Seiler, Julian ; Schäfer, Niklas ; Zhao, Guoping ; Latsch, Bastian ; Grimmer, Martin ; Beckerle, Philipp ; Kupnik, Mario (2024)
Human-Exoskeleton Interaction Force Estimation Based on Quasi-Direct Drive Actuators.
10th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob). Heidelberg, Germany (01.09.2024-04.09.2024)
doi: 10.1109/BioRob60516.2024.10719722
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

Exoskeletons have emerged as a promising tech-nological solution to provide functional compensation, training support, physical enhancement, and rehabilitation for an aging population. High torque density and control bandwidth are essential for exoskeleton actuation. In addition, monitoring and control of human-exoskeleton interaction forces is essential for the effectiveness, safety, and comfort of exoskeletons. Quasi-direct drives (QDD) have the potential to address the actuation requirements by enabling proprioceptive actuation with low mechanical output impedance. We present a test bench system to identify and validate the system dynamics, including torque constant, inertia, and friction properties of the QDD Cube-Mars AK10-9 V1.1. The actuator is employed in a hip exoskeleton to evaluate open-loop torque tracking and interaction force estimation in an assisted gait scenario with one participant walking on a treadmill. The model estimates interaction forces with a mean absolute error (MAE) of 2.78±0.58N (6.4% of rated force output). Model-based open-loop control improves the torque tracking MAE by 23%. Our analysis indicates that considering the soft coupling between human and exoskeleton has the potential to further improve the torque tracking and interaction force estimation accuracy. Altogether, the findings demonstrate that QDD actuation enables backdrivability, re-liable force estimation and controllable assistance, thereby providing an effective solution for exoskeleton actuation.

Frage zum Eintrag

Redaktionelle Details anzeigen