Structured Control and Learning for Networked Systems

ABSTRACT

Learning-based approaches are increasingly popular for discovering optimal control policies in a model-free manner. However, their lack of provable guarantees has limited their adoption in high-stakes real-world applications such as power systems and traffic networks. This talk will describe how to bridge the gap between learning and safety-critical constraints through structured neural networks guided by control theory and the physics of networked systems. Using Lyapunov theory, Dr. Cui will show how we can extract stabilizing controller structures for transient stability problems, and show how to parameterize the structures by neural networks. The structured approach provides end-to-end stability guarantees that are independent of the learning process, which in turn enables greater flexibility in designing decentralized and efficient learning algorithms.

BIOGRAPHY

Wenqi Cui is the Claire Booth Luce Assistant Professor in the Department of Electrical and Computer Engineering at New York University. Before joining NYU, she was a postdoctoral fellow in the Department of Computing + Mathematical Sciences at the California Institute of Technology. She completed her PhD in Electrical and Computer Engineering at the University of Washington in 2024. Her research interests are in the power and networked systems, from the perspective of control, machine learning, and optimization. She was a recipient of the Pioneer Postdoctoral Fellowship, PIMCO Postdoctoral Fellowship, Rushmer Innovator Fellowship, Sarala Vadari Award, etc. She has participated in the Rising Stars in EECS Workshop in 2022 and the Rising Stars in Cyber-Physical Systems Workshop in 2023.