Ship behavior prediction via trajectory extraction-based clustering for maritime situation awareness

• Published in: Journal of ocean engineering and science Vol. 7; no. 1; pp. 1 - 13
• Main Authors: Murray, Brian, Perera, Lokukaluge Prasad
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.02.2022; Elsevier
• Subjects: AIS; Anvendt matematikk: 413; Applied mathematics: 413; Collision avoidance; Machine learning; Marin teknologi: 580; Marine technology: 580; Maritime situation awareness; Matematikk og Naturvitenskap: 400; Matematikk: 410; Mathematics and natural science: 400; Mathematics: 410; Ship navigation; Technology: 500; Teknologi: 500; Trajectory prediction; Unsupervised learning; VDP; Trajectory prediction; Maritime situation awareness; Ship navigation; Machine learning; AIS; Collision avoidance; Unsupervised learning
• ISSN: 2468-0133; 2468-0133
• DOI: 10.1016/j.joes.2021.03.001

•This study presents an approach to predict the future behavior of a vessel using historical AIS data. Such predictions can subsequently be utilized to evaluate the potential future collision risk, supporting proactive collision avoidance. In this manner, enhanced situation awareness can be facilitated.•By utilizing machine learning techniques, local behavior clusters can be extracted from historical AIS data to describe the possible future 30 min behavior of a vessel. These clusters are discovered using the Karhunen–Loeve transform and Gaussian Mixture Models.•The developed technique allows the vessel to be classified to a cluster of behavior, and conducts a trajectory prediction with respect to the behavior in this cluster.•The method is evaluated using test cases from a region surrounding Tromsø, Norway. The results indicate that the technique has good performance in predicting future ship behavior. This study presents a method in which historical AIS data are used to predict the future trajectory of a selected vessel. This is facilitated via a system intelligence-based approach that can be subsequently utilized to provide enhanced situation awareness to navigators and future autonomous ships, aiding proactive collision avoidance. By evaluating the historical ship behavior in a given geographical region, the method applies machine learning techniques to extrapolate commonalities in relevant trajectory segments. These commonalities represent historical behavior modes that correspond to the possible future behavior of the selected vessel. Subsequently, the selected vessel is classified to a behavior mode, and a trajectory with respect to this mode is predicted. This is achieved via an initial clustering technique and subsequent trajectory extraction. The extracted trajectories are then compressed using the Karhunen–Loéve transform, and clustered using a Gaussian Mixture Model. The approach in this study differs from others in that trajectories are not clustered for an entire region, but rather for relevant trajectory segments. As such, the extracted trajectories provide a much better basis for clustering relevant historical ship behavior modes. A selected vessel is then classified to one of these modes using its observed behavior. Trajectory predictions are facilitated using an enhanced subset of data that likely correspond to the future behavior of the selected vessel. The method yields promising results, with high classification accuracy and low prediction error. However, vessels with abnormal behavior degrade the results in some situations, and have also been discussed in this study.