Towards Perturbation-free Identification of Human Standing Balance

Abstract

Balance disorders affect millions in the United States alone. Despite the large body of literature in the field, we do not yet fully understand the neuromuscular control of balance. A method to estimate natural, unperturbed standing balance dynamics would provide a necessary tool to examine the apparent control mechanisms employed by healthy individuals, and lead to insight and inspiration for developing effective rehabilitation and assistive technologies for those with balance impairments.

In this thesis, a correlation-based system identification method has been investigated as a candidate for perturbation-free identification of human standing balance. The method was tested in simulation to understand its strengths and limitations, and was successfully validated on a hardware system. However, existing human quiet standing data revealed that the posture control process cannot be modelled by a stationary process at the time scales of interest, as required by the system identification method. Accordingly, the perturbation-free system identification of balance dynamics and control remains an area for ongoing research.