ECE Seminar: Advancing Control Theory for Autonomous Systems

Shaoshuai Mou, Purdue University

Monday, March 3 at 4:00pm

DL 514 or Zoom (https://yale.zoom.us/j/93238198353)

Hosted by: Professor Steve Morse

Abstract:

Modern society relies on a vast number of intelligent and autonomous systems, ranging from individual systems (such as robotic arms that can operate independently in manufacturing), human-autonomy teaming (such a robot to work as a teammate of human operator for sophisticated tasks) , to multi-agent systems (such as swarms of drones that can work as a cohesive whole for search and rescue). In this seminar we will present our recent research in advancing control theory with progress in optimization, networks and learning to address fundamental challenges in enabling autonomous systems to be optimal, adaptive, cooperative and swarming. The seminar will include fundamental progress to solving inverse optimal control for learning objective functions; a theoretical framework to adjust closed-loop optimal control with respect to parameter change; an end-to-end control/learning framework for autonomous systems to be adaptive to additional missions, a fundamental framework for autonomous systems to integrate inputs from human operators, and a series of distributed algorithms for computation, optimization and resilience in multi-agent systems.

Bio:

Shaoshuai Mou is the Elmer Bruhn associate professor in the School of Aeronautics and Astronautics at Purdue University. He received a Ph.D. in Electrical Engineering at Yale University in 2014, and then worked as a postdoc researcher at MIT for a year. He joined Purdue University as a tenure-track assistant professor in 2015, and was promoted to be Associate Professor with Tenure in 2021. His research group Autonomous & Intelligent Multi-agent Systems (AIMS) lab has been focusing on advancing control theory with recent progress in optimization, networks and machine learning for autonomous and robotics systems, with particular research interest in inverse optimal control for learning-from-demonstrations in robotics, parameter adaptation in optimal control, integration of control with learning, human-robot teaming, and distributed algorithms for control and optimization in multi-agent systems. Mou co-directs Purdue’s Institute for Control, Optimization and Networks (ICON) , consisting of about 100 faculty members from more than 15 departments across Purdue University, which aims to provide a research and education platform for control of autonomous and robotics systems.