Efficient iterative learning model predictive control for uncertain nonlinear discrete-time systems

Abstract

This work focuses on the iterative learning model predictive control (ILMPC) design for nonlinear discrete-time batch systems. Different from the existing results, a novel efficient two-dimensional (2-D) ILMPC approach is firstly proposed based on the 2-D system theory, which is able to guarantee the $H_{\infty }$ tracking performance with lower computation load. Furthermore, based on the newly established event-triggered mechanisms, an event-triggered 2-D ILMPC is developed to reduce the occupation of the network resources while ensuring the $H_{\infty }$ tracking performance. For the proposed ILMPC schemes, the sufficient conditions for the $H_{\infty }$ tracking performance are provided explicitly by employing the linear matrix inequalities (LMI) techniques. Finally, the effectiveness of the proposed ILMPC strategies are demonstrated through numerical simulations.