二阶统计量盲分离技术在表面肌电信号分解过程中的应用

The application of blind source separation techniqnes using second order statistics in the decomposition of surface EMG

采用基于二阶统计量的盲源分离算法对多导表面肌电信号进行处理,实现噪声的分离和表面肌电信号的初步分解。实验结果表明,无论是对仿真表面肌电信号还是真实表面肌电信号,二阶盲分离方法具有良好的处理结果,其中,SEONS算法的分解性能最佳。

Some methods of SEMG signal pretreatment based on blind source separation using second order statistics were proposed for noise separation and the elementary decomposition of multi-channel SEMG. The experiment results showed that these methods of BSS, SOBI, Robust whitening and SEONS, especially the SEONS algorithm, could decompose the SEMG effectively, no matter for the simulated or recorded SEMG signal.