摘要
在前向模糊神经网络的归一化层和输出层之间加入递归层,形成的一种新型动态模糊神经网络(DFNN)具有动态映射能力,从而对动态系统有更好的响应.文章还推导了基于BP的反传学习算法.运用DFNN对船舶动力定位控制进行的仿真实验结果证明了该方法的有效性.
A new dynamical fuzzy neural network （DFNN） is constructed by adding a recurrent layer between the normalized layer and the outpout layer of the forward fuzzy neural network. The new DFNN has ability of dynamic mapping and better response to dynamic system. Meanwhile, the back propagation learning algorithm is given based on BP. A simulating test of dynamic positioning control of ships is done by using DFNN. The results show that the method is effective.
出处
《船舶工程》
CSCD
北大核心
2006年第2期43-46,共4页
Ship Engineering
关键词
船舶
动力定位
动态模糊神经网络
动态系统
ship
dynamic positioning system
dynamic fuzzy neural network（DFNN）
dynamic system
