Page - 62 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2019,12, 164 Author Contributions: Conceptualization, A.A.; Formal analysis, A.M. and M.A.; Investigation, A.A.; Methodology,A.A.andN.J.; Software,A.A.;Supervision,N.J.;Validation,N.J.,A.M.andM.A.;Writing—original draft,A.A.;Writing—review&editing,N.J. andZ.A.K. Funding:This researchreceivednoexternal funding. Conflictsof Interest:Theauthorsdeclarenoconflictsof interest. Nomenclature SG Smartgrid DAL Day-aheadload DALF Day-aheadloadforecast(ing) AN Artificialneuron ANN Artificialneuralnetwork MARA Multivariateautoregressivealgorithm ARMA Autoregressiveandmovingaverage EDE Enhanceddifferentialevolutionalgorithm mEDE ModifiedversionofEDEalgorithm NIST National instituteofstandardsandtechnology MSE Minimumsquareerror P Historical loaddatamatrix TDP Historicaldewpoint temperaturedatamatrix TDB Historicalboilingpoint temperaturedatamatrix DTYP Historicaldewpoint temperaturedatamatrix phm,dn Loadvalueatmthhourof thenthday pcimax LocalmaximaforeachcolumnofP Pnrm LocallynormalizedP TDP,nrm LocallynormalizedTDP TDB,nrm LocallynormalizedTDB MI(K,G) Relativemutual informationbetweeninputKandtargetG pr(K,G) JointprobabilitybetweenKandG pr(K) IndividualprobabilityofK Sf Selectedfeatures ST Trainingsamples SV Validationsamples MAPE Meanabsolutepercentageerror pa Actual load pf Forecasted load Ith Irrelevancythresholdvalue Rth Redundancythresholdvalue y ′t i,j jth trialvectory ′ for ith individual ingeneration t xti,j jthparentvectorx for ith individual ingeneration t uti,j jthmutantvectoru for ith individual ingeneration t yti,j jthoffspringvectory for ith individual ingeneration t rnd Randomnumber FFN(.) Fitness function EF Forecasterror References 1. Gelazanskas,L.;Gamage,K.A.Demandsidemanagement insmartgrid:Areviewandproposals for future direction.Sustain. CitiesSoc. 2014,11, 22–30. [CrossRef] 2. Yan,Y.;Qian,Y.; Sharif,H.;Tipper,D.ASurveyonSmartGridCommunicationInfrastructures:Motivations, RequirementsandChallenges. IEEECommun. Surv. Tutor. 2013,15, 5–20. [CrossRef] 62