Hybrid Systems: Modelling and Control

Dr John Lygeros (Cambridge University Engineering Department)

Abstract

Hybrid systems are dynamic systems that involve the interaction of continuous dynamics (typically modeled by differential or difference equations) and discrete dynamics (typically modeled by finite state machines or other models of computation). Hybrid systems have been the topic of intense research activity in recent years, primarily because of their importance in applications. These include hardware and embedded software verification, mobile communication networks and large scale, multi-agent systems, such as air traffic management, automated highways, power networks, and Uninhabited Aerial Vehicles. In this talk, a mathematical language for modeling hybrid systems will be introduced. Based on this language, a procedure for designing hybrid controllers will then be developed. The control objective is to guarantee that the closed loop system will satisfy a set of desired safety specifications. The proposed design procedure builds on results from game theory and optimal control for purely discrete and purely continuous systems. The discussion will be motivated by examples from Automated Highway Systems and Air Traffic Management.


Back to Control Seminars Page.