基于动态阈值的肌电假手动作控制方法研究

摘要

由于构成肌电信号采集电路的电子元器件性能不可能完全对称及干扰信号的存在，有时会导致两路肌电信号发生阈值不一致。这时仍采用固定阈值来对两路肌电信号控制的动作进行判别，会导致动作的误判率增加。为了提高对假手动作判别的正确率，本文提出了利用动态阈值对假手动作进行判别。实验结果表明，利用动态阈值对假手动作进行判别，能够提高对动作判别的正确率达约10％。%Due to the function asymmetry of electrical components used for EMG signal acquisition circuit and the existing of undesired signals, sometimes thresholds for judging EMG's occur may be incongruent. If a fixed threshold is used to recognize Prosthetic Hand's movement,the matching rate of prosthetic hand's movement pattern recognition will be increased. In order to improve prosthetic hand's movement pattern recognition, a method of dynamic threshold was used to recognize the prosthetic hand's movement. The exprimental results showed that the method of dynamic threshold can obviously improve Prosthetic Hand's movement pattern recognition by about 10%.