基于小波分析与神经网络相结合的表面肌电信号识别的研究

The Classification of Surface EMG Signal Based on Wavelet Transform and Neural Networks

表面肌电信号是从人体骨骼肌表面通过电极记录下来的神经肌肉活动发放的生物电信号,具有非平稳性和复杂性的特点。本研究通过使用小波分析与神经网络相结合的方法,识别正常肌电信号与疲劳肌电信号。实验表明,将小波分解后的肌电信号代替原始肌电信号,能明显提高神经网络对肌电信号的识别准确率。

Surface EMG （sEMG） signal is the noninvasive recording of electrical activity of muscle, which poses the characters of nonstationary and complexity. A method of using wavelet transform and neural networks to classify the normal and fatigue sEMG is provided. Experimental result shows that replacing the raw SEMG by wavelet decomposition sEMG can enhance the correctness greatly.