Decreasing-horizon Robust Model Predictive Control with Specified Settling Time to A Terminal Constraint Set

Robust model predictive control for discrete-time linear systems with norm-bounded disturbances is investigated in this pape1r. The control objective is to steer the system state t a terminal constraint set within specified number of steps. Meanwhile, the performance of the closed-loop control system is optimized. A decreasing-horizon predictive control strategy is proposed. Moreover, affine state-feedback control laws with memory of prior states are adopted over the prediction horizon. To optimize the system performance, an ∞-type cost function is considered in this paper. It is shown that finite settling time is achieved, if the optimization problem in the proposed control strategy is initially solvable. Some simulations are presented to show the effectiveness of the proposed control strategy.