Combining physics-based and data-driven modeling and simulation for power systems

Physics-based methods have been the fundamental modeling and simulation approaches in power systems for decades. However, the increasing scale, complexity, uncertainties in power systems lead to grand challenges in meeting the increasing planning and operation requirements of the grid with existing physics-based methods. Recent advancements of data-driven based methods provide promising solutions to achieve computationally efficient modeling and simulation with good accuracy. Yet, they usually suffer from data sampling efficiency, robustness and interpretability issues. Fusing the physical models and data-driven approaches can help leverage the advantages of the two, and it have achieved promising results in many domains, including power systems. In this panel, we invite researchers and engineers to present the latest research and progress in this area for different power system applications, ranging from power flow, dynamics to transient modeling and simulation. It will help share these new techniques, inspire new ideas and collaboration, and ultimately facilitate their adoptions in the industry to help better plan and operate the future grid.