Athletes muscle fatigue recognition model based on KPCA and SVM

The muscle fatigue is the inevitable phenomenon occurring in the process of athletes in sports training, usually after intense exercise or sustainability movement, characterized by muscle soreness, fatigue and so on. When the muscle fatigue to a certain extent can cause human body damage, to avoid this kind of circumstance, can choose reasonable physiological signals, determine the level of fatigue of human body, scientific and effective means of fatigue recovery can help athletes maintain good state of movement, reduce the risk of sports injury, and electromyographic signal because for observation and has high real-time performance, in terms of evaluation of muscle fatigue Attention. This paper in order to build the classification model of muscle fatigue, athletes improve the correct recognition rate of the model, first of all, analyzes the generation mechanism, features and electromyographic signal denoising method, the dimension of feature set, and reduce the redundancy between characteristics; Finally after dimension reduction and Fisher linear discriminant analysis, a new feature set K neighbor and three kinds of support vector machine classifier combination, nine fatigue classification model is set up, on muscle relaxed state, transition state of fatigue and fatigue state classification of three states, and the results show that the kernel principal component analysis and support vector machine classification model for the fatigue of the average recognition rate is highest, at 91.5%, higher than the other fatigue classification model, this method can obtain better athletes the classification results of muscle fatigue, for athletes muscle fatigue degree of judgment has important research significance.