驾驶动作对安全的影响英文版.docx

驾驶动作对安全的影响英文版.docx

《驾驶动作对安全的影响英文版.docx》由会员分享，可在线阅读，更多相关《驾驶动作对安全的影响英文版.docx（14页珍藏版）》请在冰豆网上搜索。

驾驶动作对安全的影响英文版

InfluenceofReachingActionsonDrivingPerformance

TakafumiAsao1,KentaroKotani1,KenHorii1

1Dept.ofMechanicalEngineering,FacultyofEngineeringScience,KansaiUniversity,Osaka,Japan

Abstract

Drivershavevariousadditionalactionswhiledriving,whichdonotbelongtooperationsofavehicle,forexample,eating,drinking,pressingsomebuttons,andexchangingcompactdiscsinacaraudiosystem.Theseactionscausetrafficaccidents,becausetheyusedtodistractdriver’svisionanddecisionmaking.Moreoverpostureofthedriverduringtheactionsinfluencesdrivingoperationseriously.Therefore,wefocusedonreachingactionswhiledriving.Inthispaper,weinvestigateddrivingperformancewithreachingtasksusingsimplifieddrivingsimulator.Furthermore,weexaminedinfluencesoftheadditionalactionsonthedrivingperformancewithparametersofadriver-vehiclemodel.

Keywords:

DrivingPosture;SteeringManeuverability;DriverModel

1、INTRODUCTION

Awiderangeofdriversupportsystemhasbeenproposedandanumberofoperationsforthesedevicesareincreased[1]-[3].Besideprimarydrivingtasks,wehavevariousactions,forexample,wepushsomebuttonsandlookforpositionsofthem.Theseactionsdistracthumandriving.

Adrivingrequirestorepeatcognition,decisionmaking,andoperation.Ifworkloadsofthecognitions,thedecisionmakings,andtheoperationsforadriverbecomegreater,theydistractthedriverfromdriving.Iftheworkloadsarehigherthanmanagementallimitofthedriver,trafficaccidentsmayoccurbecauseofhumanerrors.

Theworkloadsthatmayinfluenceonthedriverareclassifiedintothreeasfollows[4];1）visualworkloads,2）mentaworkloads,and3）executionworkloads.Thevisualworkloadsarecausedbyvisualsearching.Thementalworkloadsarepsychologicalworkloadscausedbythinking.Theexecutionworkloadsarecausedbyexecutionofanyoperationinacar.Whenthedriverswanttopushsomebuttons,theylookatthebuttonsandconsiderhowtooperatetheequipment,whichcontainsthevisualandmentalworkloads.Thentheyreachtheirhandtothebuttontopush.Itisimportanttoclarifyinfluenceoftheexecutionworkloadsonthedrivingbehaviors,becausetheexecutionworkloadsmayresultsinaddingthevisualandmentalworkloads.

Althoughtrains,airplanes,andshipsarecontrolledbyexpert

operators,thecarsareoperatedbyordinarypeople.Theordinarydriversmayhaveunexpectedlydangerousactionwhiledriving,forexample,pickingacompactdisc（CD）fromadashboardtoexchangeanotherCDinacaraudiosystem.Inthiscase,first,thedriversconfirmalocationoftheCDinthedashboardwiththevisualworkloads.Second,theyhavetoselectoneCDwiththementalworkloads,whichtheywanttolisten.Finally,theyreachtheirhandtothedashboardchangingtheirdrivingposturewiththeexecutionworkloads.Thereforeitisnecessarytoeducatethedriverstheriskofdoingsuchaction,andtodevelopsupportsystemsofdecreasingtheexecutionworkloadsaswell.

Severalresearcheshavedealtwithcomfortabledrivingpostureandmaneuverabilityinfluencedbythesittingpostures[5][6].Howeverdrivingperformanceaccordingtotheposturewasnotinvestigated.Therefore,inprecedingstudy,influencesofthereachingactionswiththevisualandexecutionworkloadsondrivingperformanceswereinvestigated[7][8].Astheresults,itwasmoredangerousforconditionswithonlytheexecutionworkloadsthanthosewith

thevisualandexecutionworkloads.Inthispaper,wefocusedonlyonexecutionworkloadstoclarifytheinfluenceofthepostureofdriversonthedrivingperformances.Theperformanceswereexaminedbyusingadrivingsimulator（DS）underemergencydriving.Moreover,weanalyzedthe

influenceofthereachingactionsbyusingadrivermodel.Todescribethedrivingbehaviorintothedrivermodel,quantitativelyevaluationoftheperformanceasvaluesofparametersofthemodelcanbeconducted.

2、DRIVINGSIMULATOR

Inthisstudy,adrivingsimulator（DS）wasemployedforexperiments.TheDSwasconsistedfromacomputer,adisplay,andasteeringwheelasshowninFigure1.ThesteeringwheelwasGTForceRX（logicool）.

Asthesteeringwheelwasforracegame,maximumangleofthesteering

wheelwassmall.Also,accelerationandbrakepedalswere

setup.A19inchLCDdisplaywasplacedat900mmfromthe

drivers.Graphicsofadrivingsceneduetodriver’soperations

wasprogrammedbyusingC/C++languagelibrariesofWorld

ToolKit（Sense8）.

2.1Vehiclemodel

AvehiclemodelcalculatedintheDSwasequivalenttotwowheelsmodelasshowninFigure2.Thevehiclemodelhadsimpletwodegreesoffreedom,whichwerealateralmovementandayawrotation,asfollows[9];

wherethevariableywasalateralpositionandψwasayawangleofthevehicle,whichwereinagroundcoordinatesystem.TheconstantMandIdenotedvehiclemassandyawmomentofinertiaofsprungmass,respectively.TheconstantLfandLrdenotelongitudinaldistancesfromsprungmasscenterofgravitytofrontandrearaxles.TheconstantFfandFrweresideforcesoffrontandrearaxlescombiningsideforcesofleftandrighttiresasfollows;

wherethevariableδdenotedsteeringangleasdriverinput.TheconstantCfandCrwerecorneringpowersofthefrontandreartires,whichmeantstiffnessbetweentiresandground.TheVwasvehiclevelocityasconstantvalue.Inexperiments,specificationsofthevehiclemodelweredeterminedasshowninTable1,whichwereadjustedassameasasmallsedancar.InTable1,steeringgearratiomeantaratioofsteeringangledividedbysteeringwheelangle.Theratiowassmallerthanusualvalue,becausethemaximumsteeringwheelanglewastoosmall.

3DRIVERMODEL

Adriver’ssteeringoperationmodelisassumedtoreducethedeviationofthecoursefromafrontviewpoint,inthefollowingexpressionshowninFigure3.AmaneuvermodelconsistsofagainK.Thetransferfunctionofawholeofthemodelisshownbyanequationasfollows;

wherethevariablesyandψdenotecarlateralpositionandyawangle,respectively.Thevariableyrislateralpositionofthetargetcourse.TheconstantvalueVmeansvelocityofthevehicle.TheTdenotespreviewtime.Thedriversviewatthepreviewpointfromthecockpitforcontrollingthevehicle.Apreviewpointisdefinedasthepointlocateddistance,expressedasV*T,aheadfromthecenterofgravity.Ifthedriver’smaneuvermodelissimpleasequation（5）,itiseasiertoestimatetheparametersKandTfromexperimentaldataandtoconsidertheresultsoftheestimationthanacomplexmodel.Asthewhole,thedriver-vehicleclosedloopmodelisasshowninFigure4.Ifthedriversperformreachingtasks,thevariablesKandTchangecomparedwiththatwithno

reachingtask.

4EXPERIMENTSOFEMERGENCYAVOIDANCE

4.1Reachingtasks

Thereweresixtargetsofthereachingtasks,asshowninFigure5.ThetargetsofPS1,PS2andPS3,denotedasSideTargets,wereonaleftsideofadriver’sbody.ThepositionsoftheFrontTargets,shownasPF1,PF2andPF3inthefigure,were300mmfromthefrontoftheSideTargets.Theheightofalltargetswassetto820mmfromthefloor.Intheexperiments,driversgraspedthetopofthetripodstandasshowninFigure6.Thentheyoperatedsteeringwheelwithmaintainingtheirposturesasreachingtasks.Ascontroltaskconditions,twoofnoreachingtasksweretested,whichweresingle-handedandtwo-handeddriving.

4.2Drivingconditions

Thedrivingconditionswereacar-followingandemergencyavoidancewiththesteering.Figure7showsdrivingconditions.Iftheexperimentalrunningwasstarted,subject’scarandaforwardcarrunat40km/hwith15moftheheadwaydistance.Thedrivercanoperateonlythesteeringwheel.Ifsubject’scarrunthroughalineA,100mfromthestartposition,theforwardcarsuddenlydeceleratedatarandom

distancebetween0mand200mfromthelineA.Afterthat,theforwardcarstoppedat10mfromthelineBwheretheforwardcarstartedtodecelerate.Thenthedriversmustavoidtheforwardcarbythesteeringoperation.Finally,thedriversrunstraightfor150mfromthelineB.Thesequenceabovewassettoonetrial.

4.3ExperimentalProcedure

1）Driversreachedtheirlefthandtothetarget,andthenmaintainedtheirposture.

2）Thedriversstartedtofollowacaralongalongstraightroad.Theheadwaydistancewas15m,andspeedsofbothcarsweresetto40km/h.Thedriverfollowedtheforwardcaronlyoperatingsteeringwheelwithout

requiringanyfootpedaloperations.

3）Theforwardvehiclesuddenlydeceleratedattherateof0.6*9.8m/s2afterarandomlygivenduration.Thenthedriverdeflectedtheforwardcartotherightbyusingthesteering.Atthistime,thedriverchangedtheposturetonormaldrivingposture.

4）Thedriversrun150malongthestraightroad.

5）Returnto1）.Thereachingtargetwassettootherone.

Onesetoftheexperimentsincludedsixtargets’conditionsandtwocontrolconditions.Therewere5setsofexperimentsforeachsubject.Sixuniversitystudentsparticipatedintheexperiments.

4.4Results

Figure8showstheaveragesteeringanglesforeachsubjectbythedifferenttarget.Thesteeringanglesincaseofthecontrolconditionsareshowninthefigureaswell.Accordingtothefigure,amplitudeofthesteeringanglesbecamelarge,whenthetargetspositionbecamefarfromthedrivers.Moreover,whenthedriverschangedlaneswithreaching

tasks,amountofovershootofsteeringangleswasfoundtobelargerthanthatwithoutthetasks.Thisimpliedthatmagnitudeofthesteeringoperationsbecamelargeduetothegenerationofthe