Exploring Feasibility of Machine Learning for Grid Resilience Assessment

In recent years, extreme events such as droughts, wildfires, polar vortex, hurricanes, and cyberattacks have caused stressful conditions and even catastrophic failures of energy systems. Meanwhile, machine learning has been applied in many areas in power systems to address challenges which were difficult to solve in the past. This panel will focus on the application and feasibility of machine learning techniques to address extreme events from modeling, analysis, and mitigation perspectives to enhance power grid resilience. The topics in this panel include all participants in power system operation and planning, including but not limited to generation, transmission, distribution, microgrid, and customers. Presentations in this panel session: - Deep Reinforcement Learning-Based On-Line Dynamic Multi-Microgrid Formation to Enhance Resilience (23PESGM0849) - Data-Driven Modeling and Assessment of Outage and Resilience in Distribution Grids (23PESGM0852)