实时肌电信号采集系统设计与探究

Design and Verification of a Real-time SEMG Acquisition System

针对残疾人要完成的绝大多数任务都是由多个连续的动作组成的实际情况,仅仅对动作信号的稳态值来分析处理肌电信号的数据已经不能够满足实时性的要求。考虑到肌电信号的特性以及实验室的实际环境,搭建了一个肌电信号的硬件采集平台,并设计了3个包含连续动作的实验。利用肌电信号采集平台采集到两路正常人前臂上的表面肌电信号,应用信号幅值作为特征,并通过极值的方法分别得到3个实验中各个动作的切换点位置,为后续对连续动作的识别打下基础。

For almost all tasks are composed of several continuous actions,dealing with steady SEMG can not fits with the requirement of real-time capability.A data acquisition platform,which contains the design of hardware and human-computer interface is obtained firstly and then SEMG from flexor carpi radialis and flexor digitorum profundus are measured through three experiments with continuous actions.Relative to power spectrum,amplitude is simpler for myoelectric classifier.The extremum method is used after mapping the initial sequence into a smaller dimension vector and the place of the alter phases is obtained accurately,which can be further used in the future classification.