Research On Node Recognition In Marine Traffic Complex Networks

The complexity of marine traffic is one of the important research directions in the field of Marine traffic situation analysis research.Most of the existing studies on the complexity of marine traffic has overlooked the differences in ship relative azimuth distribution and ship individual behavior complexity,making it difficult to reasonably and accurately describe ship encounter scenarios in different situations,consider the complexity differences between different ships and conducts a thorough analysis of existing marine traffic complexity models and complex network models,which is of great significance for the research of maritime traffic complexity recognition,the main work of this thesis is summarized as follows:(1)The research on ship spatial clustering based on DBSCAN.With the issue of inaccurate identification of ship encounter using fixed Eps neighborhood parameters in existing methods,this thesis has proposed an improved DBSCAN ship spatial clustering method based on a comprehensive consideration of ship size and handling characteristics.(2)Establishment of a marine traffic complexity model.In response to the problem that existing waterborne traffic complexity models cannot comprehensively and accurately describe marine traffic unit in different situations,considering factors such as ship length,relative distance,DCPA,relative motion trend,and relative azimuth distribution,a marine traffic complexity model based on the perspective of the own ship is constructed.(3)Establishment of the key node’s identification model for Maritime Traffic Complex Networks.To identify highly complex vessels within the entire regulatory waters,this thesis is based on the marine traffic complexity model,a marine traffic complexity network is constructed.This thesis comprehensively considers the node degree and comprehensive strength characteristics of ship nodes,as well as the direction of complex network connectivity,and proposes a method for identifying key vessels in marine traffic complex networks based on node importance contribution.(4)This thesis takes the east side of Laotieshan Waterway as the research area.Firstly,based on the improved DBSCAN spatial clustering method,all ships in the water area are spatially clustered.The node contribution recognition method was used to identify key ships in the water area,and the effectiveness of the recognition model was verified through the analysis of the resilience of the complex network.This thesis proposes a marine traffic complexity model that can effectively describe the degree of mutual influence between two ships under different encounter situations.By constructing a marine traffic complex network and effectively identifying key vessels in the complex network,it provides decision support for regulatory personnel to control the traffic of vessels with complex traffic relationships in the water area.