Multiscale Analysis and Control of Civil Infrastructure Systems: A Cyber-Physical Perspective

Friday, April 20, 2018, 11:00 am - 12:00 pm PDTiCal
6th floor large conference room
This event is open to the public.
AI Seminar
Ketan Savla, USC
Video Recording:

Increasing sensing, autonomy and connectivity are presenting novel opportunities for control of infrastructure systems at multiple spatial and temporal scales. Examples range from private incentives and controlling driverless vehicles at the individual level to coordinated traffic signal control at city-scale, and from controlling cascading failure over a short time horizon to network operations that deliver robust performance over the long run. We first provide an overview of certificates that provide insight into fundamental performance limits under these emerging control paradigms. Specifically, we advocate a cyber-physical approach to revisit and extend basic settings in a network flow, game theory, queuing systems, and combinatorial optimization in light of canonical physical constraints underlying civil infrastructure systems to construct such certificates.

We then consider the specific problem of computing optimal control of power networks during cascading failure. The underlying multistage problem is challenging due to the presence of discrete and continuous variables, corresponding to network topology and physical quantities respectively. We provide a methodology to construct an equivalent finite abstraction, similar to a decision tree. This construction leverages and extends concepts from the arrangement of hyperplanes – a computational geometric tool typically used in robot path planning.


Ketan Savla is an assistant professor and the John and Dorothy Shea Early Career Chair in Civil Engineering at USC. Before joining USC, he was a research scientist in the Laboratory for Information and Decision Systems at MIT. He received his Ph.D. in Electrical and Computer Engineering from the University of California at Santa Barbara. His current research interest is in distributed robust and optimal control, dynamical networks, state-dependent queueing systems, and incentive design, with applications in civil infrastructure and autonomous systems. His recognitions include NSF CAREER, an IEEE CSS George S. Axelby Outstanding Paper Award, and AACC Donald P. Eckman Award.


« Return to Upcoming Events