پیش بینی انرژی خورشیدی و تایید با استفاده از پیش بینی های مدل عملیاتی و اندازه گیری انرژی خورشیدی روی زمین
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|55788||2015||13 صفحه PDF||سفارش دهید||9065 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Energy, Volume 93, Part 2, 15 December 2015, Pages 1918–1930
The present study focuses on the predictions and verification of these predictions of solar energy using ground-based solar measurements from the Hellenic Network for Solar Energy and the National Observatory of Athens network, as well as solar radiation operational forecasts provided by the MM5 mesoscale model. The evaluation was carried out independently for the different networks, for two forecast horizons (1 and 2 days ahead), for the seasons of the year, for varying solar elevation, for the indicative energy potential of the area, and for four classes of cloud cover based on the calculated clearness index (kt): CS (clear sky), SC (scattered clouds), BC (broken clouds) and OC (overcast). The seasonal dependence presented relative rRMSE (Root Mean Square Error) values ranging from 15% (summer) to 60% (winter), while the solar elevation dependence revealed a high effectiveness and reliability near local noon (rRMSE ∼30%). An increment of the errors with cloudiness was also observed. For CS with mean GHI (global horizontal irradiance) ∼ 650 W/m2 the errors are 8%, for SC 20% and for BC and OC the errors were greater (>40%) but correspond to much lower radiation levels (<120 W/m2) of consequently lower energy potential impact. The total energy potential for each ground station ranges from 1.5 to 1.9 MWh/m2, while the mean monthly forecast error was found to be consistently below 10%.