The improved extrapolated center of mass enhances the safety of exoskeleton system.

Journal: Scientific reports
PMID:

Abstract

Maintaining the balance and safety of the exoskeleton human-robot coupling system is a prerequisite for realizing the rehabilitation training function. Therefore, research on the balance of lower limb exoskeleton robots has attracted much attention. When the exoskeleton human-robot coupling system reaches the critical state of falling, there is an issue with inaccurate detection in the extrapolated center of mass (XCoM) balance index. This paper firstly establishes an inverted pendulum model after the system is disturbed, fully considering that the external environment may exert a disturbance force on the system at any time, and proposes an improved extrapolated center of mass(PXCoM) balance index based on XCoM. Secondly, we aim to the problem of balance recovery in the critical state of the human-exoskeleton system falling. Using an ankle joint under-actuated exoskeleton robot as a research object, we study the balance recovery method of the exoskeleton based on active stepping strategies under large disturbance states. Finally, a lower limb exoskeleton robot experimental platform was built, on which the balance sensing method and balance recovery mechanism studied in this article were transplanted, and its feasibility was verified. The experiment results show that the PXCoM balance evaluation index proposed in this paper has a better perceptual effect than the XCoM balance evaluation index in the critical state of the human-exoskeleton system falling. The proposed gait recovery strategy can effectively restore the balance of the human-exoskeleton system under significant disturbances, thereby ensuring safety under disturbances while wearing the exoskeleton.

Authors

  • Chang Wang
    Key Laboratory of the plateau of environmental damage control, Lanzhou General Hospital of Lanzhou Military Command, Lanzhou, China.
  • Jian Cao
    Department of Mechanical Engineering, Northwestern University, Evanston, IL, 60208, USA.
  • Jianhua Zhang
  • Junhui Wang
    Institute of Artificial Intelligence and Robotics, College of Artificial Intelligence, Xi'an Jiaotong University, Xi'an 710049, China.
  • Qiang Yang
  • Song Men
    Inspur Yunzhou Industrial Internet Co., Ltd., Jinan, 250101, China.