首页 | 本学科首页   官方微博 | 高级检索  
     

基于功率谱分析和RBF网络的表面EMG多模式分类
引用本文:张清菊,罗志增,叶明.基于功率谱分析和RBF网络的表面EMG多模式分类[J].机电工程,2005,22(11):35-38.
作者姓名:张清菊  罗志增  叶明
作者单位:杭州电子科技大学,机器人研究所,浙江,杭州,310018
基金项目:国家自然科学基金资助项目(60474054);浙江省自然科学基金资助项目(RC02070).
摘    要:提出了一种以功率谱K值法和RBF网络相结合的表面肌电信号处理方法.首先,将采集到的肌电信号进行预处理,计算互功率谱比值作为其特征值;其次,将其特征值作为训练样本输入RBF神经网络进行网络训练,并对手臂的各种动作进行多运动模式分类.实验表明,这种方法不仅简化了计算工作量,而且取得了比较理想的识别效果.

关 键 词:表面肌电信号  功率谱比值  RBF神经网络
文章编号:1001-4551(2005)011-0035-04
收稿时间:2005-07-05
修稿时间:2005-08-22

Based on Power Spectrum and RBF Neural Network to Classify Surface Electromyography
ZHANG Qing-ju,LUO Zhi-zeng,YE Ming.Based on Power Spectrum and RBF Neural Network to Classify Surface Electromyography[J].Mechanical & Electrical Engineering Magazine,2005,22(11):35-38.
Authors:ZHANG Qing-ju  LUO Zhi-zeng  YE Ming
Affiliation:Robotics Research Institute, Hangzhou Dianzi University, Hangzhou 310018, China
Abstract:In this paper, a method to process surface electromyography is presented, which bases on Power Spectrum and RBF neural network. First, we calculate Power Spectrum coefficient with the original signal that is pretreated as its eigenvector. Second, using the Power Spectrum coefficient to train the RBF neural network and classify the muscle movement of forearm with power spectrum coefficient. The experiment indicates this measure can reduce workload and get the relatively good results.
Keywords:surface electromyography signal  power spectrum  RBF neural network
本文献已被 CNKI 维普 万方数据 等数据库收录!
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