At EKO Instruments, we proudly support research that improves the understanding and prediction of solar energy. In the article “All-Sky Imaging-Based Short-Term Solar Irradiance Forecasting with Long Short-Term Memory Networks”, published in Solar Energy (Vol. 272, 2024), researchers from Utrecht University — N.Y. Hendrikx, K. Barhmi, L.R. Visser, T.A. de Bruin, M. Pó, A.A. Salah, and W.G.J.H.M. van Sark — collaborated with EKO Instruments. They explored how all-sky imaging solar irradiance forecasting can enhance renewable energy reliability and grid management.
The team used EKO’s ASI Sky Cameras at the Plataforma Solar de Almería (Spain). These devices captured high-resolution images of the sky. They analyzed cloud movements and patterns to gain insights into solar irradiance. Additionally, the images were combined with data from EKO solar sensors measuring Global Horizontal Irradiance (GHI), temperature, and humidity. This created a robust dataset for training Long Short-Term Memory (LSTM) neural networks. As a result, the model successfully predicted short-term solar irradiance up to 20 minutes ahead. This achieved a notable improvement in forecast accuracy compared with traditional persistence methods.
This study shows how EKO Instruments’ all-sky imagers and solar sensors support the development of next-generation solar forecasting systems. By providing accurate, high-frequency irradiance and sky image data, EKO products enable researchers and energy professionals to model and anticipate solar power fluctuations. Consequently, this leads to more efficient integration of photovoltaic systems and a more stable, sustainable energy grid.
Read the full paper here: All-Sky Imaging-Based Short-Term Solar Irradiance Forecasting with Long Short-Term Memory Networks – Solar Energy (2024)
