基于小波和非线性指标的表面肌电信号动作特征分析

Characterization of surface electromyography signal based on wavelet analysis and non-linear exponent

由于表面肌电信号的复杂性,使得寻找到适合其分类的特征是非常困难的事情,因此,特征选取在表面肌电信号的模式识别中有着决定性的作用。表面肌电信号具有非平稳,非线性的特性。小波分析是一种分析非平稳信号的有效工具,而最大Lyapunov指数已广泛应用于判定非线性指标中。文章基于小波分析和Lyapunov指数的特点,提出了算法上的改进方案,利用小波分析和最大Lyapunov指数相结合的方法对表面肌电信号进行分类识别,取得了很好的分类效果。

Due to the complexity of surface electromyography （SEMG） signal, it is difficult to find features for its classification. Thus feature extraction from SEMG signal is critical to its pattern recognition. The characteristics of non-stationary and non-linearity have been found in some researches. Wavelet analysis is effective in analyzing non-stationary signal. And the largest Lyapunov exponent was extensively used to testify the non-linearity of a system. Based on the characteristics of wavelet analysis and Lyapunov exponent, this study modified the algorithm to classify and recognize the SEMG signal by binding wavelet analysis and the largest Lyapunov exponent, which has achieved a better classification.