Name
Abstract | ||
|---|---|---|
Trajectory analysis and extraction of trajectory patterns are crucial to enhance marine safety and marine status awareness. The major data source for such analysis is Automatic Identification System (AIS), which publishes data related to movement of the ship while cruising. AIS broadcasts information including type of ship, identity number, state, destination, estimated time of arrival (ETA), location, speed, direction, and cargo. In this paper, we focus on extracting a variety of trajectory patterns for maritime vessels. The first group of analysis we focus on is arrival port, arrival time, and next position prediction on AIS messages, which are useful to aid maritime operators. We propose three different approaches for the prediction of marine vessel movement. As the second type of analysis, anomaly detection on marine vessel trajectory is studied. For vessel movement prediction, the experiments show that the proposed solutions brought improvement against conventional supervised learning approaches. The proposed anomaly detection technique is demonstrated on a case study. |
Year | DOI | Venue |
|---|---|---|
2021 | 10.1016/j.iot.2021.100436 | Internet of Things |
Keywords | DocType | Volume |
AIS,Trajectory,Supervised learning,Arrival port prediction,Arrival time prediction,Next position prediction,Anomaly detection,LSTM,Maritime | Journal | 16 |
ISSN | Citations | PageRank |
2542-6605 | 0 | 0.34 |
References | Authors | |
0 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Gözde Boztepe Karatas | 1 | 0 | 0.34 |
| Pinar Karagoz | 2 | 0 | 0.34 |
| Orhan Ayran | 3 | 0 | 0.34 |