基于能量比小波网络的sEMG虚拟识别系统设计  

Virtual Reorganization System of sEMG Based on Loose Wavelet Packet NNT and Energy Ratio

作　　者：李静[1] 赵丽[1] 耿丽清[1] 

机构地区：[1]天津职业技术师范大学天津市信息传感与智能控制重点实验室,天津300222

出　　处：《机床与液压》2012年第10期108-110,共3页Machine Tool & Hydraulics

基　　金：中国科学院支持天津滨海新区建设科技行动计划项目(2007AA04Z254);天津市科技发展支撑重点项目(08ZCKFSF03400);天津职业技术师范大学科研发展基金预研项目(KJY11-11)

摘　　要：研究了基于松散小波神经网络及能量比法的表面肌电信号分类原理和方法,提出利用小波包变换获得表面肌电信号各频段系数,并将化简处理后的能量比值作为特征向量输入后续的松散型小波神经网络用于分类表面肌电信号的方法,搭建了虚拟识别系统平台,进行了详细的实验研究。仿真结果表明:松散型小波神经网络及能量比法能够有效的从桡侧腕屈肌和总指伸肌采集的肌电信号中识别握拳、松拳、腕外展、腕内收4种运动模式,具有较高的识别率(90%以上)。sEMG signal classification method and principle based on energy ratio and loose wavelet packet neural network was re- searched. Wavelet packet transform was used to obtain the sEMG coefficient of each band, which was processed by energy ratio as the feature vectors to be input to loose wavelet packet neural network. A virtual recognition system was set up, and detailed experimental study was carried out. Simulation results show that the loose wavelet neural network and energy ratio can be used to effectively recog- nize four movement patterns of fist, loose fist, wrist abduction, wrist adduction, which were gotten from the radial wrist flexor and figure extensor muscles, with a high recognition rate （90%）.

关 键 词：能量比值 松散小波神经网络 表面肌电信号 虚拟识别 

分 类 号：TP319.9[自动化与计算机技术—计算机软件与理论]