Convergence of AI- and Physics-based Approaches in Power System Analysis, Optimization, and Control

For several decades, power systems have relied on physics-based methodologies for their analysis, optimization, and control. However, the recent surge in renewable energy sources, energy storage, behind-the-meter resources combined with grid digitalization has instigated a transformative shift in the dynamics and operation of these systems. Consequently, there's a burgeoning interest in data-driven and machine learning techniques to address the escalating scale, intricacy, and unpredictability of power systems. Yet, solely relying on data-driven strategies can pose challenges related to data quality, robustness, and interpretability. Integrating both physics-based models and data-driven or AI-based methods can harness the strengths of each approach and have resulted in notable enhancements in power system analysis, optimization and control. This panel seeks to share insights and recent progresses on the amalgamation of these two methodologies with speakers from academia and industry, enabling a reliable and smooth transition from contemporary power systems to the new electric system.