|
|||||
|
|
| 一种改进的基于扩展的RBF模糊神经网络 | |
| 引用本文： | 胡蓉,徐蔚鸿,夏烨.一种改进的基于扩展的RBF模糊神经网络[J].宁夏大学学报(自然科学版),2010,31(3):239-242. |
| 作者姓名： | 胡蓉 徐蔚鸿 夏烨 |
| 作者单位： | 1. 南京理工大学计算机科学与技术学院,江苏南京,210094;长沙航空职业技术学院计算机与信息工程系,湖南长沙,410003 2. 南京理工大学计算机科学与技术学院,江苏南京,210094;长沙理工大学计算机与通信工程学院,湖南长沙,410076 3. 长沙金荣集团,湖南长沙,410013 |
| 基金项目： | 教育部重点科研基金资助项目，湖南省教育厅重点项目 |
| 摘 要： | 针对Wu Shiqian等提出的动态模糊神经网络（D-FNN）,在参数估计时,对中心的调整直接采用输入数据会导致网络收敛速度慢、乏化能力不好的情况,提出一种改进的基于RBF的动态模糊神经网络（RBDFNN）.该网络采用连续学习的方法和分级学习的思想,在进行前提参数的估计时,高斯隶属函数的中心动态调整充分考虑了宽度调整对中心的影响,算法能自动地确定模糊规则.仿真实验表明,算法在收敛速度和乏化能力方面优于原来的算法. |
| 关 键 词： | 模糊神经网络 参数调整 泛化能力 |
An Improved Fuzzy Neural Network Based on RBF |
|
| Hu Rong,Xu Weihong,Xia Ye.An Improved Fuzzy Neural Network Based on RBF[J].Journal of Ningxia University(Natural Science Edition),2010,31(3):239-242. | |
| Authors: | Hu Rong Xu Weihong Xia Ye |
| Affiliation: | 1.School of Computer Science and Engineering,Nanjing University of Science and Technology,Nanjing 210094,China;2.Department of Computer Information and Engineering,Changsha Aeronautical Vocational and Technical College,Changsha 410003,China;3.College of Computer and Communications Engineering,Changsha University of Science and Technology,Changsha 410076,China;4.Changsha Kinglory Enterprises Group Co.,Ltd.,Changsha 410013,China) |
| Abstract: | The D-FNN presented by Wu Shiqian,which adjust centers directly using input data,leads to slow convergence and lack of capacity of generalization.To solve this problem,a new algorithm for creating RBDFNN is proposed.In this neural network,hierarchical on-line self-organizing learning is used.During the parameters estimation phase,we adjust the Gaussian centers according to the adjustment of widths.Neurons can be recruited or deleted dynamically.Simulation studies demonstrate the presented algorithm is superior in terms of learning efficiency and capacity of generalization. |
| Keywords: | fuzzy neural network adjustments of parameters capacity of generalizations |
| 本文献已被 维普 万方数据 等数据库收录！ | |
