Hybrid system theory, applications, and experiments

点击翻译按钮获取中文摘要

This proposal concerns two areas of research: hybrid system theory and multivehicle systems. Hybrid systems are systems which exhibit both continuous time and discrete event behavior. This research area has emerged as a synergy between Computer Science and Control Theory in the last 15 years, and is motivated by a wide number of engineering applications that exhibit mixed continuous and discrete behavior, including: communication networks with dynamic agents at the physical layer, a discrete event supervisor regulating a manufacturing or chemical process; a continuous system with internal switching such as an automotive engine; a real-time system such as a paper copier with transitions marking events; a continuous process such as a temperature regulator with safety constraints that, when violated, trigger a mode switch, and so forth.  While there are manifest applications of the model, what is needed is a more versatile and mature theory for analysis and control synthesis.  The main objective of my research is to develop mathematical and computational tools for synthesis of controllers that satisfy so-called reachability specifications. Reachability specifications differ from the traditional control specifications of stabilization, fast response, and tracking in that they allow one to specify - in an untimed way - a sequence of behaviors that the system should exhibit. One immediate application area of the above proposed research is multivehicle systems,  which is the second component of the research. Here we are interested in investigating how distributed controllers achieve desired global behaviors for a collection of vehicles. The theoretical research is complemented by software development, experiments, and pedagogical demonstrations using an experimental testbed of ten vehicles, as well as pendulum and coupled tank experiments.