Research on the Automatic Classification of Ship’s Navigational Status
Maritime traffic analysis has been attracted increasing attention due to their importance for the safety and efficiency of maritime operations. The first step of maritime traffic analysis is the identification of ships’ navigational status, and various analysis tasks are started based on the status information. It should be considered the complex traffic characteristics of the harbor and ships. These tasks depend on the expert’s experiences, however, it becomes difficult to classify manually as the amount of traffic volume increases. Therefore, in this paper, we proposed a new model to identify the ship’s navigational status automatically. The proposed method generated traffic pattern model using accumulated AIS trajectories and then classified using the clustering algorithm. This method based on semi-supervised machine learning and the proposed clustering method using the pre-classified dataset. Finally, we review experimental results using the actual trajectory data to verify the feasibility of the proposed method.
KeywordsNavigational status AIS Machine learning Clustering
This research was supported by a grand from Endowment Project of “Development of core technology for the analysis and reproduction of maritime accidents through simulations” funded by Korea Research Institute of Ships and Ocean Engineering (PES9350).