Vessels Trajectory Clustering And Its Application Based On C-OPTICS Algorithm

With tihe rapid development of the world economy,the maritime transportation has increased sharply and vessel transportation has been increasingly busier,which lead the traffic of important maritime arteries and port water becoming heavier,and bring the unprecedented pressure to the safe navigation of the vessel,the construction of the port and the maritime management.For the sake of ensuring the safety of maritime personnel and property,protecting the maritime environment and reducing the labor intensity of supervisors,it seems very necessary that use the information technology to perceive the maritime traffic situation and discover the abnormal situation of the vessel.The vessels trajectory clustering is the foundation of detecting vessel abnormal behavior and this article clusters the vessels trajectory based on C-OPTICS algorithm and detects three abnormal behavior of the vessels.The main works are as follows:(1)Decoding and preprocessing of AIS data.On the basis of introducing AIS information types,content and time updated of information,this article analyzes the composition of the original information and the main message structure of AIS,and decodes AIS information,then constructs AIS database,designs the scheme of cleaning and filtering AIS data,which ensure that we can obtain the clean and available AIS information.(2)Posing the algorithm of vessels trajectory clustering which uses the C-OPTICS algorithm based on the course constraints.Using Euclidean distance to measure trajectory similarity,designs the algorithm of identifying the moving point and stopping over point of the vessel and extracts AIS data of vessel moving point and stopping over point after cleaning;considering course factor,establishes the track mining process and implementation method of the vessel based on the algorithm of C-OPTICS;The AIS data of the waters outside the jiaozhou bay in Qingdao are used to verify the algorithm of this article,and eleven clusters is obtained through clustering,and that is corresponded with the actual vessel traffic flow pattern which indicating that the method is correct and feasible.(3)The application of vessels trajectory clustering in the vessel abnormal behavior detection.Combined with the electronic chart simulator platform,use the result of vessel trajectory clustering as the alarm threshold of the electronic chart simulator platform and detect the location,course and speed of the vessel.When an abnormal behavior occurs,an early warning can be provided.The Experiments show that the results of clustering can be used for vessel abnormal behavior detection.