On Stabilizing Terminal Costs and Regions for Configuration-Constrained Tube MPC

This paper introduces a novel class of terminal regions and cost functions for tube model predictive control (TMPC). Our focus is on polytopic configuration-constrained TMPC schemes, which offer flexibility by introducing a significant amount of variables to model the shape of the propagated sets. This flexibility, however, comes with a challenge, namely, to enforce stability efficiently. To address this challenge, we propose tailored terminal regions and cost functions enabling efficient and stable TMPC implementations, without relying on regularity assumptions about the control system or configuration templates. Numerical case studies demonstrate the effectiveness and performance of the proposed control scheme.