A novel constraint tightening approach for nonlinear robust model predictive control

Abstract

Listen

Translate article

In this paper, we present a novel constraint tightening approach for nonlinear robust model predictive control (MPC). This approach uses a simple constructive constraint tightening based on growing tubes. Contrary to other approaches, we require no complex offline computations to obtain a stabilizing control law. Instead, we consider the notion of incremental stabilizability and design tubes based on an estimate of the achievable exponential decay rate. In addition, we show how this tightening can be used as an ad-hoc modification to improve the robustness of MPC without terminal constraints. We study the system theoretic properties of the resulting closed-loop system, including bounds on the region of attraction and the minimal robust positively invariant (RPI) set. Within an MPC framework without terminal constraints, the proposed constraint tightening leads to a nonlinear robust controller without complex design procedures, which makes it appealing for practical applications.