基于AIS信息的船舶轨迹聚类模型及应用  

Ship Trajectory Clustering Model Based on AIS Data and its Application

作　　者：肖潇[1] 邵哲平[1] 潘家财[1,2] 纪贤标[1,2] 

机构地区：[1]集美大学航海学院,福建厦门361021 [2]厦门大学信息科学与技术学院,福建厦门361011

出　　处：《中国航海》2015年第2期82-86,共5页Navigation of China

基　　金：福建省教育厅科技计划项目(JB11103);浙江省交通厅科技项目(2012W14)

摘　　要：为获得船舶典型的运动模型,及时发现船舶异常轨迹并对其进行有效监控和管理,进而实现海上智能交通,基于船载AIS蕴藏着大量的海上交通特征的特点,从中获取能够反映船舶行为规律的有效的、潜在的信息。根据海上交通工程理论和数据挖掘技术,利用AIS信息并结合轨迹聚类算法,完成对已有轨迹的聚类,从中获取船舶典型的运动轨迹。以厦门港主航道及闽台直航船为实例,通过构建相应的AIS数据库并对船舶轨迹进行聚类结果展示,获得该海域船舶典型的运动轨迹。To monitor the traffic and timely find abnormal ship trajectory is important for intelligent maritime traffic management. In order to do this, to know the typical ship motion trajectory is the premise. Huge AIS data collected from ship broadcasts contain the information we need. The typical ship trajectory from AIS information is determined by means of the trajectory clustering algorithm according to the marine traffic engineering theory and data mining techniques. The trajectories of direct navigating ships between Fujian and Taiwan on the main channel of Xiamen are processed to show the feasibility of proposed method.

关 键 词：水路运输 船舶自动识别系统信息 数据挖掘 轨迹聚类 

分 类 号：U675.7[交通运输工程—船舶及航道工程] TP391[交通运输工程—船舶与海洋工程]