Research on CNN-based Fatigue State Assessment Method for Muscle Pressurization Training

Pressure training is a popular fitness programme, and monitoring muscle fatigue state during training can be beneficial for those who perform muscle pressure training. Bioimpedance detection technology has been developed in response to the need for fatigue monitoring in muscle compression training. Its principle is based on the individual anisotropy of impedance exhibited by muscles in different states. In light of this, this paper set out to investigate the characteristics of the blood flow impedance signal of the muscle under the condition of pressurised training. The aim was to use the waveform characteristic parameters of the muscle blood flow impedance signal to monitor the muscle fatigue state during pressurised training. In addition, the paper explored the potential of deep learning technology to assess the fatigue state during the pressurised training process. A classification method of muscle fatigue state during pressurised training was designed based on the convolutional neural network model. The cycle segments of muscle blood flow impedance signals of non-fatigue state and fatigue state were input into the convolutional neural network model for identification and classification. Ultimately, the method achieved a classification accuracy of up to 89.7%. The results indicate that the method proposed in this study has the potential to classify muscle blood flow impedance signals during pressurised training. It may therefore be of benefit in guiding pressurised training and preventing excessive muscle fatigue.