首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于混合学习算法的模糊小波神经网络控制
引用本文:彭金柱,王耀南,孙炜.基于混合学习算法的模糊小波神经网络控制[J].湖南大学学报(自然科学版),2006,33(2):51-54.
作者姓名:彭金柱  王耀南  孙炜
作者单位:湖南大学,电气与信息工程学院,湖南,长沙,410082
基金项目:国家自然科学基金资助项目(60375001)
摘    要:采用小波函数作为模糊隶属函数,将模糊控制与神经网络相结合,利用神经网络实现模糊推理.针对BP算法易陷入局部极值点的缺点和简单遗传算法局部搜索能力差的不足,提出了一种混合学习算法,即首先利用遗传算法全局搜索的特点来离线优化神经网络的参数,再利用BP算法较强的局部搜索能力对网络参数进行在线调整.仿真结果表明,该网络能对不同的对象实施有效控制,且具有快速、适应性强等特点.

关 键 词:模糊神经网络  小波  遗传算法  BP算法
文章编号:1000-2472(2006)02-0051-04
收稿时间:04 28 2005 12:00AM
修稿时间:2005-04-28

Fuzzy Wavelet Neural Networks Control Based on Hybrid Learning Algorithm
PENG Jin-zhu,WANG Yao-nan,SUN Wei.Fuzzy Wavelet Neural Networks Control Based on Hybrid Learning Algorithm[J].Journal of Hunan University(Naturnal Science),2006,33(2):51-54.
Authors:PENG Jin-zhu  WANG Yao-nan  SUN Wei
Institution:College of Electrical and Information Engineering, Hunan Univ, Changsha, Hunan 410082, China
Abstract:Using wavelet basis function as membership function,fuzzy control and neural network were combined,and the fuzzy inference was realized by neural network.To counteract the defects of BP algorithm and the chances of simple genetic algorithm premature convergence,a hybrid learning algorithm was proposed.First,the genetic algorithm was used to optimize the fuzzy neural network's parameters off-line.Then,because of the strong capability of local search,the BP algorithm was used to adjust the parameters on-line.The simulation results showed that the network could control different objects effectively,and had the characteristics of speediness and adaptability.
Keywords:fuzzy neural network  wavelet  genetic algorithm  BP algorithm  
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《湖南大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《湖南大学学报(自然科学版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号