Power systems and machine learning
The transformation of modern power systems, driven by the evolving generation mix, presents significant challenges for grid stability, efficiency, and market operations. Our research addresses these challenges across four key areas. We develop models to assess failures in transmission grids, identify critical infrastructure, and design robust expansion strategies. We address the integration of distributed energy resources in power distribution grids, analyzing how their flexibility could be used for network expansion planning and ancillary services provision. Then, we study and model electricity markets to optimize bidding strategies, design new products, and assess market structures' impact on system security. Finally, we leverage advanced machine learning techniques to improve fault detection, uncertainty management, and optimization in energy networks.
Transmission grid analyses, planning, and reliability assessment
Modern power transmission systems face unique challenges facilitated by the changing generation mix, consumer profile, and weather patterns. Understanding the impact of these changes is crucial for ensuring the efficient operation of the system and the security of supply. To address these concerns, we develop models that simulate the propagation of failures across the transmission grid, identify critical components, and devise comprehensive expansion strategies.
Lab members: Anna Varbella, Ambra Van Liedekerke, Wanhong Yu, Dr. Andrej Stankovski, Dr. James Qin
Distribution grid analyses and planning
Our research on distribution grid analysis supports the energy transition by enabling greater integration of distributed energy resources (DERs) and unlocking flexibility opportunities. We explore how DERs can contribute to grid support, incorporating them into distribution network planning and ancillary service markets.
Lab members: Lorenzo Zapparoli, Alfredo Oneto
Electricity markets
The liberalized electricity market is the cornerstone of modern power systems. Governed by basic economic principles, the market has wide-spanning effects on power system security and economic prosperity. We develop models that capture the behavior of the electricity markets (day-ahead, futures, and ancillary services), construct optimal bidding strategies for market participants, and assess the impact of different market structures on system security.
Lab members: Lorenzo Zapparoli, Dr. Andrej Stankovski
Applied machine learning for engineering systems
Machine learning (ML) is a powerful tool for understanding complex interactions through observation. It enables challenging assessments, predictions and optimizations where conventional methods fail or are not affordable. We apply and develop advanced ML algorithms to handle faults, uncertainties, hazards and optimality in power grids, energy networks and critical engineering systems.
Lab members: Anna Varbella, Wanhong Yu, Dr. Enrico Ampellio, Dr. Laya Das