| 神经网络用于船舶操纵动态特性学习的研究 |
| |
| 引用本文： | 黄梓瑜,张炎华.神经网络用于船舶操纵动态特性学习的研究[J].中国造船,1998(4):13-18. |
| |
| 作者姓名： | 黄梓瑜 张炎华 |
| |
| 作者单位： | 上海交通大学 |
| |
| 摘 要： | 本文针对BP网络存在的一些问题，提出一种单层神经网络结构，通过大量的仿真试验，对两种网络在学习、预报和适应参数变化等方面的特性进行了对比。结果表明单层神经网络用于学习船舶操纵动态特性，具有收敛速度极快、预测误差较小、能快速跟踪参数变化等优点，可用于船舶操纵特性的实时学习。
|
| 关 键 词： | 船舶操纵 神经网络 自适应控制 |
STUDY ON LEARNING OF SHIP'S OPERATIONAL CHARACTERISTICS WITH NEURAL NETWORKS |
| |
| HUANG Ziyu,ZHANG Yanhua.STUDY ON LEARNING OF SHIP''S OPERATIONAL CHARACTERISTICS WITH NEURAL NETWORKS[J].Shipbuilding of China,1998(4):13-18. |
| |
| Authors: | HUANG Ziyu ZHANG Yanhua |
| |
| Institution: | Shanghai Jiaotong University |
| |
| Abstract: | The neural networks are used to learn the operational characteristics of a ship. Thesimulation results show that the single layer metwork presented in this paper has many advantages over a BP network. Its fast learning speed and high predict precision make itprominent in real-time control area. |
| |
| Keywords: | Neural network Ship maneuver Adaptive control |
| 本文献已被 CNKI 维普 等数据库收录！ |
