中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊

重庆交通大学学报(自然科学版) ›› 2018, Vol. 37 ›› Issue (10): 117-122.DOI: 10.3969/j.issn.1674-0696.2018.10.18

• 交通运输工程 • 上一篇    下一篇

基于时空密度的船载AIS数据聚类分析方法研究

李永攀1,2,刘正江1,郑中义1   

  1. (1. 大连海事大学 航海学院,辽宁 大连 116026;2. 舟山海事局 船舶交通管理中心,浙江 舟山 316000)
  • 收稿日期:2017-05-02 修回日期:2017-10-16 出版日期:2018-10-08 发布日期:2018-10-08
  • 作者简介:李永攀(1981—),男,河南开封人,博士研究生,主要从事海上交通安全方面的研究。E-mail: liyongpan@dlmu.edu.cn。 通信作者:刘正江(1959—),男,江苏如皋人,教授,博导,主要从事海上交通安全方面的研究。E-mail: liuzhengjiang@dlmu.edu.cn。

Clustering Analysis Method of Shipborne AIS Data Based on Spatio-Temporal Density

LI Yongpan1, 2, LIU Zhengjiang1, ZHENG Zhongyi1   

  1. (1. Navigation College, Dalian Maritime University, Dalian 116026, Liaoning, P. R. China; 2. Vessel Traffic Management Center, Zhoushan Maritime Safety Administration, Zhoushan 316000, Zhejiang, P. R. China)
  • Received:2017-05-02 Revised:2017-10-16 Online:2018-10-08 Published:2018-10-08

摘要: 船载AIS数据包含位置、时间和其他属性,属于典型的时空数据,对其开展时空聚类分析有助于挖掘海上交通特征。结合AIS数据的具体特 征,提出时间切片化方法,有效约简AIS数据并处理报告间隔不等的问题。在DBSCAN算法基础上综合考虑时间和空间要素,提出船载AIS数据时空 聚类算法,并对实际数据开展分析。该方法能更好顾及船舶交通流的时空耦合特征,识别隐含的时空模式,为主管机关开展船舶交通管理、优化 通航秩序、保障航行安全等提供一种新途径。

关键词: 交通运输工程, 时空密度, 聚类分析, DBSCAN算法, 船载AIS

Abstract: Shipborne AIS data contains location, time and other attributes, which belongs to typical spatio-temporal data. And the spatio-temporal clustering analysis on shipborne AIS data contributes to the data mining of maritime traffic characteristics. Considering the specific characteristics of AIS data, a time slicing method was proposed to effectively simplify the AIS data and deal with the problem of different report intervals. Based on DBSCAN algorithm, combining time and space attributes, a spatio-temporal clustering algorithm of shipborne AIS data was put forward, and the actual AIS data was taken for a case study. The proposed method can better take account of the spatial-temporal coupling characteristics of ship traffic flow and identify the hidden spatio-temporal pattern, which provides a new way for the competent authorities to carry out ship traffic management, optimize navigation order and ensure navigation safety.

Key words: traffic and transportation engineering, spatio-temporal density, clustering analysis, DBSCAN algorithm, shipborne AIS

中图分类号: