Artificial Intelligence for Microgrid Resilience: A Data-Driven and Model-Free Approach

Extreme weather events, which are characterized by high impact and low probability, can disrupt power system components and lead to severe power outages. The increasing adoption of renewable energy resources in the power sector, as part of decarbonization efforts, introduces further system operation challenges because of their fluctuating nature, potentially worsening the impact of these extreme weather events. To address the challenges from these high-impact and low-probability events, the concept of resilience has been introduced into the power industry. Considering the potential serious disruptions, the primary goal of resilient power system operation during extreme events is to ensure the continuous supply of critical loads, such as hospitals, police stations, data centers, traffic lights, etc., across various power sectors, which constitutes a system-wide load restoration problem.