Seite - 299 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2018,11, 1561 2016a on a PCwith the configuration ofWindows 7 64-bit, Inter Core i5-4590 CPU@ 3.30GHz, 8GBRAM. Figure3.Datadescriptionofexperiments. (a)Locationofsamplesites; (b)Divisionof trainsetand test set; (c)Structureof inputsetandoutputset; and(d)EntropyofeachIMF). Table1.Relatedparameters inhybridmodel. SubmodelsandParameters Value ElmanNeuralNetwork(ENN) Inputnum 6 Hiddennum 13 Outputnum 3 Train.epoch 500 Train.lr 0.1 Train.func “Adam” Completeensembleempiricalmode decompositionwithadaptivenoise (CEEMDAN) Nstd 0.2 NR 200 Maxiter 100 Multi-objectivesalpswarmalgorithm(MOSSA) Dim 754 Lb −2 Ub 2 Obj_no 2 Pop_num 50 299