Data-driven Learning and Control - DDLC

Model-based control has played a vital role in many branches of engineering and science. The purpose of this talk is to present a different paradigm for control systems design, particularly in situations where accurate models are difficult to obtain. Inspired by emerging applications in artificial intelligence and autonomous systems, learning-based control aims to develop computationally efficient and analytically tractable reinforcement learning based control algorithms with rigorous guarantees of stability and robustness.

In this talk, I will first discuss early developments in learning-based control for continuous-time linear and nonlinear systems with unknown dynamics. I will then present recent results on the robustness and resilience of learning-based controllers. Finally, I will illustrate the effectiveness of learning-based control through applications to autonomous vehicles and biological motor control.