Communications in Information and Systems

Volume 22 (2022)

Number 4

Special Issue Dedicated to Toshio Fukuda

Guest Editors: Stephen S.-T. Yau, Hisao Ishibuchi, and Naoyuki Kubota

ZMP-based fall prevention assist for lower-limb exoskeletons during dynamic walking

Pages: 499 – 526

DOI: https://dx.doi.org/10.4310/CIS.2022.v22.n4.a4

Authors

Oybek Rashidov (Department of Mechanical Engineering, Kyushu University, Nishi-ku, Fukuoka, Japan)

Kazuo Kiguchi (Department of Mechanical Engineering, Kyushu University, Nishi-ku, Fukuoka, Japan)

Abstract

A lower-limb power-assist exoskeleton robot is a wearable device that assists persons in their daily activities. Although the main purpose of the lower-limb exoskeleton robot is to assist the intended motion of the user, the stability of the robot’s posture is still one of the biggest challenges. This paper focuses on the balance aspect of the exoskeleton robot in terms of control considering Zero Moment Point (ZMP) that is widely used in biped robots. A ZMP based fall prevention assist method, that prevents the user from falling down by giving equivalent external motion modification force to the user during dynamic walking on even terrains, is proposed in this paper. Fall prevention assist strategy is changed between a double support phase and a single support phase in the proposed method. Position of swing leg, velocity of swing leg, and change rate of ZMP are considered as an index to stabilize the posture of the exoskeleton during dynamic walking in the proposed method. In the proposed method, fuzzy control approach is applied to keep the ZMP inside the support polygon by generating the equivalent motion modification force at the chest or back of the user by the lower-limb exoskeleton for fall prevention assist during walking without making additional steps. The effectiveness of the proposed method is evaluated through experiments.

Received 9 November 2020

Published 18 January 2023