Hierarchical robust control for autonomous berthing of MASS under dynamic switching of multiple propulsion modes

To address the coupled effects of dynamic propulsion-mode switching and strong environmental disturbances during the berthing of Maritime Autonomous Surface Ships (MASS), this paper proposes a hierarchical robust control architecture. First, a quintic-polynomial-based reverse spatial-temporal planning algorithm reformulates the two-point boundary-value problem as an initial-value manifold-tracking problem, avoiding the convergence difficulties of conventional forward-search methods and ensuring terminal-pose reachability. Second, a decoupled adaptive integral line-of-sight (DAI-LOS) guidance law with a virtual time reference decouples the longitudinal speed command from the unintended influence of lateral drift. Third, a dynamic-truncation sequential quadratic programming (SQP) allocation strategy addresses rank deficiency in the thrust allocation matrix caused by bow-thruster deactivation and supports smooth actuator-command transitions upon reactivation. Linear active disturbance rejection control (LADRC) provides disturbance rejection in the motion-control layer, while a control barrier function (CBF)-based safety filter provides hull-to-quay collision avoidance. Under the tested wind-current conditions with a Beaufort force 7 crosswind, the proposed architecture achieves sub-meter terminal berthing accuracy, reduces azimuth-angle total variation and yaw-rate standard deviation by approximately 21.5% and 13.8%, respectively, relative to conventional SQP allocation, and prevents shoreline contact under a sudden onshore disturbance.