Framework for detecting abnormal behaviors of passenger ships: A case study from the Yangtze River Estuary

Navigation safety of passenger ships has a great social impact due to the potential of large casualties, and understanding their behavior patterns is essential for identifying potential threats. This paper proposes a framework to detect passenger ships’ abnormal behaviors based on multi-source data, by considering typical classification of abnormal behaviors for passenger ships navigating in complex waters. The framework includes data collection and processing, abnormal behavior detection and recommendation. The characteristics of ship behavior patterns are analyzed through the synchronization of AIS data, ship type data and environmental data, trajectory simplification and regional segmentation. Then, four types of abnormal behaviors of passenger ships in specific waters, namely abnormal course, abnormal ship spacing, abnormal speed and time anomaly, are detected, respectively. A case study conducted in the Yangtze River Estuary demonstrates the effectiveness of the proposed framework. Based on the findings, this paper suggests several control measures such as crowded waters control, speed supervision, overtaking limitation, etc., to reduce the occurrence of abnormal behaviors of passenger ships and improve navigation safety.