Skip to main navigation Skip to search Skip to main content

A Single-Source Dual-Drive Flexible Knee Assistive Exoskeleton for Elderly Daily Locomotion

  • Shisheng Zhang
  • , Yang Zhang
  • , Yanzong Xu
  • , Jinke Li
  • , Yuquan Leng*
  • , Xinyu Wu
  • *Corresponding author for this work
  • Shenzhen Institute of Advanced Technology
  • University of Chinese Academy of Sciences
  • Shenzhen Technology University
  • Shenzhen Milebot Robotics Company Ltd.
  • School of Biomedical Engineering, Harbin Institute of Technology Shenzhen

Research output: Contribution to journalArticlepeer-review

Abstract

The knees, as important parts in daily locomotion, carry the body weight and provide stability and flexibility. With the degradation of muscle and endurance, the elders get weak knee strength, which heavily affects their life quality. To address this issue, a single-source dual-drive flexible knee assistive exoskeleton is explored to assist the elders in their daily locomotion. Firstly, a single-source dual-drive architecture is designed. The left and right knee are controlled by single electric motor with clutch. It reduces the weight of exoskeleton and has the potential to enhance interaction safety and comfort with cable-driven and elastic elements. Secondly, a cyclic locomotion pattern recognition (LPR) methodology for daily walking situation is provided, adopted with a dual detection strategy fused with finite state machine and fuzzy control. Daily walking situations, e.g., level walking (LW), ramp ascending (RA), ramp descending (RD), stair ascending (SA), and stair descending (SD), could be recognized accurately and precisely. Finally, a time-shared assistive torque control strategy with finite state is conducted to effectively provide suitable auxiliary torque in various walking environments based on locomotion patterns. Three young participants and 3 elder participants participate in outdoor LPR experiments in real-world environment and indoor energy loss experiments. The initial experiment results indicate that the average precision of LPR reaches 98.89%, and the average metabolic cost and surface electromyography signals reduce as much as 28.20% and 67.33%, respectively, with the assistance of exoskeleton.

Original languageEnglish
Article number0576
JournalCyborg and Bionic Systems
Volume7
DOIs
StatePublished - Jan 2026
Externally publishedYes

Fingerprint

Dive into the research topics of 'A Single-Source Dual-Drive Flexible Knee Assistive Exoskeleton for Elderly Daily Locomotion'. Together they form a unique fingerprint.

Cite this