Nonlinear adaptive controller for omni-directional walker: Dynamic model improvement and experiment

This paper describes an omni-directional walker (ODW), which is being developed for both walking rehabilitation and walking support for people with walking disabilities. The ODW needs to accurately follow the path referring to the exercise prescription from physical therapists. However, path tracking error happens during walking rehabilitation. This paper is part of an effort aiming to improve path tracking accuracy of the ODW considering the center-of-gravity shifts (COG) and load changes caused by user. In previous studies, a nonlinear adaptive controller was proposed based on a gravity model of the ODW. In this paper, in order to precisely describe the plant, a new center-dynamic model of the ODW and an improved nonlinear adaptive control method are derived. We conducted experiments in which the ODW was controlled to follow a linear path as a common rehabilitation path. Experiment results showed that the improved strategy is effective to deal with the COG shifts and load changes.