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

BMAC网络及其在函数学习中的应用
引用本文:刘昭度,马岳峰,张景波,齐志权,宋明.BMAC网络及其在函数学习中的应用[J].北京理工大学学报,2004,24(10):858-861.
作者姓名:刘昭度  马岳峰  张景波  齐志权  宋明
作者单位:北京理工大学,机械与车辆工程学院,北京,100081;北京理工大学,机械与车辆工程学院,北京,100081;北京理工大学,机械与车辆工程学院,北京,100081;北京理工大学,机械与车辆工程学院,北京,100081;北京理工大学,机械与车辆工程学院,北京,100081
摘    要:论述了将n阶B样条函数运用到CMAC网络中形成BMAC网络的过程,并讨论了BMAC网络感受域函数和网络输出特性.BMAC网络与CMAC网络相比,克服了输入状态空间和输出状态空间离散化的缺点,具有连续的输入状态空间和输出状态空间.在函数学习中通过与CMAC网络的对比,体现出了BMAC网络具有逼近速度快,精度高的特性,同时也得出了学习参数及网络权值初始化对学习速度及逼近精度的影响规律.

关 键 词:B样条小脑模型神经网络  B样条函数  感受域函数
文章编号:1001-0645(2004)10-0858-04
收稿时间:2003/11/3 0:00:00
修稿时间:2003年11月3日

BMAC Neural Network and Its Application in Function Learning
LIU Zhao-du,MA Yue-feng,ZHANG Jing-bo,QI Zhi-quan and SONG Ming.BMAC Neural Network and Its Application in Function Learning[J].Journal of Beijing Institute of Technology(Natural Science Edition),2004,24(10):858-861.
Authors:LIU Zhao-du  MA Yue-feng  ZHANG Jing-bo  QI Zhi-quan and SONG Ming
Institution:School of Mechanical and Vehicular Engineering, Beijing Institute of Technology, Beijing100081, China;School of Mechanical and Vehicular Engineering, Beijing Institute of Technology, Beijing100081, China;School of Mechanical and Vehicular Engineering, Beijing Institute of Technology, Beijing100081, China;School of Mechanical and Vehicular Engineering, Beijing Institute of Technology, Beijing100081, China;School of Mechanical and Vehicular Engineering, Beijing Institute of Technology, Beijing100081, China
Abstract:BMAC neural network has been established by the introduction of B-splines into CMAC neural network. Ways of establishing BMAC and the specialties of BMAC output and its receptive field functions are described in details. BMAC has continuous input and output space compared with CMAC whose input and output space are dispersed. In function learning, BMAC approaches faster and is more accurate than CMAC. At the same time, rules of influence of the studying parameters and weights initialization in function learning are obtained.
Keywords:BMAC  B-splines function  receptive field functions
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《北京理工大学学报》浏览原始摘要信息
点击此处可从《北京理工大学学报》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