| Penulis/Author |
TRIWIYANTO (1); Prof. Ir. Oyas Wahyunggoro, MT., Ph.D. (2); Prof. Ir. Hanung Adi Nugroho, S.T., M.Eng., Ph.D., IPM., SMIEEE. (3); Prof. Dr.Eng. Ir. Herianto, S.T., M.Eng., IPU., ASEAN Eng. (4) |
| Abstrak/Abstract |
Fatigue is a state that the muscle could not able to
maintain the contraction. Electromyography signals can be used
to determine the state of muscle fatigue. Quantization of muscle
fatigue needs to be defined clearly so that it can be used as an
indicator or a compensator in the control system.
Electromyography signal which is produced during dynamic
motion during muscle fatigue assessment is a non-stationary
signal. In this study, the discrete wavelet transforms method was
used to analyze electromyography signals. The electromyography
signal was decomposed up to 5th scale. Features extraction, root
mean square, mean average, standard deviation and power signal
were used to observe the correlation coefficient and the slope of
the regression line. Our finding from the results of the analysis
with multi-resolution discrete wavelet transforms showed that the
frequency ranged 62.5 Hz up to 125 Hz was the dominant
frequency of the EMG signal and feature of power was better
among other features to observe the muscle fatigue. In this study,
the correlation coefficient was 0.852 and the best slope was 0.728
mV/s. These value can also be used as an indicator or an estimator
of the state of the fatigue and used as a compensator in the system
based myoelectric control. |
