In this study, we checked experimentally whether anterior-posterior accelerations of the head during quiet human stance are usually below or above known thresholds of the otolith sensor. Thereto, we measured head kinematics with high spatial resolution. Furthermore, we used both these experimental data and computer simulations of two double inverted pendulum (DIP) models in order to verify the validity of DIP models in general. The results are clear cut. First, not only are acceleration thresholds regularly exceeded about once a second but also are velocity thresholds exceeded, albeit probably less frequently. Second, COM and head movement predicted by interwoven DIP model dynamics can not reproduce the mean measured amplitudes at once. Thus, neither the formerly promoted single inverted pendulum nor any DIP model can causally explain the dynamics of quiet human stance. Instead, we suggest to factor in at least three mechanical degrees of freedom. Due to a couple of reasons discussed, the triple inverted pendulum (TIP) model seems to be a promising abstraction implying potential to better understand the dynamics of quiet human stance.
List of symbols(time) sequence An array of values of measured variables (positional or force components) sampled discretely versus time trial Acquisition of one (consistent and synchronised) data set containing all (time) sequences of the measured variables P Number of subjects (9) N Number of trials per subject (10) M