Framework Extraction Of Vessel Traffic Flow Based On AIS Data

Maritime transportation has always played an important role in global economic and trade activities.Since the implementation of the strategy of maritime power,maritime transportation safety has been paid more and more attention by people.With the rise of artificial intelligence and big data technology,people are eager to process and analyze the AIS trajectory data,in order to mine the ship motion characteristics and the potential laws of the marine traffic environment contained in it,so as to ensure the safety of navigation.In order to improve the safety of ships entering and leaving traffic-intensive areas and solve the problem of insufficient AIS data mining,this thesis uses clustering algorithm and data processing technologies to propose a diversified method for extracting ship traffic flow framework.The main research contents are as follows:(1)Preprocess the decoded AIS trajectory data,including data cleaning and coordinate transformation.The ship trajectory data of Laotieshan is analyzed,the abnormal trajectories are screened out,the characteristics of different noise points are studied,and they are classified.After processing,clean AIS data is obtained,which improves the data quality.(2)A density clustering algorithm based on trajectory points is designed.A distance threshold is set,the number of points in the neighborhood of the trajectory points is counted to obtain the point density.In order to avoid the clustering of trajectory points exceeding the threshold,the method of preventing multiple iterations in the clustering process is adopted.After multiple clustering,the density point network of ship traffic flow is obtained.(3)Extraction of turning point and route intersection point: using Douglas-Peucker compression algorithm to extract track turning point.According to the mathematical vector theory,the method of judging the intersection of two line segments is designed;for many ship track lines in Laotieshan,the three-level judgment mechanism of line-to-line,line-to-segment,and segment-to-segment is adopted,and finally the vector method is used to judge the two tracks.Segments have no intersections.This method avoids taking up too many resources and improves the algorithm efficiency of extracting intersection points.(4)The weighted fusion of feature points to condense the framework of ship traffic flow:The threshold selection in the clustering process of different feature point densities is studied,the weight adjustment and basis of different feature points in the fusion process are analyzed;Through weighted fusion,multiple feature points are obtained,and their importance is represented by the size of the points to generate the ship traffic flow framework in the Laotieshan waters.The framework integrates a variety of track feature points,can display the important track distribution in the nearby waters,and fully reflects the overall situation and dense areas of ship traffic flow;the framework condenses the habitual routes of ships in this water area from a statistical point of view.The route has a good airworthiness,which can be used for route planning and can also provide a reference for the maritime department to select a recommended route.