Second-generation wavelet electromyographic signal noise eliminating method based on ensemble empirical mode decomposition

Abstract

The invention relates to a second-generation wavelet electromyographic signal noise eliminating method based on ensemble empirical mode decomposition. The method provided by the invention comprises the following steps of: obtaining electromyographic signal sample data of the human body; performing empirical modal decomposition after adding white noise in original eletromyoraphic signals; then performing secondary-generation wavelet decomposition and threshold processing on high-frequency intrinsic mode function components; performing wavelet reconstruction to high-frequency components; finally overlying processed high-frequency components and low-frequency components; and obtaining reconstituted signals as noise eliminating signals. According to the invention, as the signals are adaptively decomposed to different scales, the second-generation wavelet electromyographic signal noise eliminating method is suitable for processing nonlinear and nonstationary signals, not only has all advantages of wavelet analysis and but also has a more clear and accurate spectrum structure; in addition, the method can improve extreme point distribution of the signals, has anti-mixing decomposition capacity, can keep useful signals as far as possible, can effectively eliminate noise and can greatly improve the signal-to-noise ratio of the signals.

Description

technical field [0001] The invention belongs to the field of signal denoising, and relates to a second-generation wavelet myoelectric signal denoising method based on overall average empirical mode decomposition. Background technique [0002] Electromyography (EMG) is a comprehensive reflection of bioelectrical signals emitted by neuromuscular activities during voluntary movement of the human body on the body surface, and has been widely studied in the fields of clinical and sports medicine. At the same time, because there are different degrees of correlation between it and the active state and functional state of the muscle, it reflects the activity of the neuromuscular to a certain extent, so it has become an ideal control signal for artificial limbs and functional nerve electrical stimulation. [0003] However, since the EMG signal is usually recruited on the skin surface, it is a nonlinear and non-stationary signal, and the useful energy is mainly distributed between 0-5...

AI Technical Summary

Problems solved by technology

However, wavelet analysis is essentially a window-adjustable Fourier transform, which has problems such as difficult selection of wavelet bases, parameter sensitivity, and stationarity assumptions. Once the wavelet base is selected, the entire signal must be analyzed with the wavelet base. lack of adaptability

[0010] 3. The Empirical Mode Decomposition (EMD) method is a new adaptive time-frequency analysis method, which can perform adaptive time-frequency decomposition according to the local time-varying characteristics of the signal, and is more suitable for nonlinear , non-stationary EMG signal processing, but its important defect is the phenomenon of mode confusion

[0011] In summary, the existing EMG signal denoising methods all have deficiencies.

Images