Monitoring Accumulative Fatigue of Finger by Autoregressive Modeling of Physiological Tremor

抄録

In this study, block data structured autoregressive (AR) method is used to evaluate fatigue, based on physiological tremor during and after loading a weight mass on the index finger. The temporal changes in the prediction coefficients and the reflection coefficients are determined. AR spectral estimation with the ninth-order is obtained and presented in graphical form. The results indicate that the first prediction coefficient a₁ can be used to characterize the state of fatigue of finger muscle and the other prediction coefficients do not show any tendency for the finger load. The coefficient a₁ can be applied to monitor the accumulative fatigue induced by the weight loading for a duration of time.