A deep learning model using graph neural networks to forecast renewable energy production by modeling spatial and temporal relationships between energy plants and meteorological data.

A system that uses satellite imagery and convolutional neural networks to predict solar photovoltaic generation over short time horizons (minutes to two hours).