Research On Major Characteristics Of Marine Traffic Based On AIS Data

Due to the economic development,the marine traffic is more and more busy.Because of the limited resources in navigable waters,the increase in the number of ships and the increase in the traffic complexity will intensify the risk of marine traffic.However,it is not easy for the competent authorities to quickly and accurately identify the marine traffic situation and effectively carry out traffic organization.With the mandatory deployment and extensive use of AIS equipments,ship dynamic data has become a rich and cheap information source.Therefore,it is necessary to analyze the massive dynamic data of ships and fully explore the marine traffic characteristics,so as to provide theoretical basis and technical support for the competent authorities to improve the efficiency of traffic organization and to reduce the risk of marine traffic.Based on the AIS data,this paper takes the major characteristics of marine traffic as the research object and computer data mining as the research means to analyze the macro characteristics of marine traffic flow,spatio-temporal pattern and complexity of marine traffic.The main research contents and methods include:(1)The characteristics and parameters of marine traffic flow are analyzed.By the constraint of attributes such as speed,heading and MMSI,DBSCAN algorithm is used to perform constraint clustering analysis on a large number of AIS data to distinguish different types of traffic flows,the traffic flow parameters are calculated and the relationship between the parameters is studied.(2)AIS data are reduced by time slices to deal with the problem of unequal report interval.Taking temporal and spatial attributes into consideration,ST-DBSCAN algorithm is used to carry out spatio-temporal density clustering of AIS data.Multi-ship encounter situation recognition and spatio-temporal correlation analysis based on Apriori algorithm are carried out.(3)Considering factors such as ship length,relative distance,movement trend and crossing angle,the complexity model of traffic relationship unit is proposed to define the order distance.The AIS data are clustered using the OPTICS algorithm to calculate the complexity of marine traffic.The studies on the macro characteristics,spatio-temporal pattern and complexity of marine traffic enrich the theories and technologies of marine traffic.The case analysis based on AIS data in zhoushan area shows that the constrained clustering can distinguish traffic flows of different types more precisely and flexibly,and extract the macro characteristics of marine traffic flows.Spatio-temporal clustering and correlation analysis are more helpful to find the hidden spatio-temporal patterns.The complexity of marine traffic calculated by clustering after considering ship encounter situation,truly reflects the cognitive load of VTS operators.The research results are important and valuable for traffic organization and safety guarantee.