新型太阳能汽车路线优化外文翻译中英文.docx

新型太阳能汽车路线优化外文翻译中英文.docx

《新型太阳能汽车路线优化外文翻译中英文.docx》由会员分享，可在线阅读，更多相关《新型太阳能汽车路线优化外文翻译中英文.docx（18页珍藏版）》请在冰点文库上搜索。

新型太阳能汽车路线优化外文翻译中英文

新型太阳能汽车路线优化外文翻译中英文

英文

CriteriaforSolarCarOptimizedRouteEstimation

MehrijaHasicic，DamirBilic，HarunSiljak

Abstract

ThispapergivesathoroughoverviewofSolarCarOptimizedRouteEstimation（SCORE）,novelrouteoptimizationschemeforsolarvehiclesbasedonsolarirradianceandtargetdistance.Inordertoconducttheoptimization,bothdatacollectionandtheoptimizationalgorithmitselfhavetobeperformedusingappropriatehardware.Herewegiveaninsighttobothstages,hardwareandsoftwareusedandpresentsomeresultsoftheSCOREsystemtogetherwithcertainimprovementsofitsfusionandoptimizationcriteria.ResultsandthelimitedapplicabilityofSCOREarediscussedtogetherwithanoverviewoffutureresearchplansandcomparisonwithstate-of-the-artsolarvehicleoptimizationsolutions.

Keywords：

Vehiclerouting，Electricvehicle，Solarvehicle，Navigation，Routeoptimization

Introduction

Developmentofnavigationsystemshasbeenanimportanttopicinoptimizationonceportableelectronicdeviceswerefeasibleandrouteselectionandoptimizationhavebeenthevitalpartofit,aimingatfuelconsumptionreductionanddriversatisfaction,whichisingeneralamulti-objectiveoptimizationproblem.Withtheadventofelectricandautonomouscars,attentionindevelopmentofoptimizationalgorithmsturnedtothem,utilizingpropertiesofthesenewvehicles.Navigationforautonomousvehiclesalsoallowsuseofalgorithmspreviouslydevelopedformobilerobotics.

GeographicInformationSystem（GIS）integratesvariousdatatypes,manyofwhichareinstrumentalinnavigation,leadingtoextensiveuseofGISinrouteoptimization.Fromourperspective,itisalsoimportanttonoteutilizationofGIS（solarradiationmaps）insolarenergymanagement asitallowsustouseGISforouroptimizationaswellasaninputprovider.

Inthespiritofoptimizationwithrespecttofuelconsumption,powermanagementoptimizationinelectricandsolarcarshasbeeninvestigatedintheoryandpractice,andforsolarhybridcarsthepowermanagementschemesfocusonthequestionofswitchingenergysourcesandoptimizationofresources.Sunshineforecastasanoptimizationinputhasbeenrecentlyintroduced andusedmainlyforparkingplanning.

ThispaperpresentsSolarCarOptimizedRouteEstimation（SCORE）,anovelrouteoptimizationsystembasedonproposingsunniestroutesandsunniestparkingspots,thereforeutilizingtheoptionsofchargingwhiledrivingandchargingonparkinglots.Thedataonsolarirradianceforroutesandparkingspotsisafusionofpreviouslycollected,realtimeandforecasteddata.

Describingthewholeprocessfromdatacollectiontorouteselection,thispaperprovidesboththeoreticalandpracticaltreatiseofSCORE,givingageneralstructureandtherealworldimplementationofit.ThepaperitselfisanextensionoftheworkundersametitlepresentedinMECO2016conference,presentingtheimplementationchallengesandsolutions.Inadditiontopreviouswork,thispaperalsoextendstheanalysisofcostfunctionsusedinSCOREforselectionofroutesandparkingplaces,aswellthemathematicalmodelofdatafusion,togetherwithmoredetailsontheresults.

Stateoftheart

Routeplanningandselectionforroadvehicleshasbeenasubjectofinterestfordecades,withthefirstcommercialdigitalmapnavigationsystemappearingmorethan30yearsago.Sincethen,variousrouteplanningsystemshavebeenproposed,basedondifferentalgorithmsandinputs.

Dijkstra'salgorithmhasbeenthesimplestalgorithmforimplementationandservesasabenchmarkforstorageandtimeconsumptioninrouteplanningforcars whencomparedwithotheralgorithms,frombidirectionalandA*searchtospecialgoal-driven,hierarchicalandbounded-hopalgorithms.AlthoughDijkstra'salgorithmistheslowestoptioninbignetworks,itcanstillbeusedasaproofofconcept.

Whileinthebeginningtherouteplanningandselectionsystemshadasingleobjective:

namely,fuelconsumptionorjourneytimeminimization,soonaftertheirinceptionmulti-objectiveoptimizationmodelsweredeveloped,oftenusingartificialintelligencetechniquestocombinedifferentgoalsThesegoalsoftenincludepersonalizedchoicesofdrivers andtheirpersonalattitudetowardspossibleroutes.

Parkingselectionhasbeenstudiedextensivelyaswell.Ithasbeenmodeledasamulti-inputproblemmeasuringtheutilityofaparkingspacebyaccountingforavailability,drivingduration,walkingdistancetothedestination,parkingcost,trafficcongestion,etc.

Intermsofsolarcaroptimization,thepowermanagementtechniquesmentionedintheintroductionhavebeenextendedtominimizetotalenergyconsumptionbyplanningspeedonpartsofthepathdifferentlyexposedtothesun.Thiswork,publishedatthesametimeasthefirstSCOREresults buildsuponcloselyrelatedworkonsolarpoweredrobots.

In authorsproposeasolarracecaroptimizationbasedonweatherforecastandvelocityprofile,determiningtheneedforaccelerationanddecelerationthroughouttheracecourseinordertomaximizetheaveragevelocity.Similartaskisdonein aswell,butthelatteralsoincludessomewhatmorecomplexweathermodelandsolarpositionalgorithm.Theweathermodelisarandomwalkaroundexpectedirradiancecurve,whilesolarpositionisdeterminedthroughtheNREL（NationalRenewableEnergyLaboratory）model.

Incomparisontothesesolutions,SCOREhassomewhatdifferentpurpose.Aimingatofferingaframeworkforoptimizedcityandintercitytravel,SCORE'smodelreliesondatawhichcanbecollectedinamoreregularfashionthanthedataforaracetrack.Moreover,SCORE'soutputistheroute,unliketheoutputsofpowermanagementoptimizationsystemshavingthevelocityprofileorengine-generatorpowertrajectory（incaseofhybridcars）astheoutput.ThisdoesmeanthatSCOREshouldbeextendedsoitcoversthevelocitycontroland/orhybridvehiclepowerswitching,butatthispointthefocusisonrouteselection.Thisisalongthelinesoftheideain,wherethepathselectionisfollowedbyspeedprofileselection.

SCOREsystemdescription

SCOREsystemconsistsofthreeseparableparts,indicatedinnamely:

1.Mobilesensordatatransmitter,transmittingsolarirradiancedatathroughwirelesschannelinrealtimefromtheroads.Althoughwewillusethetermmobilesensorsthroughoutthispaper,theycanbestationaryaswell,placedatselectedplacesbytheroad.Whenmobile,thesetransmittersarenotnecessarilyplacedonsolarcarsusingSCOREastheirnavigationsystem.Theycanbeplacedonfossilfuelandelectriccarsaswell.Preferably,carscarryingthesensorswouldbeofteninmotion,coveringalargearea（e.g.taxis,publictransportation）.

2.Serverfordatafusion,collectingreadingstransmitterssendfromthefieldandthirdpartysources,processingthemandcombiningwithofflinedata（whichcanincludeweatherforecastandhistoricreadings,asthenextsectionwillshow）andallowingthecarcomputerclientstofetchtheprocesseddatainappropriatematrixformat.

3.Embeddedcarcomputerclientinthesolarcar,takingtheprocesseddatafromserver'scloudserviceandcustomizeitonitsownbyusingreadingsfromitsownsensors.Built-inlightsensorcanbeusedfornormalizationofdata,andelectricmeasurementsfromthecarcanbeutilizedforstateestimation.Finally,theusercanenterthedestinationandobtaintheproposedroute,whichshoulddynamicallychangebasedonweatherupdates.

Datacollection

Theoreticalconsiderationsofirradiationdataanalysis

Inordertoselectrouteswithhighestsolarenergygain,SCOREsystemhastohaverelevantsolarirradiationdata.Sinceitisnotpossibletoalwayshaveuptodatedatainrealtimeforeveryroadsegmentconsideredbythealgorithm,itisimportanttousedifferentsourcesofinformation.Inthiswork,wehavedividedtheirradiationdataintotwocategories:

1.Onlinedata,gatheredbythemobilesensordatatransmittersandupdatedinregularfashion.Thedataforeachlocationisrepresentedbyrealnumbersbetween0（noirradiance）and1（maximumirradiance）withatimestampfordatasamplecollection.Inthispaper,timestampsareintegersdenotinghoursstartingfromareferencetime（beginningoftheyear）.

2.Offlinedata,generatedusingnumericalsunshineforecast,CAD（computeraideddesign）andGISmodelsforpredictionofsolarirradianceforaparticularlocation.CADdataisgeneratedfromCADstreetmodelsandsimulatingsunmovement,whileGISdataistakenfromtheGISservicesdoingsolarirradiancemeasurementsforareasofinterest.ThisdataisprovidedinanaggregatedformbyGoogleaswellthroughtheirProjectSunroofforhousingsolarpanelplanning.

Thesetwonumericalvalues,denoted ron （normalizedvalueofonlinedatairradiance）and roff （normalizedvalueofirradianceinferredfromofflinedata）arecombinedforeachgeographicallocation（intheoptimizationpart,wewillrefertotheseasgraphnodes）ontheserver.Detailsofthisfusionwillbediscussedinaseparatesection.

Implementationofsensordatacollectionandtheserver

Mobiledevicedevelopedwithinthisprojectiscompactandautonomous,whichenablesitsplacementonavehiclemovingthroughthecitytocollectirradiationdatawithoutcustomizationofthecaritselforitsroutes.OnceagainitisemphasizedthatthevehiclescarryingthesemobiledevicesdonothavetobevehiclesusingSCOREfornavigation,i.e.traditionalfossilfuelcarsmayservethepurposeofdatacollecting“crawlers”.

Whileanywirelessprotocolcouldbeusedfortransmissionfromthesemobiledevices,weproposetheuseofpacketradio.ItseasiestimplementationisAPRS（Auto