eco-driving

201212thInternationalConferenceonITSTelecommunicationsIntelligenteco-drivingsuggestionsystembasedonvehicleloadingmodelWei-YaoChou,Yi-ChunLin,Yu-HuiLinandSyuan-YiChenVehicularInformation&ControlDepartmentIndustrialTechnologyResearchInstituteHsinchu,Taiwan,RepublicofChina{wychou;chun;Linuh;chensy}@itri.org.twAbstract-Eco-drivingskillhasbeengettingmoreandmoreattentionsbecauseofglobalwarningandincreasingoilprice.Sofar,existingeco-drivingassistancesystemsmainlyofferedrawinstantaneousfueleconomytodriver.However,inexperiencedriverstillhadthedifficultytoturnrawfueleconomyinformationintopropereco-drivingbehavior.Forthissituation,anintelligenteco-drivingsuggestionsystembasedonvehicleloadingmodelwasdeveloped.Theinstantaneousfueleconomywascomputedaccordingtotheinformationfromvehicleonboarddiagnosticsystem.Inaddition,fuzzyinferencesystemwasappliedtoestimateeco-Ievelandfuzzyruleswereutilizedtoestablishavehicleloadingmodel.Theappropriateeco-drivingsuggestionwasanalyzedbybuilt-inartificialintelligenceandcanbedisplayedonanyAndroidportabledevice.Finally,thedevelopedeco-drivingsuggestionsystemwasportedonSmartVehicleInformationGateway,installedonrealvehicleandtestedonrealtrack.Theexperimentalresultsprovedthat7%fueleconomycanbeimproved.Keywords-Eco-DrivingAssistanceSystem;SmartVehicleInformationGateway;Fueleconomy;On-BoardDiagnostic.I.INTRODUCTIONEco-drivingskillhasbeengettingmoreandmoreattentionsbecauseofglobalwarningandincreasingoilprice.Howtoincreasefuelefficiencybecomesaglobalchallenging.Generallyspeaking,vehiclefueleconomyishighlycorrelatedtodrivingbehavior.Inordertodriveecologically,combinationofvehiclecharacteristics,roadconditionsanddrivingbehaviorsisrequiredtobeoptimized.Improperdrivingbehaviorsmaycauseevenworsefuelefficiency.Therefore,moreandmoreEco-DrivingAssistanceSystem(EDAS)havebeenbeingproposed[1-3].ThefunctionofEDASwastopromptfueleconomyinformationoreco-indicationtodriver.Throughtheinformation,theultimategoalwastosaveenergyandreduceCO2emissions.AccordingtothestatisticaldatafromiSuppli,vehiclewithbuilt-inEDASwillrisefrom0.8%in2010to16.8%in2015asshowninFigureI[4].ThetrendshowedthatEDASsbecomesmoreimportantandcommon.Inthelastdecade,severalwell-knownEDASshavebeencommercialized.Forexample,aneco-drivingindicatorwasadoptedinNissanTiida/Livina[5].TheinstantaneousfueleconomywaspromptedonHeads-upDisplay(HUD).Besides,PLXDevicesCorporationproposedaneco-drivingdevicenamedKiwiin2008[6].Vehiclespeed,enginerevolution,978-1-4673-3070-1/12/$31.00©2012IEEE558engineloadingandKiwiscorewereoutput.TheKiwiscorewasevaluatedbasedonsmoothness,drag,accelerationanddeceleration.ThehigherKiwiscorerepresentedhigherfuelefficiency.Moreover,MosomotoCorporationinKoreadevelopedintelligentEnvironmentalDrivingSystem(iEDS),fueleconomy,accumulatedfueleconomyandtraveldistanceandetcwasoutputbyLCD[7].Alltheabovesystemsweredesignedtoimprovefuelefficiency,butnoneofthemspecificallyguidedriverhowtodriveecologically.Therefore,wedevelopedaneco-drivingsuggestionsystemwithspecificguidebasedonvehicleloadingmodel.Theinformationfromvehicleonboarddiagnosissystemwascapturedtocalculateinstantaneousfueleconomy.Inaddition,realtimeeco-Ievelwasprovidedtodriverandfuzzysetswereappliedtoconstructvehicleloadingmodel.Thevehicleloadingwasestimatedaccordingtothecorrelationofthrottlepositionandacceleration.Underdifferentvehicleloadinglevels,driverwasguidedtoapproachtheoptimaleco-drivingspeed.Inadditiontostandalonemode,throughICTtechnology,allfrontendvehiclesensoringinformationcanbetransmittedintobackendTelematicsserverforfurtheranalysis.Forpracticalbenefit,assumeonevehiclenormallyconsumes1,000litersofgaseveryyear.If10%fueleconomycanbeimproved,100USdollars(basedonIUSdollarperliter)and110kilogramsofCO2emissioncanbereducedeveryyear.II.SYSTEMOVERVIEWA.SystemplatformTheproposedeco-drivingsuggestionsystemwasimplementedandportedonSmartVehicleInformationGateway(SVIG)asshowninFigure2.SVIGrunonFreeRTOSembeddedsystemandwascapableofcollectingthe3Sr---------------------------------��t-------------------------------_,�2St---------------------------------�20r_------------------------�._---­i,Sr_----------------------��10r_--------------------20072008200920102011201220132014201520162020Figure1.Forecastofbuilt-inEDASgrowthpercentagefrom2007to2020.SmartVehicleInformationGatewayr--------------------------------�IIIFigure2.SVIGsystemarchiteture.datafromOBD-II(On-BoardDiagnostic),CANbus,gyroscope,3-axisaccelerator,temperature,compassandpositioninginformation.The32-bitcoremicroprocessorwasinchargeofschedulingandcomputing.TherawsensoringinformationandanalyzedinformationnotonlycanbetransmittedtoportabledevicebybluetoothbutalsotobackendTelematicsserverbyWiFiconnection.B.SystemfunctionsThemajorfunctionsoftheproposedeco-drivingsuggestionsystemincludedcollectionofdrivinginformation,fueleconomyestimationandeco-drivingsuggestion.Theyaredescribedasfollowing.(1)Collectionofdrivinginformation:3-AxisacceleratorandOBD-IIinformationcanbecollectedforcalculationandanalysis.(2)Fueleconomyandeco-Ievelanalysis:Theinstant