不応期長さにランダム性を考慮した振戦モデル

抄録

Even in a state that the person is resting, the person vibrates with a small amplitude. This vibration is called human tremor. For example, when performing precise work such as watch assembly or surgical operation, the tremor will have an influence on the accuracy of work. In order to suppress the tremor, frequency spectrum analysis of the tremor is necessary. However, only dominant frequency of the tremor have been calculated in previous works. In this paper, human tremor model for frequency spectrum analysis is proposed. In the model, the tremor is regarded as forced vibration. Forearm is modeled as a rigid link, which is antagonistically driven by two Hill's muscle models. The stiffness of muscle model is considered as non-linear. This makes it possible to estimate the difference of frequency spectrum of the tremor occurs due to different angle of the elbow. Moreover, the force driving the forearm is generated from muscle by using leaky Integrate-and-Fire cell model. In a previous study, it was confirmed that neural cell had a refractory period, and there was randomness in the period. Therefore, in this study, the refractory period of the cell model is set to the normal distribution. Because the force is intermittently generated by the random period, the forearm is vibrated. The vibration spectrum obtained from the simulations and that obtained from the experiments quantitatively agreed.