Seite - 322 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2018,11, 1449 Figure 2. The relationship between meteorological conditions and power load. (a) the highest temperatureandload; (b) theweatherandload; (c) theaveragewindspeedandload; (d) theaverage humidityandload. Similar to traditionalpower loads, thedaily loadof thechargingstation increasesowingto the useofairconditionersone-buseswhenthetemperaturechangeofcoldnessandwarmthisaggravated. Since temperature has an important influence onbattery capacity, aswell as on the charging and dischargingprocess, thecharging time isdiverseatdifferent temperatures,whichalso leads todistinct trendsof load. Thedaily loadcurves fromSeptember12 to14,2017are takenasanexample, inwhich the totalnumberofchargede-buses in these threedayswasabout60andthemaximumtemperature droppedfrom35to24.Asseen inFigure3, theviolentfluctuationofair temperature inadjacentdays causesgreatchanges indaily loadcurves. Thus, it isnecessary to take temperatureasaninfluential factor in theselectionofsubsequentsimilardaysamples. Figure3.Relationshipbetweentemperatureanddaily load. Taking the daily load curves on August 29 and August 30 in 2017 as an example, weatherconditionscanbedividedintosunnydaysandrainydays. Figure4 illustrates therelationship betweenweatherconditionsandthedaily loadof thechargingstation. Itproves thatdailymaximum 322