On Variation of Infinite Horizon Performance of Model Predictive Control with Varying Receding Horizon

Abstract

Abstract This paper investigates variation of infinite horizon (IH) performance of Model Predictive Control (MPC) without constraints as the optimization horizon changes. By exploring properties of the Difference Riccati Equations (DRE), an upper bound and a lower bound of the ratio between variation of IH performance of MPC and finite horizon (FH) optimal cost are obtained. The result shows the dynamic behavior of IH performance of closed-loop MPC systems as the optimization horizon varies.