Receding-horizon robust online communication scheduling for constrained networked control systems

Abstract

This paper proposes a novel online scheduling approach for perturbed Networked Control Systems (NCS) over lossy networks, subject to state and input constraints and limited network capacity. The paper shows how an online communication schedule can be found for the considered class of systems, as the solution of a constrained optimization problem, which is solved in receding horizon and is proven to be persistently feasible. The resulting online schedule is proven to preserve the system invariance w.r.t. state and input constraints. A communication scheduler is also derived for lossy networks, starting from a feasible baseline schedule calculated offline. Numerical examples are given to illustrate the proposed contributions.