Clouds moving in front or away from the sun are the leading cause of irradiance variability. These variations have a repercussion on the electricity production of photovoltaic systems. Predicting such changes is essential for proper control of these systems and for maintaining grid stability. Images from the sky have proven to help with short-term solar irradiance forecasting, especially when combined with artificial intelligence. Nevertheless, these models tend to smooth the irradiance fluctuations. We propose a forecasting model to predict the clear-sky index in a forecast horizon of 20 min with a 1-minute resolution. Our model, based on a classifier to determine the sky conditions and, on an optical flow, applies an artificial intelligence model explicitly trained on each class of sky conditions. This strategy has an equivalent performance to an unclassified model and a forecast skill between 5 and 20% with respect to the smart persistence model for most classes of sky conditions while requiring considerably less training data. Although our model reduces the overall predicting error, it still has difficulties predicting irradiance changes and mainly overcast days. Our classifying strategy can be applied to other models targeting different objectives to predict sudden changes in either irradiance or power related to photovoltaic systems.

Short-term solar forecasting is crucial for large scale implementation of solar energy and plays an important role in grid balancing, energy trading, and power plant operation. Cloud movement is the main source of unpredictability within solar forecasting and can be recorded using All-Sky Imagers. Conventional cloud modelling methods using image analysis techniques are unable to extract the spatial configuration and the temporal dynamics of clouds, resulting in poor predictions of the interaction with solar radiation. The goal of this study is to create a deep learning model for short-term irradiance forecasting between 0 and 21 minutes into the future using all sky images combined with auxiliary data. The model performance was assessed by comparing the deep learning model with the persistence model and showed that the deep learning model outperforms the persistence model with 24.8%. A sensitivity analysis to data usage is performed showing that besides using more data, also the variation of using multiple years of data results in better performance. Furthermore, the sensitivity of the model to input variables is assessed, showing that using the clear sky irradiance as input improves model performance with 16% and that meteorological data does not improve performance. Additionally, the model performance was evaluated during different sky conditions showing that the deep learning model outperforms the persistence model for all sky conditions, except overcast conditions. An example of the model behavior is extensively described, showing that the deep learning model tends to predict the trend of the irradiance fluctuations rather than the actual fluctuations. Next to that is in this study shown that the current deep learning model occasional miss important weather events, like obscuration of the Sun, resulting in large irradiance prediction errors. A pathway for future improvements for deep learning models to forecast the short-term irradiance is provided.