Muscle Fatigue Recognition Based on sEMG Characteristics

Abstract

In order to provide the best rehabilitation strategy for patients with neurological hemiplegia, this study collects the surface electromyography signal (sEMG) of the human body during exercise through MYO device to identify the muscle fatigue state. Firstly, the sEMG is filtered by a 4-step Butterworth filter with zero lag, and the 50 Hz power line interference is eliminated by the power line notch method. Then, the muscle fiber motion information contained in the multicomponent EMG signal is separated into the IMF components by empirical mode decomposition (EMD). The multi-scale entropy feature of sEMG is used to identify the fatigue of muscles, and it is found that the IMF5 component entropy of sEMG is most suitable for evaluating muscle fatigue. Finally, by comparing the IMF5 component entropy values of the biceps sEMG in four different exercise modes, it is found that the elbow joint movement has the most obvious training effect on the biceps.