Automatic berthing control for ASD tugs based on system identification and super-twisting sliding mode control

Harbor tugs play a crucial role in port operations, including pushing, towing, and assisting the berthing of large cargo vessels. However, autonomous berthing remains a complex and demanding task, as it requires precise control of speed, heading, and position under challenging conditions such as reduced rudder effectiveness at low speeds, ship–bank interactions, and environmental disturbances including wind and currents. For Azimuth Stern Drive (ASD) tugs in particular, the development of automatic control strategies remains challenging due to the high cost of obtaining high-fidelity models and the risk of actuator wear induced by control chattering. To address these issues, this paper proposes a deployment-oriented automatic berthing control framework. First, a low-speed maneuvering model is constructed and identified using a minimally sufficient dataset, demonstrating that reliable modeling can be achieved with reduced data acquisition effort. Second, a multi-stage trajectory planning method is introduced to generate reference trajectories, which are transformed into control references via the Line-of-Sight (LOS) strategy. Third, a human-inspired control structure is introduced to decouple the dual-azimuth thrusters, thereby simplifying the thrust allocation problem. A Super-Twisting Sliding Mode Controller (ST-SMC) is then developed to deal with environmental disturbances while eliminating chattering, resulting in smooth and actuator-friendly control actions. Full-scale simulations conducted on a high-fidelity marine simulator demonstrate the accuracy and robustness of the proposed approach, indicating its applicability to practical marine engineering scenarios.