BEJARANO PELLICER, GUILLERMO, Gantiva-Osorio M. , MILLÁN GATA, PABLO, Camacho E.F.
No
Ocean Eng.
Article
Científica
1
1
01/01/2026
001826438700001
2-s2.0-105044690547
This work presents a hierarchical path-following control strategy for underactuated autonomous surface vehicles based on non-linear model predictive control. Traditional cascade controllers, including Line-of-Sight variants, often treat lateral velocity (sway) as an auto-induced disturbance, resulting in oscillations and degraded performance for vessels with non-diagonal dynamic matrices. This work proposes exploiting sway as an active variable to enhance convergence to the desired path. Acknowledging the significant computational demands of non-linear predictive controllers, the proposed approach divides the control problem into two layers, each with a simpler hierarchical predictive controller, thereby substantially reducing the required computational load. By leveraging prior knowledge of the desired path and forward velocity (or surge) at the high-level layer, and by enforcing constraints on control inputs and their rates at the low-level layer, the proposed control strategy ensures feasible low-level actions compatible with real actuators. The advantages and performance improvements offered by the proposed strategy, compared to traditional cascaded and state-of-the-art controllers, are demonstrated through curvilinear path-following simulations under realistic environmental disturbances and noisy measurements, including a hardware-in-the-loop implementation that also accounts for modelling parametric uncertainty. © 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Autonomous vehicles; Cascade control systems; Controllers; Predictive control systems; Uncertainty analysis; Autonomous surface vehicles; Cascade controller; Control strategies; Marine robotics; Model-predictive control; Nonlinear model predictive control; Path following; Path following control; Predictive control; Underactuated; control system; noise; nonlinearity; parameterization; path analysis; performance assessment; robotics; ship design; vessel; Model predictive control