基于神经网络的船舶碰撞危险度模型  

A Model Based on Neural Networks for the Risk of the Collision between Ships

作　　者：朱晓琳[1] 许汉珍[1] 

机构地区：[1]华中理工大学交通科学与工程学院

出　　处：《华中理工大学学报》2000年第10期52-54,共3页Journal of Huazhong University of Science and Technology

基　　金：湖北省自然科学基金资助项目! (96J0 1 2 )

摘　　要：采用模糊理论方法建立船舶碰撞危险度的数学模型 ,并对其中的关键因素安全域 (Domain)和动界(Arena)采用BP神经网络根据专家知识进行学习 .该模型综合考虑了环境、操纵性、会遇势态等多种因素的影响 ,能快速跟踪状态变化 。A fuzzy model is presented to express the risk of collision between ships. The key variables “Domain” and “Arena” in the model are obtained, using the neural networks learned from the experts. The effects of the various factors are considered. Following rapidly the change of the situation, the model can be applied to the consultation on line in a collision avoidance expert system.

关 键 词：船舶避碰 碰撞危险 神经网络 模糊决策 

分 类 号：U661.33[交通运输工程—船舶及航道工程] U675.9[交通运输工程—船舶与海洋工程]