Sampled-Data Model Predictive Control

This article focuses on model predictive control (MPC) design in the context of sampled-data control systems with full-state measurements. It is shown that recent results on this area can be successfully generalized to cope with sampled-data MPC. The open-loop plant is subjected to polytopic parameter uncertainty and at sampling times, a controlled output variable satisfies a set of convex constraints. A guaranteed H-2 performance index with infinity horizon is minimized such that the feedback control preserves asymptotic stability and feasibility. The design conditions are expressed through differential linear matrix inequalities. Continuous-time systems are treated with no kind of discrete-time modeling approximation. Comparisons with classical methods from the literature dealing with continuous-time systems are presented and discussed. Examples are included for illustration.

Automated summary

This article focuses on the design of model predictive control (MPC) in sampled-data control systems with full-state measurements. The results show that recent research in this area can be successfully applied to cope with sampled-data MPC. By minimizing a guaranteed H-2 performance index with infinity horizon, the feedback control preserves asymptotic stability and feasibility. The design conditions are expressed through differential linear matrix inequalities and no discrete-time modeling approximation is used for continuous-time systems. Comparisons with classical methods from previous literature are presented and discussed.