文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 自然科学 > 数学 > 2104数学建模美赛A 交通流 安全性 参考资料 交通量
侵权投诉

2104数学建模美赛A 交通流 安全性 参考资料 交通量

2104数学建模美赛A 交通流 安全性 参考资料 交通量

RelationshipsAmongUrbanFreewayAccidents,Traf c

Flow,Weather,andLightingConditions

ThomasF.Golob1andWilfredW.Recker2

Abstract:LinearandnonlinearmultivariatestatisticalanalysesareappliedtodeterminehowthetypesofaccidentsthatoccuronheavilyusedfreewaysinSouthernCaliforniaarerelatedbothtothe owoftraf candtoweatherandambientlightingconditions.Traf c owismeasuredintermsoftimeseriesof30-sobservationsfrominductiveloopdetectorsinthevicinityoftheaccidentpriortothetimeofitsoccurrence.Resultsindicatethatthetypeofcollisionisstronglyrelatedtomediantraf cspeedandtotemporalvariationsinspeedintheleftandinteriorlanes.Hit-objectcollisionsandcollisionsinvolvingmultiplevehiclesthatareassociatedwithlane-changemaneuversaremorelikelytooccuronwetroads,whilerear-endcollisionsaremorelikelytooccurondryroadsduringdaylight.Controllingforweatherandlightingconditions,thereisevidencethataccidentseverityisin uencedmorebyvolumethanbyspeed.DOI:10.1061/ ASCE 0733-947X 2003 129:4 342

CEDatabasesubjectheadings:Traf csafety;Traf c ow;Nonlinearanalysis;Urbanareas;Weather;Light.

Introduction

Ourobjectiveistoquantifyrelationshipsbetweenthetypeoftraf caccidents crashes thatoccuronurbanfreewaysandthecon gurationofthetraf c ow,whiletakingintoaccountweatherandlightingconditions.Wehavedataonapproximately1,200crashesthatoccurredonsixfreewayroutesinSouthernCaliforniaduring1998.Thesecrashesarecharacterizedby: a thetypeandlocationoftheprimarycollision, b themovementoftheinvolvedvehiclespriortocollision,thenumberofvehiclesinvolved,and c theaccidentseverity intermsofinjuryversuspropertydamageonly .Traf c owismeasuredintermsoftimeseriesof30-secondobservationsfrominductiveloopdetectorsinthevicinityoftheaccidentpriortothetimeofitsoccurrence.Thereisstrongempiricalevidenceoffunctionalrelationshipsbetweenaccidentratesandtraf c ow,conditionaluponroadwaycharacteristics e.g.Gwynn1967;CedarandLivneh1982;FrantzeskakisandLordanis1987;SullivanandHsu1988;HallandPendleton1989;GarberandGadiraju1990;Sullivan1990;StokesandMutabazi1996;Aljanahietal.1999;SandhuandAl-Kazily1996;ZhouandSisiopiku1997 .Aseriesofstudieshavealsodealtwithquanti cationofthesafetycomponentofthemar-ginalcostsofroadwayuse,asafunctionoftraf cspeed, ow,anddensity Vickery1969;Newberry1988;Jones-Lee1990;Vi-talianoandHeld1991;Jansson1994;O’Reillyetal.1994;Jo-hansson1996;SheferandRietveld1997;Dickerson,Peirson,and

Researcher,InstituteofTransportationStudies,Univ.ofCalifornia,Irvine,CA92687-3600.E-mail:tgolob@uci.edu2

Professor,Dept.ofCivilandEnvironmentalEngineeringandDirec-tor,InstituteofTransportationStudies,Univ.ofCalifornia,Irvine,CA92687-3600.E-mail:wwrecker@uci.edu

Note.DiscussionopenuntilDecember1,2003.Separatediscussionsmustbesubmittedforindividualpapers.Toextendtheclosingdatebyonemonth,awrittenrequestmustbe ledwiththeASCEManagingEditor.Themanuscriptforthispaperwassubmittedforreviewandpos-siblepublicationonNovember27,2001;approvedonJuly17,2002.ThispaperispartoftheJournalofTransportationEngineering,Vol.129,No.4,July1,2003.©ASCE,ISSN0733-947X/2003/4-342–353/$18.00.

1Vickerman2000 .Insituationsofcongested ow,moststudieshaveshownthatbothaccidentriskandthecostperaccident howeverde ned arenonlinearlyrelatedtotraf cspeedandden-sity.

Previousstudiestypicallyusedsuchaggregatetraf c owdataasdailyorhourlytraf ccountsandvolumetocapacitymeasures.Typesofcollisionsaregenerallynotdistinguished,exceptintermsofseverity.Speci cationofafunctionalrelationshipbe-tweenaccidentprobabilitiesandtheambienttraf c owatthetimeoftheaccidents,asmeasuredbycommonlyavailabletraf cmonitoringdevices,hasremainedelusive.MensahandHauer 1998 citetwokeyproblemsofaveragingassociatedwithusingaggregateddata—argumentaveragingandfunctionaveraging.Argumentaveragingrelatestotheuseofaveragetraf c owdata,ratherthandatameasuringtraf cconditionsatthetimeoftheaccident.Thesecondproblem,functionaveraging,iscausedbyusingthesamefunctionalrelationshipforalltypesofcollisionsunderallconditions e.g.,dayornight,dryorwetweather .Byusingtraf c owdataprevailingjustpriortothetimeofeachaccidentandbyincludingtheconditionsoftheaccidentintheanalysis,weareabletoavoidthesetwoproblems.

Ouranalysismethodinvolvesseveralsteps.First,toreducecollinearityinthetraf cdata,principalcomponentsanalysis PCA isperformedtoidentifyrelativelyindependentmeasure-mentsof owconditions.Nonlinear nonparametric canonicalcorrelationanalysis NLCCA isthenconductedwiththreesetsvariables.The rstsetiscomprisedofaseven-categorysegmen-tationvariablede ninglightingandweatherconditions;thesec-ondsetismadeupofaccidentcharacteristics collisiontype,location,andseverity;andthethirdsetmadeupofthetraf c owvariablesidenti edusingPCA.NLCCAisaformofcanonicalcorrelationanalysisinwhichcategoricalvariablesareoptimallyscaledasanintegralcomponentin ndinglinearcombinationsofvariableswiththehighestcorrelationsbetweenthem.Theseanalysesshowclearpatternsrelatingaccidentcharacteristicsandprevailing owconditions.

342/JOURNALOFTRANSPORTATIONENGINEERING©ASCE/JULY/AUGUST2003

第1页

你可能喜欢

  • 道路通行能力
  • 元胞自动机
  • 美国数学建模
  • 美赛论文
  • 数学建模美赛
  • 篮球教练
  • 交通安全

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email