基于自适应神经网络智能舵的自动避碰研究  

Research on Automatic Collision Avoidance Based on Adaptive Neural Network Intelligent Rudder

作　　者：宁君[1] 陈汉民 鲁峰[1] 李春泽 NING Jun;CHEN Han-min;LU Feng(Navigation College of Dalian Maritime University,Dalian 116026,China)

机构地区：[1]大连海事大学航海学院,辽宁大连116026

出　　处：《浙江海洋大学学报（自然科学版）》2021年第1期71-79,共9页Journal of Zhejiang Ocean University：Natural Science

基　　金：国家自然科学基金青年科学基金项目(61803064);辽宁省自然科学基金(20170540098);中央高校基本科研业务费项目(3132021135)。

摘　　要：通过优化船舶避碰过程中涉及的数学模型,设计了一种基于自适应神经网络智能舵的自动避碰算法,用于船舶避碰决策过程。该算法可以判断会遇情况,计算碰撞危险度,并自动确定转向时间和转向幅度。此外,在船舶自动避碰过程中,还考虑了COLREGs,航行经验和自动避碰方法。基于动态面技术和神经网络方法,解决了船舶运动控制系统中由后推法引起的“计算量膨胀”和“维度灾难”问题。通过对“YUKUN”号和“YULONG”号远洋船的Matlab仿真实验,验证了所提出的自动避碰决策系统的有效性。In this paper,by studying and optimizing the mathematical model involved in the collision avoidance process of ships,an automatic collision avoidance algorithm based on an adaptive neural network intelligent rudder is designed for the ship collision avoidance decision process.The algorithm can determine the meeting situation,calculate the risk of collision,and automatically determine the time and range of action.In addition,during the ship's automatic collision avoidance process,COLLEGs,navigation experience,and automatic collision avoidance methods were also considered.Based on the dynamic surface control technology and neural network method,the algorithm can also solve the"Calculation volume expansion"and"Dimensional disaster"problems caused by the backstepping method in ship motion control systems.The validity of the proposed automatic collision avoidance decision system is verified by Matlab simulation experiments on the ocean-going ships"Yukun"and"YULONG."

关 键 词：智能舵 自动避碰 神经网络 动态面技术 

分 类 号：U675.96[交通运输工程—船舶及航道工程]