摘要
针对在船舶操纵控制系统中 ,船舶模型参数会随着外界条件 (如航速、吃水、风等 )的改变而变化的情况 ,提出了如何用神经网络模型使船舶在这些不断变化的外界条件下保持良好的状态 ,描述了前向网络原理与反向传播学习算法 (BP)的步骤 .
A ship steering control system is of non-linear because the parameters of the ship model change with operating conditions (such as forward speed of the ship, depth of water etc.). In this paper, we examine how to obtain satisfactory performance with a neural network under these varying conditions. Then describe the principle adopted in training a feedforward type of network and the back-propagation algorithm.
出处
《武汉理工大学学报(交通科学与工程版)》
北大核心
2003年第3期291-293,共3页
Journal of Wuhan University of Technology(Transportation Science & Engineering)
基金
国家自然科学基金项目资助 (6 0 10 30 2 1)
关键词
人工神经网络
控制系统
前向网络
反向传播学习规则
artificial neural networks
control system
feedforward network
back-propagation learning rule