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

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.