalgorithmbasedonnonparameterregressionsplinefittingandthetrafficincidentd

7、etectionalgorithmbasedonwaveletmultirasolutionanalysis，arepresentedaccordingtothecharacteristicoftrafficflowTheparametersofalgorithmsarecalibratedandtwoofflinetestsaremadeaccordingtothedataundertheconditionofincidentandn

8、onincidentobtainedfromTransportationResearchGroup(TRG)ofUniversityofSouthamptonofUKandthesimulationTheresultsoftestsprovedthatbothalgorithmsperformbetterintrafficflowconditionrecognitionKeywords：trafficflowtheoryconditio