人工神经网络和H_∞优化控制用于航迹舵的实现研究  

The Implementation Study of Artificial Neural Network(ANN) and the H Infinite(HINF) Controller Applying to a Track Autopilot

作　　者：贾欣乐[1] 朱利民[1] 蒋丹东[1] 张显库[1] 杨承恩[1] 崔海深[2] 李永启 

机构地区：[1]大连海事大学轮机工程研究所 [2]大连远洋运输公司

出　　处：《中国航海》1998年第2期60-68,共9页Navigation of China

摘　　要：本文概述一种新型航迹自动舵的系统结构，重点讨论航迹和航向控制器的设计方法，报告航迹舵的海上试验概况，给出神经元网络、Ｈ∞混合灵敏度及多重增益规划ＰＩＤ三种航向控制算法的实航结果比较，这些经验对航迹舵的进一步研究和产品化有重要意义。The system structure of a new autopilot for ships is outlined briefly. The stress of the paper is laid on the design principles of the track and course keeping controllers. The sea trial results of the track autopilot are reported, and the comparisons are made among the performances of the neural network controller,the H ∞ mixed sensitivity controller and the multiple gain scheduling proportional integral differential (PID) controller. These experiences will be beneficial to the future study and the production of the autopilot.

关 键 词：航迹自动舵 H∞优化控制 神经网络 舵 

分 类 号：U664.36[交通运输工程—船舶及航道工程]