Dynamical analysis of human standing model with cyclic motion

Human standing is characterized by large body sway, which cannot be explained by linear control. In past researches, sway has been considered as an uncontrolled biological noise. In contrast, we consider the sway to be a cyclic motion generated by continuous proportional-integral-derivative (PID) control with weak nonlinearlity. Through mathematical analysis of nonliner PID control, cyclic motion is shown to be generated by a stability-gain dependent Hopf bifurcation, and biological noise is shown to help the smooth transition between stationary stable state and cyclic state. The relevance of the proposed sway generation mechanism is verified through human experiment on floors with different stability. As a result, the existence of Hopf bifurcation, i.e., apparent expansion of sway with small decrease of control parameters, realized by the destabilization of floor, is observed.