摘要
针对复杂水域船舶碰撞风险较高的问题,选取航速、航向、距离、能见度、风、浪、流等碰撞影响因素,设计碰撞危险度的计算方法,提出环境危险度的概念及计算方法;根据二者计算结果,构建神经网络样本,利用神经网络模型进行训练;基于Matlab设计复杂水域智能避碰专家系统。在对遇、追越、交叉相遇3种会遇局势下进行仿真验证,实验表明,该系统可以直接求解出碰撞危险度和环境危险度,将目标船和环境因素造成的碰撞危险区别对待,根据实际情况给出航行提示和建议。
Aiming at the problem of high collision risk of ships in complex waters,the ship intelligent collision avoidance expert system was constructed. The collision factors such as speed,heading,distance,visibility,wind,wave and flow were selected to design a calculation method of collision risk. The concept and calculation method of environmental collision risk were proposed. According to results of the collision risk and the environmental risk,the samples of neural network were constructed and trained. The intelligent collision avoidance expert system for complex waters was designed by Matlab. The simulations were carried out in three encounter situations: confrontation,chasing and crossover. The experiment showed that this system can directly get the collision risk and environmental collision risk,and treat the collision risk differently caused by the target ship and environmental factors,giving navigation tips and suggestions according to the actual situation.
作者
汤国瑞
谢新连
潘伟
TANG Guo-rui;XIE Xin-lian;PAN Wei(Integrated Transport Institute, Dalian Maritime University, Dalian Liaoning 116026, China)
出处
《船海工程》
北大核心
2019年第3期160-165,共6页
Ship & Ocean Engineering
基金
国家重点研发计划(2017YFC0805309)
中央高校基本科研业务费专项资金(3132016358)
关键词
复杂水域
智能避碰
专家系统
模糊神经网络
碰撞危险度
环境危险度
complex waters
intelligent collision avoidance
expert system
fuzzy neural network
collision risk
collision risk of environment
作者简介
第一作者:汤国瑞(1994-),男,硕士生研究方向:交通运输规划与管理.
