oalib

Publish in OALib Journal

ISSN: 2333-9721

APC: Only $99

Submit

Any time

2019 ( 46 )

2018 ( 4910 )

2017 ( 4728 )

2016 ( 4662 )

Custom range...

Search Results: 1 - 10 of 262517 matches for " <br>韩红桂 "
All listed articles are free for downloading (OA Articles)
Page 1 /262517
Display every page Item
Optimal Structure Design for RBFNN Structure
RBF神经网络的结构动态优化设计

QIAO Jun-Fei,HAN Hong-Gui,<br>乔俊飞,韩红
自动化学报 , 2010,
Abstract: Due to the fact that the conventional radial basis function (RBF) neural network cannot change the structure on-line, a new dynamic structure RBF (D-RBF) neural network is designed in this paper. D-RBF is based on the sensitivity analysis (SA) method to analyze the output values of the hidden nodes for the network output, then the hidden nodes in the RBF neural network can be inserted or pruned. The final structure of D-RBF is not too large or small for the objectives, and the convergence of the dynamic process is investigated in this paper. The grad-descend method for the parameter adjusting ensures the convergence of D-RBF neural network. The structure of the RBF neural network is self-organizing, and the parameters are self-adaptive. In the end, D-RBF is used for the non-linear functions approximation and the non-linear systems modelling. The results show that this proposed D-RBF obtains favorable self-adaptive and approximating ability. Especially, comparisons with the minimal resource allocation networks (MRAN) and the generalized growing and pruning RBF (GGAP-RBF) reveal that the proposed algorithm is more effective in generalization and finally neural network structure.
Dynamic optimization structure design for neural networks: review and perspective
神经网络结构动态优化设计的分析与展望

QIAO Jun-fei,HAN Hong-gui,<br>乔俊飞,韩红
控制理论与应用 , 2010,
Abstract: The influence of the structure design to the capabilities of neural networks is discussed in this paper. The development history of the dynamic structure neural networks, especially the growing and the pruning neural networks are introduced. The substantial results on the computing capabilities, learning theories, stability of neural networks are then analyzed. Finally, the research on the dynamic optimization structure design is summarized and several views are put forward.
On Structure Design for RBF Neural Network Based on Information Strength
基于信息强度的RBF神经网络结构设计研究

HAN Hong-Gui,QIAO Jun-Fei,BO Ying-Chun,<br>韩红,乔俊飞,薄迎春
自动化学报 , 2012,
Abstract: Based on the systemic investigation on the feedforword neural network, for the problem of the structure design of the RBF neural network, a new flexible structure design method is used for RBF neural network in this paper. By computing the output-information (OI) of the hidden neurons and the multi-information (MI) of the hidden nodes and output nodes, the hidden nodes in the RBF neural network can be inserted or pruned, thus the topology of the network can be modulated. This method can effectively solve the structure design of the RBF neural network. The grad-descent method for the parameter adjusting ensures the exactitude of the flexible RBF neural network (F-RBF). The structure of the RBF neural network is self-organizing, and the parameters are self-adaptive. In the end, the proposed F-RBF is used for approximating the classical non-linear functions and modelling key parameters of the wastewater treatment process. The results show that the F-RBF obtains a favorable dynamic character response and the approximating ability. Especially, comparied with the minimal resource allocation networks (MRAN), the generalized growing and pruning RBF (GGAP-RBF) and the self-organizing RBF (SORBF), the proposed algorithm is more effective in terms of training time, generalization, and neural network structure.
RBF神经网络的结构动态优化设计
乔俊飞, 韩红
自动化学报 , 2010, DOI: 10.3724/SP.J.1004.2010.00865
Abstract: ?针对径向基函数(Radialbasisfunction,RBF)神经网络的结构设计问题,提出一种结构动态优化设计方法.利用敏感度法(Sensitivityanalysis,SA)分析隐含层神经元的输出加权值对神经网络输出的影响,以此判断增加或删除RBF神经网络隐含层中的神经元,解决了RBF神经网络结构过大或过小的问题,并给出了神经网络结构动态变化过程中收敛性证明;利用梯度下降的参数修正算法保证了最终RBF网络的精度,实现了神经网络的结构和参数自校正.通过对非线性函数的逼近与污水处理过程中关键参数的建模结果,证明了该动态RBF具有良好的自适应能力和逼近能力,尤其是在泛化能力、最终网络结构等方面较之最小资源神经网络(Minimalresourceallocationnetworks,MRAN)与增长和修剪RBF神经网络(Generalizedgrowingandpruningradialbasisfunction,GGAP-RBF)有较大提高.
神经网络结构动态优化设计的分析与展望
乔俊飞,韩红
控制理论与应用 , 2010, DOI: 10.7641/j.issn.1000-8152.2010.3.CCTA080987
Abstract: 阐述了神经网络结构设计对神经网络性能的影响.介绍了动态结构神经网络,尤其是增长型和修剪型神经网络研究的发展过程,分析了动态设计方法研究在计算能力、学习理论和网络的稳定性等方面取得的成果.最后对神经网络动态设计的研究进行总结,给出了神经网络结构动态设计研究的发展趋势.
Research and realization of dynamic neural network navigation algorithm for mobile robot
机器人动态神经网络导航算法的研究和实现

QIAO Jun-fei,FAN Rui-yuan,HAN Hong-gui,RUAN Xiao-gang,<br>乔俊飞,樊瑞元,韩红,阮晓钢
控制理论与应用 , 2010,
Abstract: For the navigation of Pioneer3-DX mobile robot in unknown environment, we propose a self-navigation strategy with learning reinforcement, and develop the navigation algorithm based on the dynamical neural network. The dynamically self-organizing neural network can automatically adjust its structure according to the complexity of the working environments of the mobile robot to realize the mapping between environmental states and robot actions, effectively avoiding the dimension explosion in learning reinforcement. Simulations and real robot navigation experiments are carried out; results show that the proposed method is effective in applications. It gives a better navigation performance than that of the artificial potential-field method.
基于贡献率的离散Hopfield结构优化
乔俊飞,李荣,韩红
控制与决策 , 2015, DOI: 10.13195/j.kzyjc.2014.1320
Abstract: 针对离散Hopfield神经网络(DHNN)结构复杂的问题,提出一种基于贡献率的结构优化算法.该算法利用奇异值分解方法对连接权值进行设计,进而利用贡献率的方法对DHNN进行结构优化.优化后的网络降低了DHNN结构的复杂程度,使网络具有类似生物神经网络的稀疏结构,实现了DHNN网络结构的优化.最后,通过水质评价和数字识别对该算法进行验证,表明了所提出算法的有效性和可行性,同时,还验证了其对于大规模DHNN的有效性和适用性.
基于ART的RBF网络结构设计
蒙西,乔俊飞,韩红
控制与决策 , 2014, DOI: 10.13195/j.kzyjc.2013.0945
Abstract: 针对径向基函数(RBF)网络隐层结构难以确定的问题,基于自适应共振理论(ART)网络良好的在线分类特性,提出一种RBF网络结构设计算法.该算法将ART网络的聚类特性用于RBF网络结构设计中,通过对输入向量与已存模式的相似度比较将输入向量进行分类,确定隐含层节点个数和初始参数,使网络具有精简的结构.对典型非线性函数逼近的仿真结果表明,所提出的结构具有快速的学习能力和良好的逼近能力.
一种改进型离散Hopfield学习算法
李荣,乔俊飞,韩红
控制与决策 , 2014, DOI: 10.13195/j.kzyjc.2012.1523
Abstract: 针对离散Hopfield神经网络(DHNN)的权值设计问题,提出一种改进型学习算法,并在DHNN动力学分析的基础上设计该学习算法.利用矩阵分解的方法(MD)得到正交矩阵,并采用得到的正交矩阵直接计算DHNN的权值矩阵.通过该学习算法得到的权值矩阵,可以很好地存储训练样本的信息,使测试样本收敛到稳定点.该学习算法不需要进行分块计算,减少了计算步骤和计算量,降低了网络的迭代次数,从而提高了网络运行速度.最后,将该学习算法应用于水质评价,验证了其有效性和可行性.
基于信息强度的RBF神经网络结构设计研究
韩红, 乔俊飞, 薄迎春
自动化学报 , 2012, DOI: 10.3724/SP.J.1004.2012.01083
Abstract: ?在系统研究前馈神经网络的基础上,针对径向基函数(Radialbasisfunction,RBF)网络的结构设计问题,提出一种弹性RBF神经网络结构优化设计方法.利用隐含层神经元的输出信息(Output-information,OI)以及隐含层神经元与输出层神经元间的交互信息(Multi-information,MI)分析网络的连接强度,以此判断增加或删除RBF神经网络隐含层神经元,同时调整神经网络的拓扑结构,有效地解决了RBF神经网络结构设计问题;利用梯度下降的参数修正算法保证了最终RBF网络的精度,实现了神经网络的结构和参数自校正.通过对典型非线性函数的逼近与污水处理过程关键水质参数建模,结果证明了该弹性RBF具有良好的动态特征响应能力和逼近能力,尤其是在训练速度、泛化能力、最终网络结构等方面较之最小资源神经网络(Minimalresourceallocationnetworks,MRAN)、增长修剪RBF神经网络(GeneralizedgrowingandpruningRBF,GGAP-RBF)和自组织RBF神经网络(Self-organizingRBF,SORBF)有较大的提高.
Page 1 /262517
Display every page Item


Home
Copyright © 2008-2017 Open Access Library. All rights reserved.