中国安全科学学报 ›› 2020, Vol. 30 ›› Issue (1): 128-135.doi: 10.16265/j.cnki.issn1003-3033.2020.01.020

• 公共安全 • 上一篇    下一篇

基于扩展式动态博弈的多船避碰决策模型

欧阳旭东1,2, 支云翔1,2, 王腾飞3, 吴兵1,2 副教授, 汪洋**1,2 副教授   

  1. 1.武汉理工大学 智能交通系统研究中心,湖北 武汉430063;
    2.武汉理工大学 国家水运安全工程技术研究中心,湖北 武汉430063;
    3.武汉理工大学 物流工程学院,湖北 武汉430063
  • 收稿日期:2019-10-20 修回日期:2019-12-20 出版日期:2020-01-28 发布日期:2021-01-22
  • 通讯作者: ** 汪 洋(1976—),男,湖北武汉人,博士,副教授,主要从事智能交通与安全研究等方面的研究。E-mail:wangyang.itsc@whut.edu.cn。
  • 作者简介:欧阳旭东 (1995—),男,江西抚州人,硕士研究生,研究方向为智能交通与安全E-mail: 13971381893@ 163.com。
  • 基金资助:
    国家重点研发计划专项(2018YFC0810400/05);中央高校基本科研业务费专项(2019III002GX)。

Antensive form game theory based multi-ship collision avoidance scheme

OUYANG Xudong1,2, ZHI Yunxiang 1,2, WANG Tengfei3, WU Bing 1,2, WANG Yang1,2   

  1. 1. Intelligent Transportation System Research Center, Wuhan University of Technology, Wuhan Hubei 430063, China;
    2. National Engineering Research Center for Water Transport Safety, Wuhan Hubei 430063, China;
    3. School of Logistics Engineering, Wuhan University of Technology, Wuhan Hubei 430063, China
  • Received:2019-10-20 Revised:2019-12-20 Online:2020-01-28 Published:2021-01-22

摘要: 为揭示船舶避碰(SCA)决策研究中船舶操纵人员在会遇场景表现出的避碰决策动机和偏好,更准确地反映多船避碰(MSCA)场景下各个船舶运动变化的趋势,将博弈模型引入现有的MSCA分析,提出将MSCA问题转化为相关船舶间完全信息的非零和动态博弈问题的方法。首先,利用最短会遇距离(DCPA)和最短会遇时间(TCPA)等参数度量SCA的危险程度,获得避碰优先级;其次,参考国际海上避碰规则(COLREGS)选择船舶操纵性和经济偏好作为避碰决策特征,为每艘船建立扩展博弈树;最后,采用逆向归纳法求解子博弈纳什均衡。结果表明:所提出的基于博弈论的MSCA方法与传统方法相比,能使各个船舶作出更有利于化解碰撞局面的决策。

关键词: 多船避碰(MSCA), 扩展式博弈, 最短会遇距离(DCPA), 最短会遇时间(TCPA), 收益矩阵

Abstract: This paper is aimed at revealing the motives and preferences of ship operators in collision avoidance decision-making for ship collision avoidance (SCA) research, and more accurately reflecting the changing trend of each ship's motion in MSCA scenarios. The game model was introduced into the existing MSCA analysis, and a method was proposed to transform the MSCA problem into a non-zero-sum dynamic game problem with complete information between related ships. Firstly, the collision risk of the encounter situations was estimated by using parameters such as DCPA and TCPA, and the SCA priority matrix of all involved ships was established. Secondly, by considering the principles of the Conventional International Regulations for Preventing Collisions at Sea (COLREGS) and choosing ship maneuverability and economy preference as SCA decision making features, the extensive form game tree was established for each ship. Finally, the backward induction was used to obtain the subgame Nash equilibrium. The simulation results show that the presented game theory-based MSCA model provides more beneficial strategy than the conventional MSCA models in terms of solving the collision-prone situation.

Key words: multi-ship collision avoidance(MSCA), extensive form game theory, distance to closest point of approach (DCPA), time to closest point of approach (TCPA), pay-off matrix

中图分类号: