Optimal Route Selection for Berthing and Unberthing Based on AIS Data

In this paper, by leveraging the data from the Ship Automatic Identification System (AIS), a clustering model for ship berthing and unberthing trajectories is chosen. The optimal route for ship berthing and unberthing is determined via the clustering outcomes. This paper extracts representative trajectory points from a large quantity of data through the preprocessing, cleaning, trajectory interpolation, and compression of AIS data. Subsequently, it conducts a profound analysis of the ship's entry and exit trajectories and route selection strategies in specific port berths. By applying the DBSCANSD algorithm for cluster analysis of the berthing trajectories, the trajectory band of the ship berthing and unberthing is obtained. Finally, the Gaussian kernel density estimation method is utilized to select high-density data points. Thereafter, the screened data points are sorted, and the univariate spline interpolation is used to fit a smooth main trajectory line, which constitutes the optimal route for the berthing and unberthing of ships.