中国安全科学学报 ›› 2019, Vol. 29 ›› Issue (3): 168-173.doi: 10.16265/j.cnki.issn1003-3033.2019.03.028

• 公共安全 • 上一篇    下一篇

基于流推理的定线制水域船舶航行安全评估方法

魏万淇1, 高曙**1 教授, 初秀民2 教授   

  1. 1 武汉理工大学 计算机科学与技术学院,湖北 武汉 430063;
    2 武汉理工大学 智能交通系统研究中心,湖北 武汉 430063
  • 收稿日期:2018-11-19 修回日期:2019-01-13 发布日期:2020-11-26
  • 通讯作者: **高 曙(1967—),女,安徽芜湖人,博士,教授,主要从事智能计算以及交通安全智能化技术等方面研究。E-mail:gshu418@163.com。
  • 作者简介:魏万淇 (1991—),男,四川犍为人,硕士研究生,主要研究方向为智能计算。E-mail: 2385951234@qq.com。
  • 基金资助:
    国家自然科学基金资助(51479155)。

Evaluation method of ship navigation safety in routing waters based on stream reasoning

WEI Wanqi1, GAO Shu1, CHU Xiumin2   

  1. 1 School of Computer Science and Technology, Wuhan University of Technology, Wuhan Hubei 430063, China;
    2 Intelligent Transport System Research Center, Wuhan University of Technology, Wuhan Hubei 430063,China
  • Received:2018-11-19 Revised:2019-01-13 Published:2020-11-26

摘要: 为了更有效地利用船舶航行过程中人员、船舶、环境、管理等数据,实时评估船舶航行安全状态,提出基于流推理(SR)的定线制水域船舶航行安全评估方法。首先,利用粗糙集条件信息熵(RSCIE)约简评估要素,结合专家知识,设计评估本体;然后,通过语义Web技术融合实时采集的相关流数据,用SR引擎推理船舶航行安全状态;最后,运用该方法评估南京下游定线制水域的船舶航行安全状态,并与基于云理论(CT)的安全评估方法等对比分析。结果表明:本文提出的方法具有更好的时效性,而且检测率高、误报率低。

关键词: 流推理(SR), 船舶安全评估, 本体, 语义Web, 规则推理, 粗糙集条件信息熵(RSCIE)

Abstract: In order to use the data of personnel, ship, environment and management during ship navigation more effectively and conduct real-time assessment of navigation safety, this paper proposed a method for evaluating ship navigation safety based on SR in routing waters. Firstly, the safety assessment elements of ship navigation were reduced based on RSCIE and the assessment ontology was constructed by combining expert knowledge. Then the ship navigation-related streaming data collected in real time was integrated by semantic Web technology, and the SR engine was used to infer the navigation safety. Finally, the ship navigation safety in a certain routing water system in Nanjing was evaluated and the method presented in this paper was compared with the other two methods based on cloud theory. The results show that the proposed method has better timeliness, higher detection rate and lower false alarm rate.

Key words: stream reasoning(SR), ship safety assessment, ontology, semantic Web, rule reasoning, rough set conditional information entropy(RSCIE)

中图分类号: