基于层次聚类自动巡航的港区船舶碰撞危险识别方法研究

摘要

港区船舶碰撞危险的自动巡航分析和识别对于港区交通监管具有重要意义.针对船舶风险监管分析的时间提前量需求,引入DCQA和TCQA,综合选用DCPA和TCQA作为提前操作的判别参数组,结合AGNES算法,提出1种基于层次聚类自动巡航的碰撞危险分析方法,该方法对辖区内船舶实时动态数据进行聚类,最后识别出存在碰撞危险的船舶流区域,以及该会遇区域中存在碰撞风险的重点船舶.%The automatic patrol analysis and identification of ships in port area play an important role in traffic supervision and management.For the existing problems in traffic supervision, by introducing DCQA and TCQA and selecting DCPA and TCQA as the parameter set of operation ship in advance, a collision risk automatic patrol method is proposed to cluster the real-time dynamic data in precinct combined with AGNES algorithm.This method can identify the zones and ships with collision risk and provide visualization aided supervision.