Warm-started Semionline Trajectory Planner for Ship's Automatic Docking (Berthing)

Abstract

In the usual framework of control, a reference trajectory is needed as the set point for a feedback controller. This leads to a problem on how to scientifically generate the reference trajectory. Such a problem is termed trajectory optimization which is done in a trajectory planner. It is desired that this reference trajectory is kept feasible with respect to the ship's dynamics, the dynamics of the environment (wind), and the surrounding boundary. For an underactuated conventional vessel, its mathematical model can be very intricate. This causes significant computational time. This article demonstrates that the balance between the feasibility of the reference trajectory and the computational time can be achieved for an underactuated vessel in a disturbed and restricted environment. This is done by: (1) using an almost-global offline solution as a warm start in a semionline trajectory optimization to speed up the calculation, (2) including the prediction of wind dynamics, and (3) using a predefined boundary to generate the necessary spatial constraints that guarantee a collisionfree trajectory. Incorporation of these three things results in a feasible and safe trajectory, with a considerable computational time speedup. In addition, the warm start generally gives better results than that without warm start.