Seite - 151 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2018,11, 2080 howtheNNismuch lessdependenton the chosen thresholdwhile theARperformance is clearly thrownoffbyamisadjustedthreshold. Table4.Forecastingerror (RMSE)withdifferentHDDthresholdadjustment inBarcelona. TypeofDayand Model HDDThreshold 13 14 15 16 17 18 19 20 21 22 23 24 25 Overall AR 2.00% 1.98% 1.96% 1.95% 1.94% 1.92% 1.91% 1.91% 1.91% 1.91% 1.92% 1.93% 1.94% NN 1.92% 1.93% 1.93% 1.93% 1.94% 1.92% 1.92% 1.93% 1.92% 1.92% 1.93% 1.93% 1.94% Hot AR 2.62% 2.63% 2.63% 2.62% 2.60% 2.58% 2.55% 2.53% 2.50% 2.48% 2.46% 2.43% 2.45% NN 2.54% 2.55% 2.55% 2.53% 2.56% 2.54% 2.55% 2.54% 2.52% 2.51% 2.51% 2.46% 2.47% Cold AR 2.16% 2.11% 2.07% 2.03% 2.00% 1.99% 1.99% 1.99% 1.99% 2.00% 2.01% 2.02% 2.02% NN 1.65% 1.67% 1.69% 1.71% 1.71% 1.68% 1.71% 1.68% 1.72% 1.67% 1.71% 1.69% 1.69% 1.70% 1.73% 1.76% 1.79% 1.82% 1.85% 1.88% 1.91% 1.88% 1.90% 1.92% 1.94% 1.96% 1.98% 2.00% 2.02% 13 14 15 16 17 18 19 20 21 22 23 24 25 HDD Threshold (ºC) Overall Accuracy vs HDD threshold offset BAR - AR BAR - NN ZAR - AR ZAR - NN Figure10.Overall forecastingerror (RMSE)withdifferentHDDthresholdadjustment inBarcelona andZaragoza. 4.4.NumberofNeurons Thenumberof neurons in thehidden layer affects both computational burdenand theNN’s performance. Therefore,bothaspectsarereportedasresultsonthis test. Table5showstheaccuracy of theneuralnetworkas thenumberofneurons is increased. Inaddition, the forecasting timefora single24-hprofile is included. It isworthnoticing that therestof forecastingprocesses likedataaccess or treatmentalsoconsumetimeand, therefore, thereportedtimeisnot theonlyconcern inorder to obtaina timely forecast. Table5.Forecastingerror (RMSE)andexecutiontimewithdifferentnumberofneurons. TypeofDay NumberofNeurons 3 4 5 10 15 20 Overall 1.56% 1.55% 1.56% 1.55% 1.58% 1.62% Regular 1.49% 1.46% 1.45% 1.44% 1.46% 1.50% Special 2.00% 2.10% 2.28% 2.31% 2.36% 2.46% Hot 2.00% 1.93% 1.95% 1.93% 2.00% 2.04% Cold 1.55% 1.45% 1.51% 1.48% 1.51% 1.58% Time(s) 1.610 1.615 1.620 1.630 1.639 1.643 Testconditions: 7YT,10RN,5TL,12MF,7/14LAG. 151