Predictive Control for Solar Radiation Forecasting: A Case Study
Bildiriler > Tam Metin Bildiri
Yazarlar: Mehmet Çeçen, Marco Rivera, Patrick Wheeler, Emrah ZERDALİ
2025 Energy Conversion Congress & Expo Europe (ECCE Europe), , , Uluslararası, Tam Metin Bildiri, ss 1-5
DOI: 10.1109/ecce-europe62795.2025.11238877
Özet
This study proposes a model predictive control (MPC) for solar radiation forecasting. Accurate estimation of solar radiation is crucial for planning and organizing solar power systems. The accuracy of predictions is frequently influenced by the varying characteristics of the system's location, and meteorological conditions. Although MPC is traditionally used in control systems, in this study MPC principles adapt for prediction of solar radiation. The implementation uses a simplified version of MPC rather than solving the full optimization problem. MPC results are then compared with machine learning (ML) model that is convolutional neural network (CNN). By systematically exploring different parameter configurations and comparing performance across multiple metrics, the approach identifies optimal settings for accurate forecasting. The comparison with CNN models provides valuable insights into the relative strengths of different predictive modelling techniques for renewable energy (RE) applications.