内河渡船危险预警及避碰决策

Early Warning of Collision and A voidance Assitance for River Ferry

为保证渡船安全航行,基于内河渡船驾驶员的经验,运用相对运动几何分析法建立内河渡船碰撞危险预警等级量化几何模型;结合专家避让操纵知识库,运用船舶拟人智能避碰决策(Personifying Intelligent Decision-Making for Vessel Collison Avoidance,PIDVCA)的基础模型和算法设计渡船预警和避碰决策算法流程。以镇扬汽渡船为例,在船舶智能操控仿真平台(Ship Intelligent Handle and Control,SIHC)上模拟镇扬汽渡船水域的交通流,进行仿真测试。仿真结果表明:该算法能实现镇扬汽渡船不同等级的碰撞危险预警,提供的避让决策符合规则和驾驶员的习惯。该方法可为内河渡船避碰避险决策系统开发提供理论基础,对确保渡船航行安全具有现实意义。

A quantitative collision risk level model is built based on the relative motion geometry analysis and the experience/knowledge of river ferry pilots. The algorithm for early waning and maneuvering decision-making is developed based on the basic PIDVCA(Personifying Intelligent Decision-Making for Vessel Collision Avoidance) algorithm and further enhanced by taking the advantage of a ship handling expert knowledge base. The Zhen Yang car ferry is simulated on the SIHC(Ship Intelligent Handle and Control) platform. The simulation results show that the algorithm can give warning of right level and suggest collision avoidance plan in accordance with the rules and ship handling practice. The method provides a theoretical basis for the development of the decision-making system for avoiding collisions of inland river ferries, which is a practical reference in navigation safety of ferries.