基于博弈的干支交汇水域船舶汇入意图识别  

Ship Merge-in Intention Identification at River Trunk and Branch Intersection Waters Based on Game Theory

作　　者：刘奕[1,2,3] 孙玉飞[1] 袁智 魏天明 张贵平 LIU Yi;SUN Yufei;YUAN Zhi;WEI Tianming;ZHANG Guiping(School of Navigation,Wuhan University of Technology,Wuhan 430063,China;Hubei Inland Shipping Technology Key Laboratory,Wuhan 430063,China;National Engineering Research Center for Water Transportation Safety,Wuhan 430063,China;Jiangsu Maritime Safety Administration of the People’s Republic of China,Nanjing 210009,China)

机构地区：[1]武汉理工大学航运学院,武汉430063 [2]武汉理工大学内河航运技术湖北省重点实验室,武汉430063 [3]武汉理工大学国家水运安全工程技术研究中心,武汉430063 [4]中华人民共和国江苏海事局,南京210009

出　　处：《武汉理工大学学报（交通科学与工程版）》2024年第5期1011-1016,共6页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国家自然科学基金(51709219);武汉理工大学自主创新研究基金(2022IVA048)。

摘　　要：文中基于博弈理论和船舶碰撞危险度量方法,构造包含安全收益、时间收益和规则约束的汇入决策收益函数,采用船舶冲突排序运算方法量化确定汇入决策时机,构建干支交汇水域船舶汇入意图识别模型.为精准估计模型权重,以决策收益最大化为下层目标,意图识别误差最小化为上层目标,设计基于双层优化方法的模型参数估计框架.运用长江江苏段实测数据验证了所提模型的有效性.结果表明:本文所提模型的汇入意图识别准确率达到了94.64%.Based on the game theory and ship collision risk measurement method,the income function of inward decision-making was constructed,which included safety income,time income and rule constraints.The decision-making time of ship import was quantitatively determined by the method of ship conflict ranking operation,and then the identification model of ship import intention in the water area where the trunk and branch meet was constructed.In order to accurately estimate the weight of the model,with the maximization of decision-making income as the lower goal and the minimization of intention recognition error as the upper goal,a model parameter estimation framework based on double-level optimization method was designed.The validity of the proposed model is verified by the measured data of Jiangsu section of the Yangtze River.The results show that the accuracy of the model proposed in this paper is 94.64%.

关 键 词：水路交通安全 干支交汇水域 汇入意图识别 博弈论 

分 类 号：U698[交通运输工程—港口、海岸及近海工程]