Feature Extraction and Classification of sEMG Based on ICA and EMD Decomposition of AR Model

摘要

The surface EMG (sEMG) is a biological electrical signal of neuromuscular activity distribution. From the point of the non-stationary and nonlinear, the independent component analysis method is firstly used to eliminate the power frequency interference in sEMG. Secondly, the low noise signal is processed by empirical mode decomposition (EMD), then use the decomposed signal to establish AR model. The model coefficients are used as signal features and PNN optimized by particle swarm optimization (PSO) is used to classify six types of forearm motions. The experimental results demonstrate the effectiveness of the proposed method.