Detection of Complex Abnormal Ship Behavior Based on Event Stream

In recent years, maritime traffic has increased dramatically. Some ships exhibit abnormal behavior, which may indicate different dangers or threats. The speed and position parameters are often used to detect abnormal ship behavior. However, such methods only detect part of the ship's abnormal behavior, and ignore the events before or after the abnormal behavior. This paper uses event stream processing to detect the complex abnormal behavior of ships. Event stream processing treats one abnormal behavior of a ship as an event, and then determines the complex abnormal behavior through the correlation of multiple events. This method can detect a series of abnormal ship behaviors, which helps to identify illegal behaviors of ships at sea and improve the safety of maritime navigation.