Predicting load harmonics in three phase systems using neura

《Predicting load harmonics in three phase systems using neura》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、PredictingLoadHarmonicsinThreePhaseSystemsUsingNeuralNetworksJoyMazumdar,R.G.HarleyandF.LambertGaneshK.VenayagamoorthySchoolofElectricalandComputerEngineeringDepartmentofElectricalandComputerEngineeringGeorgiaInstituteofTechnologyUniversityofMissouri-R

2、ollaAtlanta,GA30332-0250,USARolla,MO65409-0249,USAEmail:joymazum@ece.gatech.eduEmail:gkumar@ieee.orgAbstract-Thispaperproposesaartificialneuralnetworkmethod[3,4],harmonicimpedancemeasurement[5],andin(ANN)basedmethodfortheproblemofmeasuringtheactualrece

3、ntyearsartificialneuralnetworks(ANN)[6-9]toharmoniccurrentinjectedintoapowersystemnetworkbymeasuretheharmoniccontentintheloadcurrent,ortothreephasenonlinearloadswithoutdisconnectinganyloadspredictit,butmostofthemassumearadialfeedersupplyingfromthenetwo

4、rk.TheANNdirectlyestimatesoridentifiestheasingleloadthroughaknownfeederimpedance,ornonlinearadmittance(orimpedance)oftheloadbyusingthemultipleloadsconnectedtoaPCCwhichhasasinusoidalmeasuredvaluesofvoltageandcurrentwaveforms.Theoutputvoltageandwithzeroi

5、mpedanceinthesupplyfeeder.ofthisANNisawaveformofthecurrentthattheloadwouldhaveinjectedintothenetworkiftheloadhadbeensuppliedThispaperproposesanovelmethodbasedonArtificialfromasinusoidalvoltagesourceandisthereforeadirectNeuralNetworks(ANN)todeterminethe

6、trueharmonicmeasureofloadharmonics.currentofanonlinearloadinathreephasepowersystem.I.INTRODUCTIONII.LOADMODELINGUSINGNEURALNETWORKSTheincreaseduseofnonlineardevicesinindustryhasArtificialNeuralNetworkshaveprovidedanalternativeresultedindirectincreaseof

7、harmonicdistortioninthemodelingapproachforpowersystemapplications.Theindustrialpowersysteminrecentyears.Allloadsservicedmulti-layerperceptronnetwork(MLPN)isoneofthemostbytheutilityaredesignedtooperateat60Hz.Howeverpopulartopologiesinusetoday.Thisnetwor

8、kconsistsofanonlinearloadsdemandnonsinusoidalcurrentandthesesetofinputneurons,outputneuronsandoneormorehiddencurrentshavedetrimentaleffectonthepowersystem.Asanlayersofintermediateneurons.Dataflowsintothenetworkexample,Fig.1showsatypical