| 基于RBF神经网络的互联网时延预测仿真 |
| |
| 引用本文： | 姚君兰.基于RBF神经网络的互联网时延预测仿真[J].兵工自动化,2006,25(4):39-41. |
| |
| 作者姓名： | 姚君兰 |
| |
| 作者单位： | 湖北经济学院,教育技术部,湖北,武汉,430205 |
| |
| 摘 要： | RBF神经网络采用正交最小平方算法(OLS)决定隐层单元数目、基函数的中心和权值.该算法以每个输入样本为聚类中心,随着正交运算次数的增加,网络的输出误差平方将逐步减小到设定误差范围内,得到隐含层节点数和网络的权值.仿真表明RBF神经网络是有效的.
|
| 关 键 词： | RBF神经网络 时延预测 正交最小平方算法 |
| 文章编号： | 1006-1576（2006）04-0039-03 |
| 收稿时间： | 2005-10-08 |
| 修稿时间： | 2005-11-06 |
Simulation of Internet Time-Delay Prediction Based on RBF Neural Network |
| |
| YAO Jun-lan.Simulation of Internet Time-Delay Prediction Based on RBF Neural Network[J].Ordnance Industry Automation,2006,25(4):39-41. |
| |
| Authors: | YAO Jun-lan |
| |
| Affiliation: | Dept. of Teaching Technology, Hubei University of Economics, Wuhan 430205, China |
| |
| Abstract: | The orthogonal least-squares algorithm (OLS Algorithm) was adopted by RBF neural network to decide the element numbers of implication layer, the center of basic function and weight means. The every input sample was taken as the clustering center. As the increasing orthogonal calculations, the output error square of Internet was reduced to designed error category. The node numbers of implication layer and weight means of network were achieved. The effectiveness of RBF neural network was proved by simulation. |
| |
| Keywords: | RBF neural network Time-delay predication Orthogonal Least-Squares Algorithm (OLS Algorithm) |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
