TUNING OF MODEL-PREDICTIVE CONTROLLERS FOR ROBUST PERFORMANCE

Abstract

This paper presents practical on/off-line tuning rules for state-space MPC controllers. Based on the frequency-domain analysis of the closed-loop behavior of MPC controllers, the effect of various tuning parameters on the closed-loop system's performance and robustness is characterized and quantitative guidelines on how these parameters are best determined are established. It is shown that the choice of tuning parameters as well as their settings play a critical role in the overall robustness of the resulting closed-loop system and the ease of design and tuning. Several examples are presented to demonstrate how the proposed tuning rules provide insights and guidelines for tuning MPC controllers so that required performance and robustness are achieved.