Globally Stable Sliding-Mode Predictive Control for ASVs Without Terminal Constraints: A Short Prediction Horizon Approach

Considering that the stability of traditional model predictive control (MPC) methods without terminal constraints depends on the prediction horizon length, this article proposes a globally stable sliding-mode predictive control (SMPC) method for autonomous surface vehicles (ASVs) that operates without terminal constraints and achieves stability with a relatively short prediction horizon. First, a novel sliding-mode surface with a saturation function is introduced to ensure the boundedness of sliding-mode variables, even when position errors exceed a predefined threshold. Second, the sliding-mode variables are incorporated into the cost function and constraints of the SMPC optimization problem, enabling stability with a relatively short prediction horizon. Third, the method for selecting the weight matrices of the cost function to achieve global stability within a finite time for the closed-loop system is provided. Finally, the effectiveness of the proposed algorithm is demonstrated through hardware experiments on ASVs.