资源描述
Confidentiality/dateline:13ptArialRegular,whiteMaximumlength:1lineInformationseparatedbyverticalstrokes,withtwospacesoneithersideDisclaimerinformationmayalsobeappearinthisarea.Placeflushleft,alignedatbottom,8-10ptArialRegular,whiteIndicationsingreen=LivecontentIndicationsinwhite=EditinmasterIndicationsinblue=LockedelementsIndicationsinblack=OptionalelementsCopyright:10ptArialRegular,whiteEE5900:AdvancedEmbeddedSystemForSmartInfrastructureSingleUserSmartHomeSmartGird2Asmartgridputsinformationandcommunicationtechnologyintogeneration,transmission,distributionandenduser,makingsystemscleaner,safer,andmorereliableandefficient.SmartHome3SmarthometechnologiesareviewedasusersendoftheSmartGrid.Asmarthomeorbuildingisequippedwithspecialstructuredwiringtoenableoccupantstoremotelycontrolorprogramanarrayofautomatedhomeelectronicdevices.Smarthomeiscombinedwithenergyresourcesateithertheirlowestpricesorhighestavailability,e.g.takingadvantageofhighsolarpaneloutput.yousharez/2019/11/20/house-of-dreams-a-smart-house-concept/SmartAppliancesSmartAppliancesCharacterizedbyCompactOSinstalledRemotelycontrollableMultipleoperatingmodes4HomeApplianceRemoteControl5ZigBeeCertifiedAppliancesandHomeAreaNetwork(HAN)zigbee.org/6System7PowerflowInternetControlflowDynamicPricingfromUtilityCompanyIllinoisPowerCompanyspricedata8Pricing for one-day ahead time periodPrice($/kwh)BenefitofSmartHomeReducemonetaryexpenseReducepeakloadMaximizerenewableenergyusage9SmartHomeControlFlow10PHEVTransitionbetweentheRenewableEnergyandPowerGridEnergyAtransfer switchisanelectricalswitchthatreconnectselectricpowersourcefromitsprimarysourcetoastandbysource.Switchesmaybemanuallyorautomaticallyoperated.11SmartSchedulingDemandSideManagementwhentolaunchahomeapplianceatwhatfrequencyThevariablefrequencydrive(VFD)istocontroltherotationalspeedofanalternatingcurrent(AC)electricmotorthroughcontrollingthefrequencyoftheelectricalpowersuppliedtothemotorforhowlongusegridenergyorrenewableenergyusebatteryornot125cents/kwh 3cents/kwh5kwh10kwhPowerPowerrTimeTime12123(a)(b)VFDImpact5cents/kwh 3cents/kwhcost=10kwh*5cents/kwh=50centscost=5kwh*5cents/kwh+5kwh*3cents/kwh=40cents13UncertaintyofApplianceExecutionTimeInadvancedlaundrymachine,timetodothelaundrydependsontheload.Howtomodelit?14ProblemFormulationGivennhomeappliances,toschedulethemformonetaryexpenseminimizationconsideringVFDwithconsideringvariationsSolutionsforcontinuousVFDSolutionsfordiscreteVFD15SolutionsforcontinuousVFDSolutionsfordiscreteVFD1234TheProcedureoftheOurProposedScheme16OfflineScheduleAdeterministicschedulingwithcontinuousfrequencyAdeterministicschedulingwithdiscretefrequencyStochasticProgrammingforApplianceVariationsOnlineScheduleforRenewableEnergyVariationsTheProposedSchemeOutline17LinearProgrammingforDeterministicSchedulingwithContinuousFrequencyminimize:subjectto:18MaxLoadConstraintToavoidtrippingout,ineverytimewindowwehaveloadconstraint19ApplianceLoadConstraintSumupineachtimewindowappliancepowerconsumptionisequaltoitsinputtotalpower20ApplianceSpeedLimitandExecutionPeriodConstraintThefrequencyisupperboundedAppliancecannotbeexecutedbeforeitsstartingtimeorafteritsdeadline21PowerResourcePowerresourcecanbevarious22SolarEnergyDistributionConstraintSolarEnergycanbedirectlyusedbyhomeappliancesorstoredinthebattery23BatteryEnergyStorageConstraintandChargingCostSolarEnergyStorageBatteryChargingCost24TheProposedSchemeOutline25DeterministicSchedulingforDiscreteFrequencyFlow26DetermineSchedulingAppliancesOrderScheduleCurrentTaskUpdateUpperBoundofEachTimeIntervalAnapplianceScheduleAppliancesNotalltheappliance(s)processedAllapplianceprocessTheProposedSchemeOutline27GreedybasedDeterministicSchedulingforTaski280t1t2t3t4TaskiPricePowerTimeTimeCannothandlenoninterruptiblehomeappliancesTheProposedSchemeOutline29DynamicProgrammingbasedDeterministicSchedulingforTaskiForasolutionintimewindowi,energyconsumptioneandcostcuniquelycharacterizeitsstate.Forpruning:e1,c1willdominatesolutione2,c2,ife1=e2andc1=c2.30(15,20)(11,22)(1,2)(2,4)(3,6)(1,1)(2,2)(3,3)0t1t2(6,9)(5,8)(4,7)(5,7)(4,6)(3,5)(4,5)(3,4)(2,3)(0,0)(0,0)(3,3)(2,2)(1,1)#ofdistinctpowerlevels=k#timeslots=mRuntime:PriceTimeDynamicProgrammingreturnsoptimalsolutionHandlingMultipleTasksAccordinganorderoftasksPerformthedynamicprogrammingalgorithmoneachtask31TheProposedSchemeOutline32VariationimpactstheSchemet2t3t4WorstcasedesignEvaluateBestcasecanbeimprovedt1BestPriceWindowCostcanbereduced33BestCaseDesignt1t2t3t434VariationAwareDesignAnadaptationvariableisintroducedtoutilizetheloadvariation.t1t2t3t435MonteCarloSimulationIttakes5000differenttasksets,toevaluateavalue.Evaluatehowmanysamplesdonotviolatetripraterequirement.Triprate=tripoutevent/totalevent36 UncertaintyAwareAlgorithmAlgorithmicFlowOutput:ScheduleInput:TasksetwithtaskswhichcanbescheduledYesupdatetaskloadbasedonGenerateappliancesschedulebysolvingtheLPDerivecurrenttriprateusingMonteCarlosimulationCurrenttriprateTargetUpdateNoCore1updatetaskloadbasedonGenerateappliancesschedulebysolvingtheLPDerivecurrenttriprateusingMonteCarlosimulationCurrenttriprateTargetUpdateNoYesupdatetaskloadbasedonGenerateappliancesschedulebysolvingtheLPDerivecurrenttriprateusingMonteCarlosimulationCurrenttriprateTargetUpdateNoupdatetaskloadbasedonGenerateappliancesschedulebysolvingtheLPDerivecurrenttriprateusingMonteCarlosimulationCurrenttriprateTargetUpdateNoupdatetaskloadbasedonGenerateappliancesschedulebysolvingtheLPDerivecurrenttriprateusingMonteCarlosimulationCurrenttriprateTargetUpdateNoYesYesYesCore2Core3Core4from0to0.25from0.25to0.5from0.5to0.75from0.75to137MonteCarloSimulationtakes5000samplesLatinHypercubeSamplingtakes200samplesCurrentS38LatinHypercubeSamplingisastatisticalmethodforgeneratingadistributionofplausiblecollectionsofparametervaluesfromamultidimensionaldistributionAlgorithmImprovementExerciseHowtogeneralizedeterministicdynamicprogrammingtoanvariationawaredynamicprogramming?39TheProposedSchemeOutline40OnlineTuningActualrenewableenergyExpectedSavetherenewableenergyasmuchaspossibleActualrenewabledemand=ExpectedFollowtheofflineschedule41ExperimentalSetupTheproposedschemewasimplementedinC+andtestedonaPentiumDualCoremachinewith2.3GHzT4500CPUand3GBmainmemory.500differenttasksetsareusedinthesimulation.Thenumberofappliancesineachsetrangesfrom5to30,whichisthetypicalnumberofhouseholdappliances1.TwosetsoftheKD200-54PseriesPVmodulesfromInc2aretakentoconstructasolarstationforaresidentialunitwhicharecost$502.Thebatterycostissetto$753with845kWthroughputistakenasenergystorage.ThelifetimeofthePVsystemisassumedtobe20years4.ElectricitypricingdatareleasedbyAmerenIllinoisPowerCorporation51M.Pedrasa,T.Spooner,andI.MacGill,“Coordinatedschedulingofresidentialdistributedenergyresourcestooptimizesmarthomeenergyservices,”IEEETransactionsonSmartGrid,vol.1,no.2,pp.134144,2019.2DataSheetofKD200-54PseriesPVmodules,availableatkyocerasolar/assets/001/5124.pdf.3T.GivlerandP.Lilienthal,“UsingHOMERsoftware,NRELsmicropoweroptimizationmodule,toexploretheroleofgen-setsinsmallsolarpowersystemscasestudy:Srilanka,”TechnicalReportNREL/TP-710-36774,2019.4LifespanandReliabilityofSolarPanel,availableatsolarpanelinfo/solarpanels/solar-panel-cost.php.5Real-TimePrice,availableathttps:/www2.ameren.42LP-basedApproachvs.TraditionalApproachEnergyCost(cents)Runtime(s)householdappliancehouseholdapplianceCosttime43Traditionalvs.ContinuousVFDvs.Greedy44CostHouseholdapplianceOnlyD.P.CanHandleNonInterruptibleTasksetCostHouseholdappliance45ComparisonofWorstCase,BestCaseDesignandStochasticDesignEnergyCost(cents)TripRate(%)10secondsHouseholdapplianceHouseholdapplianceCostRate46Onlinevs.OfflineHouseholdapplianceCost(cents)47ExampleofaTaskSet48SummaryThisprojectproposesastochasticenergyconsumptionschedulingalgorithmbasedonthetime-varyingpricinginformationreleasedbyutilitycompaniesaheadoftime.Continuousspeedanddiscretespeedarehandled.Simulationresultsshowthattheproposedenergyconsumptionschedulingschemeachievesupto53%monetaryexpensesreductionwhencomparedtoanaturegreedyalgorithm.Theresultsalsodemonstratethatwhencomparedtoaworstcasedesign,theproposeddesignthatconsidersthestochasticenergyconsumptionpatternsachievesupto24%monetaryexpensesreductionwithoutviolatingthetargettriprate.Theproposedschedulingalgorithmcanalwaysgenerateamonetaryexpenseefficientoperationschedulewithin10seconds.49MultipleUsersPricingat10:00amischeap,sohowaboutschedulingeverythingatthattime?50Willnotbecheapanymore8:00GameTheoryBasedScheduling51Thanks52END
展开阅读全文