基于独立向量分析的表面肌电信号分解研究

摘要

A variable-step gradient algorithm of independent vector analysis is proposed,which can decompose sur-face electromyography signals. According to the surface electromyography( sEMG)physiological properties,the in-dependent vector analysis( IVA )model is applied to the frequency-domain separation of convolution mixtures of sEMG to extract motor unit action potentials information which was implied in the sEMG signal. This capability of method decomposition was compared to that of the independent component analysis( ICA)decomposition. Experi-mental results show that the IVA blind source separation algorithm has been highly effective in decomposition.%提出一种适用于表面肌电信号分解的变步长的独立向量分析梯度算法，根据表面肌电信号（ sEMG）的生理学特性，将独立向量分析（ IVA）模型应用到卷积混合肌电信号的频域分离中，提取隐含在sEMG信号中的运动单位动作电位信息。并将该方法与独立分量分析（ ICA）方法的分解性能分析比较。实验结果表明，基于IVA盲源分离技术的分解方法能得到较明显的分解效果。