Muscle Fatigue Detection and Classification based on Electromyography Signal for Riding Motorcycle

In both rural and urban regions in Asian countries, where rates of motorcycle ownership are seven times greater than the global average. Especially on long rides, the majority of bikers frequently experience muscle fatigue. Electromyography (EMG) is a biological signal that detects electrical currents generated during muscle contraction. The presented study did not have a standard in measuring the degree of muscle fatigue. Thus, this paper is proposed to extract the EMG signal for muscle fatigue detection onset using time-frequency distribution, S-Transform is utillised. Muscle fatigue features are extracted from S-Transform then classified into two types of non-fatigue and three types of fatigue muscles using percentage for maximum voluntary contraction (MVC). The classifiers from machine learning is automated and classified the condition of the respondent automatically. Results have proven S-Transform to be utilised with significant parameters for measuring muscle fatigue onset with higher accuracy. Then ANN with 100% of classifiers is chosen as the best classifier to be use for auto classification performance of muscles for each respondent. As the conclusion, this finding and the guidance can be used for further study in muscle fatigue not only for riding motorcycle, but also another task involved EMG signal.