Randomness evaluation using the runs test of COP time series data when taking a step

Abstract

The purpose of this study is to evaluate the randomness of the center of foot pressure (COP) time series data using the runs test. COP is used for the evaluation of human standing balance. Some evaluation indexes obtained from COP, such as total trajectory length, sway width, and time-frequency analysis, do not reflect the information as a time series. Therefore, we evaluated the randomness of human upright COP time series data using the runs test. The subjects were two women in their 50s who were independent in their daily lives. The measurement positions were both-legged stance to one-legged stance. COP data for about 30 seconds before and after the one-legged stance were collected at a sampling frequency of 100 Hz. Many time-series data obtained from the COP data, such as distance from the mean position, movement distance, movement angle, and direction of acceleration, showed somehow regularity by rejecting randomness hypotheses.