A Novel High-Dimensional Trajectories Construction Network based on Multi-Clustering Algorithm

Abstract

A novel marine transportation network based on high-dimensional AIS data with a multi-level clustering algorithm is proposed to discover important waypoints in trajectories based on selected navigation features. This network contains two parts: the calculation of major nodes with CLIQUE and BIRCH clustering methods and navigation network construction with edge construction theory. Unlike the state-of-art work for navigation clustering with only ship coordinate, the proposed method contains more high-dimensional features such as drafting, weather, and fuel consumption. By comparing the historical AIS data, more than 220,133 lines of data in 30 days were used to extract 440 major nodal points in less than 4 minutes with ordinary PC specs (i5 processer). The proposed method can be performed on more dimensional data for better ship path planning or even national economic analysis. Current work has shown good performance on complex ship trajectories distinction and great potential for future shipping transportation market analytical predictions.

Container Ship Carbon and Fuel Estimation in Voyages Utilizing Meteorological Data with Data Fusion and Machine Learning Techniques

“…When facing diferent weather conditions, the sailing status and behavior of the ship will change to a certain extent, thus afecting the fuel consumption of the ship. In previous studies, weather data are generally collected in the form of daily newspapers, and when building the corresponding fuel consumption model, the weather data are large in granularity, and a single fuel consumption model cannot respond to the diferent fuel consumption efects brought by weather changes during ship navigation [18][19][20][21][22][23]. Since the Shanghai Meteorological Bureau provides real-time maritime weather data, it is benefcial to classify diferent weather conditions.…”