Model predictive control for discrete-event and hybrid systems – Part I: Discrete-event systems

Model predictive control (MPC) is a very popular controller design method in the process industry. A key advantage of MPC is that it can accommodate constraints on the inputs and outputs. Usually MPC uses linear or nonlinear discrete-time models. In this paper and its companion paper (“Part II: Hybrid Systems”) we give an overview of some results in connection with MPC approaches for discrete-event systems and hybrid systems. In general the resulting optimization problems are nonlinear and nonconvex. However, for some classes of discrete-event systems and hybrid systems tractable solution methods exist. In this paper we consider discrete-event systems, i.e., asynchronous systems with event-driven dynamics. In particular, we discuss MPC for a special class of discrete-event systems, viz. max-plus-linear discrete-event systems, for both the noise-free and perturbed case (i.e., with modeling errors and/or noise). In the companion paper we will discuss MPC for some classes of hybrid systems.