检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:苗雨阳 徐言民[1,3,4] 关宏旭 律建辉[1,2,3] 刘佳仑 MIAO Yu-yang;XU Yan-min;GUAN Hong-xu;LV Jian-hui;LIU Jia-lun(School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan 430063,China;Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai),Zhuhai 519000,China;Hubei Provincial Key Laboratory of Inland Shipping Technology,Wuhan University of Technology,Wuhan 430063,China;School of Entrepreneurship,Wuhan University of Technology,Wuhan 430070,China;Intelligent Transportation Systems Research Center,Wuhan University of Technology,Wuhan 430063,China;National Engineering Research Center Water Transport Safety,Wuhan 430063,China)
机构地区:[1]武汉理工大学交通与物流工程学院,湖北武汉430063 [2]南方海洋科学与工程广东省实验室(珠海),广东珠海519082 [3]武汉理工大学内河航运技术湖北省重点实验室,湖北武汉430063 [4]武汉理工大学创业学院,湖北武汉430070 [5]武汉理工大学智能交通系统研究中心,湖北武汉430063 [6]武汉理工大学国家水运安全工程技术研究中心,湖北武汉430070
出 处:《舰船科学技术》2024年第1期164-169,共6页Ship Science and Technology
摘 要:结合虚拟测试验证要点,梳理测试场景构成要素,实现场景要素跨域融合,构建测试船、目标船和环境间的博弈与对抗行为特征,针对场景表示的3种抽样层次,提出面向虚拟测试的场景生成方法。基于场景优化生成方法,在场景组合生成方法的基础上离线生成测试场景,主要针对参数空间的边缘场景,让智能算法参与场景生成过程,其场景覆盖度更低,通过测试高风险的边界场景,以测试特定系统性能效果。以单船避碰为例,在虚拟仿真测试平台进行仿真验证,结果表明优化生成后的场景文件能够有效降低场景生成数量,显著提升生成效率,且在场景交互博弈、覆盖率和可重复测试等方面具有良好性能。Combined with the key points of intelligent ship testing and verification,analyze the components of high-risk edge test scenarios,realize cross-domain fusion of edge scenario elements in the virtual simulation test system,simulate the game and confrontation behavior characteristics between the ship and other ships and the environment in the real world,and express the scene for the scene.The three sampling levels are proposed,and a virtual test-oriented edge scene construction system is proposed.The scene keywords are extracted based on the semantic model,mapped to the parameter space to generate dynamically changing logical scenes,the scene model is described as a scene dynamics model,the scene controller is constructed by using a general function approximator,and realize the optimal solution of edge scene controller with the help of reinforcement learning.Finally,the parameter space is sampled by keywords to generate specific edge scene files.Taking the simultigle-vessel encounter scene of a test site as an example,the simulation verification is carried out on the virtual simulation test platform.The results show that the generated edge scene file can meet the requirements of the automatic driving test of the ship,significantly improve the generation efficiency.And it has good performance in scene interactive game,coverage and repeatable testing,etc.
分 类 号:U692.51[交通运输工程—港口、海岸及近海工程]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.222