A HYBRID SOLAR IRRADIANCE NOWCASTING APPROACH: COMBINING ALL SKY IMAGER SYSTEMS AND PERSISTENCE IRRADIANCE MODELS FOR INCREASED ACCURACY (conference)

Abstract

New technical challenges arise due to the growing share of distributed and intermittent solar power generation. Intra-minute and intra-hour variabilities, when insufficiently accounted for, could lead to grid instabilities. Shortest-term forecasts, often referred as nowcast, can anticipate such variabilities with high resolutions in space and time. Therefore, nowcasts are suited for fine-grained control applications in solar power plants and electrical grids. Highly resolve nowcasts can be obtained by all sky imager (ASI) systems. Sky conditions are analyzed and irradiance nowcasts a derived from the hemispherical sky images. In this work, ASI based nowcasts and persistence based nowcasts are combined to a hybrid nowcasts approach. This hybrid approach archived a reduction of the root mean square deviation of 12% over a 62-day lasting validation data set.