Seite - 204 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2018,11, 1948 2.5.OptimizationofFourHybridPowerModels Basedonthegeneralhybridpowermodel, as showninFigure1,weconsidered the following fourhybridpowersystemswithdifferentcombinationsofenergysourcesandstorage: 1. Solar_Wind(SW)system:Thesystemcontains twoenergysources (asolarpanelandWT)and oneenergystoragemethod(Li-Febattery). 2. Solar_Wind_PEM_HE(SWPH)system:Thesystemcontains threeenergysources (asolarpanel, WT,andPEMFC)andtwoenergystoragemethods (aLi-Febatteryandahydrogenelectrolyzer). 3. Solar_Wind_PEM_CHG(SWPC)system:Thesystemcontains threeenergysources (asolarpanel, WT,andPEMFC)withachemicalhydrogengeneratorandoneenergystoragemethods (anLi-Fe battery). 4. Solar_Wind_PEM_HE_CHG(SWPHC)system:Thesystemcontains threeenergysources (asolar panel,WT,andPEMFC)withachemicalhydrogengeneratorandtwoenergystoragemethods (anLi-Febatteryandahydrogenelectrolyzer). ThecorrespondingSimPowerSystemmodelsare illustrated inFigure8. 6:3+&ġ6:3 6:ġ 39 DUUD\Vġ +\GURJHQ HOHFWURO\]HUġ /L32 %DWWHU\ġ /RDGġ +\GURJHQġ (OHFWULFLW\ġ :LQG WXUELQHġ :LQG FRQWUROOHUġ +\GURJH Q VWRUDJHġ )XHO FHOOġ &KHPLFDO K\GURJHQ JHQHUDWRUġ '& '& &RQYHUWHUġ '& $& ,QYHUWHUġ 6:3 Figure8.Thefourhybridpowermodels. Threestandard loadconditions,asshowninFigure7a,wereappliedto the fourhybridpower models to predict systems responses. Then, we used Equations (5)–(11) to evaluate system cost andreliabilityusingdifferentcomponentsizes. TheresultingreferenceplotsareshowninFigure9, where thenumberofWTswasset tozero,becauseusingaWTtendedto increase thesystemcosts. Theoptimalsystemcostsof the fourhybridpowersystemsare illustrated inTable3. 204