Page - 111 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2018,11, 3442 Table9.SummaryofBrownmodel. Model Statistics Ljung-Box No. ofOutliers StationaryR2 Statistics DF Sig. Industry 0.281 3.802 17 1.00 0 Agriculture 0.80 49.040 17 0.000 0 Domestic 0.432 6.242 17 0.991 0 Commercial 0.421 3.999 17 0.999 0 Traction/Railways 0.393 33.210 17 0.011 0 Others 0.411 23.559 17 0.132 0 TEC 0.067 14.569 17 0.627 0 3.4.4. Timeseriesmodeler (ExponentialSmoothing—DampedTrend) TheTECfor theyears2019,2024and2030were forecastedtobe1,162,453MW,1,442,410MWand 1,778,358MWrespectively. RMSE= √ {Σ (Yactual−Yforecast)/N} TheExpertmodel selectsdifferentmodelson itsownfordifferentvariablesandproduces the abovementionedforecastbymeansofa lowrootmeansquareerrorvalue,RMSEof10,734.649anda R2valueof0.997which iscomparativelyhigh.Andtheanalysisof theDampedtrendmodel is shown intheTable10. Table10.SummaryofDampedtrendmodel. Model Statistics Ljung-Box No. ofOutliers StationaryR2 Statistics DF Sig. Industry 0.392 2.365 15 1.00 0.392 Agriculture 0.337 44.861 15 0.000 0.337 Domestic 0.570 6.543 15 0.969 0 Commercial 0.467 3.720 15 0.999 0 Traction/Railways 0.114 34.625 15 0.003 0 Others 0.062 19.223 15 0.204 0 TEC 0.760 14.245 15 0.507 0 TheforecastedvaluesareshowninTable11 for theabovementionedyears. Table11.Summaryof theresultsandTimeline forecastedvalues for2030. Sector Model 2019 2024 2030 Industry Brown 538,089 686,457 864,498 Agriculture Brown 208,891 258,836 318,770 Domestic Brown 259,381 321,054 395,062 Commercial ARIMA 115,130 172,213 279,201 Traction/Railways ARIMA 20,554 26,906 465,523 Others ARIMA 68,072 100,343 37,404 TEC Brown 1,162,453 1,442,410 1,778,358 3.5.MovingAverage Thethreeyears’ fouryears’andfiveyears’movingaveragefor the timeperiodof1974–2014 is computedhereandshowninFigure9. Thevalueswere foundtobe830,696,796,620and759,825MW respectively for theyear2014. 111