| 基于回归神经网络自适应快速BP算法 |
| |
| 引用本文： | 张丽红,王艳.基于回归神经网络自适应快速BP算法[J].计算机测量与控制,2004,12(5):480-482. |
| |
| 作者姓名： | 张丽红 王艳 |
| |
| 作者单位： | 山西大学,物理电子工程学院,山西,太原,030006 |
| |
| 摘 要： | 动态递归网络Elman网络结构简单，运算量少，适合于实时系统辨识。以Elman网络结构推导了在线学习算法。针对于传统BP算法会产生局部收敛和收敛速度慢等缺点，提出了一种改进的自适应BP算法，运用到回归神经网络，提高了在线学习的速度与收敛速度，仿真实验表明了此算法的有效性和快速性。
|
| 关 键 词： | 回归神经网络 自适应 BP算法 仿真 收敛速度 Elman网络结构 实时控制 |
| 文章编号： | 1671-4598(2004)05-0480-03 |
| 修稿时间： | 2003年10月10 |
Adaptive Fast BP Algorithm Based on Recurrent Neural Network |
| |
| Zhang Lihong,Wang Yan.Adaptive Fast BP Algorithm Based on Recurrent Neural Network[J].Computer Measurement & Control,2004,12(5):480-482. |
| |
| Authors: | Zhang Lihong Wang Yan |
| |
| Abstract: | The construction of dynamic recursive network-Elman network is simple and easy to calculate, and it is suitable for real time system recognition. An online learning algorithm is deduced based on Elman network construction in this paper. To avoid the shortcomings of traditional BP algorithm such as part convergence and slow convergence rate, an adaptive fast BP Algorithm is provided and it is used in recurrent neural network. It has improved the studying rate on line and the convergence rate. Stimulation result has proved its validness and fastness. |
| |
| Keywords: | recurrent neural network BP algorithm stimulation convergence rate |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
