脑-机接口驱动神经义肢手系统的研究

摘要

For the upper hand can not be rehabilitated by mordern medical approach,the characteristics of EEG for hand action are analyzed and the methods of WT feature extraction and BPNN pattern recognition of EEG signal ate studied deeply. And a control system of neural prosthesis hand driven by BCl(Brain-Computer Interface) is developed, and four movement patterns of the neural prosthesis hand(free moving grabbing opening) is finished successfully with the system. Through several off line and on-line experiments, the result shows that the neural prosthesis hand system driven by BCI is reasonable and feasible, and the method of WT feature extraction and BPNN pattern recognition is effective.%本文针对现代医疗手段还无法使上臂再生的问题,深入分析了人手动作脑电信号的特点,研究了面向义肢手控制的手动作脑电小波特征提取和BP神经网络模式识别方法,开发研制了一个基于BCI(Brain-Computer Interface,简称BCI)驱动的神经义肢手驱动控制系统,并用该系统完成了义肢手四种动作(手臂自由状态、手臂移动、手抓取、手张开)的驱动.经过多次在线及离线实验,结果表明:基于脑-机接口驱动的神经义肢手系统是合理可行的,所采用的脑电信号小波特征提取方法和BP神经网络模式识别方法是有效的.