Optimal Power Flow in a highly renewable power system based on attention neural networks

The Optimal Power Flow (OPF) problem is crucial for power system operations. It guides generator output and power distribution to meet demand at minimized costs while adhering to physical and engineering constraints. However, the integration of renewable energy sources, such as wind and solar, poses challenges due to their inherent variability. Frequent recalibrations of power settings are necessary due to changing weather conditions, which makes recurrent OPF resolutions necessary. This task can be daunting when using traditional numerical methods, especially for extensive power systems. In this work, we present a state-of-the-art, physics-informed machine learning methodology that was trained using imitation learning and historical European weather datasets. Our approach correlates electricity demand and weather patterns with power dispatch and generation, providing a faster solution suitable for real-time applications. We validated our method's superiority over existing data-driven techniques in OPF solving through rigorous evaluations on aggregated European power systems. By presenting a quick, robust, and efficient solution, this research establishes a new standard in real-time optimal power flow (OPF) resolution. This paves the way for more resilient power systems in the era of renewable energy.