Robust Locomotion Control of a Self-Balancing and Underactuated Bipedal Exoskeleton: Task Prioritization and Feedback Control

This letter presents a study to understand how an underactuated bipedal exoskeleton with an arbitrary human user can exhibit robust dynamic walking behavior despite severe parameter uncertainty and external disturbances. Unlike in the case of classical bipedal robots where each leg possesses 6 or more active joints, it is very challenging to satisfy multiple constraints simultaneously to ensure robust dynamic walking. To overcome this problem, we first propose an optimization algorithm that makes use of a prioritized stack of tasks to satisfy the constraints hierarchically. Furthermore, we synthesized a locomotion controller named ZMP impedance feedback and identified two other state-of-the-art locomotion controllers (admittance control, centroidal momentum control) all of which were built on top of the proposed task prioritization algorithm. In order to verify the validity and robustness of these controllers for a thorough benchmarking, a series of simulation experiments were conducted via MSC ADAMS in which a human-robot coupled model with a 40 kg underactuated exoskeleton and 12 distinct anthropomorphic subjects (66 similar to 102 kg) was considered. As the result, all 3 controllers showed adequate performances to address balanced locomotion behavior when used with the proposed task priority-based optimization algorithm. In addition, the proposed ZMP impedance controller showed statistically significant results, indicating its comparatively more robust feature.