[1]张秀玲,张志强.基于动态RBF神经网络的板形板厚综合控制仿真研究[J].智能系统学报,2007,2(2):65-68.
ZHANG Xiu-ling,ZHANG Zhi-qiang.Simulation research on strip flatness and thickness control? based on dynamic RBF neural networks[J].CAAI Transactions on Intelligent Systems,2007,2(2):65-68.
点击复制
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
2
期数:
2007年第2期
页码:
65-68
栏目:
学术论文—机器学习
出版日期:
2007-04-25
- Title:
-
Simulation research on strip flatness and thickness control? based on dynamic RBF neural networks
- 文章编号:
-
1673-4785(2007)02-0065-04
- 作者:
-
张秀玲,张志强
-
燕山大学电气工程学院,河北秦皇岛066004
- Author(s):
-
ZHANG Xiu-ling,ZHANG Zhi-qiang
-
College of Electrical Engineering, Yanshan University, Qinhuangdao 066004, Chin a
-
- 关键词:
-
BF网络; 动态设计; 逆矩阵; 板形板厚综合控制
- Keywords:
-
RBFNN; dynamic design; inverse matrix; integrated control of strip fla tness and thickness
- 分类号:
-
TP183
- 文献标志码:
-
A
- 摘要:
-
基于RBF神经网络的特点提出了一种动态调节隐含层隐节点个数的方法,由2部分组成:首先以网络输出数据的均方误差及其变化率为标准来调节隐含层节点的数目,然后调节优化隐含层节点的中心值,根据广义逆矩阵的方法求出输出层权值.所设计的神经网络具有最少的隐含层节点数,提高了学习训练速度,构造了板形板厚综合控制的数学模型,采用新的模型处理方法,用动态RBF神经网络进行控制仿真,取得了理想的结果.
- Abstract:
-
A method to dynamically adjust the number of hidden layer nodes is pro posed based on features of the RBFNN, which includes two parts: the first part i s to adjust the number of hidden layer nodes based on the mean square error and change rate of network output data, and the second part is to optimize the centr al v alue of the hidden layer and find the output layer’s weights based on the gener alized inverse matrix. The newly designed RBFNN has least nodes of hidden layers and higher training speed. A mathematical model for controlling strip flatness and thickness is proposed. Control simulation is executed with dynamic RBF neur al network based on new model, receiving an ideal result.
备注/Memo
收稿日期:2006-11-04.
基金项目:
国家自然科学基金资助项目(50374058);
燕山大学博士基金资助项目(B70)
作者简介:
张秀玲,女,1968年生,博士,教授,主要研究方向为神经网络智能控制.获国家科技进步二等奖一项,省部级一等奖、二等奖各一项,发表论文50余篇.
E-mail:zxlysu@Yahoo.com.cn.
张志强,男,1979年生,硕士,主要研究方向为神经网络优化设计.
更新日期/Last Update:
2009-05-06