Systematic Methodology for Designing the Control and Communication Modules of a Network of Agents
Prof. Nuno Martins
 |
Dr. Nuno Martins
|
Motivated by a wide class of applications, the design of systems consisting of more than one agent with individual control and communication capabilities is a topic of intense research. Prof. Martins' current research is about the development of a systematic methodology for designing the control and the communication modules of a network of agents, where the agents are connected via a graph of point-to-point communication channels. Prof. Martins believes that a systematic design method can be obtained by extending recent results in real-time communication. The reasons for adopting real-time communication as a starting point are twofold: (1) The design paradigm adopted in this proposal can be cast as an extension of the real-time communication formulation; (2) New approaches have recently been reported that lead to optimality results for certain real-time communication schemes. This research will provide new methods of systematic design that ultimately will lead to a deliverable software toolbox, while contributing to the understanding of problems involving control and communication.
Prof. Martins' long terms goals are to establish a research program in the interface between control and information theory, with applications to decentralized and networked control, biological control systems, and applications of control to information theory. Recently, Prof. Martins has established the equivalence between fundamental limits of control and a new conservation law of entropy in feedback systems. By adopting an information theoretic framework, Martins has extended the aforementioned results, leading to the characterization of the fundamental limits of networked control systems containing feed-forward as well as feedback loops.
 return to spotlight on research
|
