基于辅助变量盲辨识方法的肌电信号识别分类  

Identification Classification of Electromyography Signals Based on Blind Identification of Auxiliary Variable

作　　者：尚小晶[1] 

机构地区：[1]吉林建筑大学城建学院,长春130111

出　　处：《长春大学学报》2014年第4期435-438,共4页Journal of Changchun University

基　　金：吉林省教育厅"十二五"科学技术研究项目(2014596)

摘　　要：表面肌电信号(sEMG)是神经肌肉活动发放的生物电信号。本文从信号产生的根源出发,建立FIR模型,该模型的输入为信号源是不可测的,为了克服该不足本文提出采用盲辨识的方法辨识模型系数,以该系数作为表面肌电信号的特征,对6种手势动作识别分类,实验表明,该方法运算量小,分类能力较好。Surface Electromyography Signals（ sEMG） is a biological signal of nerve and muscle activity. This paper,starting from the root of signal generation,establishes FIR model,in which the input of the signal source is immeasurable. In order to overcome the deficiency,it presents a method of identifying model coefficient by blind identification,identifies and classifies six kinds of gestures by taking the coefficient as the features of sEMG. The experiment shows that the method is characterized by light computational load and good classification ability.

关 键 词：表面肌电信号 盲辨识 识别分类 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]