Robust neural path-following control for underactuated ships with the DVS obstacles avoidance guidance

摘要

In this note, one focuses on the waypoints-based path-following control of underactuated surface ships with the mechanism of multi-static or slow time-varying obstacles avoidance. In the scheme, an improved dynamical virtual ship (DVS) principle is initially developed to programme the real-time attitude guidance for the underactuated ship in marine practice, providing a smooth transition of heading angle and velocity based on the principle of proximity. The scheduler is applied in the path-following and obstacles avoidance missions. Furthermore, to ensure the effectiveness of the obstacles avoidance manoeuvering, a practical robust neural control is proposed by fusion of neural networks and the robust neural damping technique. It requires less (or no) information of the system parameters and structure, and only four adaptive parameters require to be updated online. These designs would facilitate the implementation of the algorithm in the practical engineering. Considerable efforts are made to obtain the semi-global finite-time uniformly bounded (SGFTUB) stability by employing the Lyapunov theory. The comparative experiments have been presented to verify the effectiveness of the proposed scheme.