Markov Actuator Assignment for Networked Control Systems

Abstract

This paper investigates control systems in which the number of actuators activated for performing control task at any time is limited. The activation status of the actuators is assigned in accordance to a random event (modeled by a Markov chain). Then under what condition is the system stable and how to design a controller to meet certain performance requirement? By taking into account the Markov chain and the transmission delay in the control signal, we establish a novel framework for stability analysis and controller synthesis. Specifically, a easily testable mode- and delay-dependent controller design method is given for uncertain linear systems. A numerical example is provided to show the advantage and usefulness of the proposed method.