Skip to main navigation Skip to search Skip to main content

Dynamics compensation of impedance-based motion control for LHDS of legged robot

  • Kaixian Ba
  • , Yanhe Song
  • , Bin Yu*
  • , Xiaolong He
  • , Zhipeng Huang
  • , Chunhe Li
  • , Lipeng Yuan
  • , Xiangdong Kong
  • *Corresponding author for this work
  • Yanshan University

Research output: Contribution to journalArticlepeer-review

Abstract

Aimed at the negative effect of dynamics characteristics of leg hydraulic drive system (LHDS) on the accuracy of motion control of the hydraulic drive legged robot, a dynamics compensation control method is proposed. First, according to the mechanical structure of LHDS, the kinematics and statics models of LHDS are analyzed and obtained respectively. Based on the principle of force-based impedance control of LHDS, an impedance based motion control simulation model of LHDS is built and analyzed. The simulation results show that the dynamics characteristics have a great influence on the accuracy of impedance based motion control. Then, a dynamics compensation method considering gravity and inertia force is proposed to solve this problem. Finally, the effect of the dynamics compensation method is verified on the robot single leg test platform. The experimental results show that the compensation method reduces the negative effect of the dynamics characteristics of LHDS on impedance based motion control accuracy, and the position tracking accuracy of the robot's foot end can be improved by more than 65%. The theory proposed in this paper provides a theoretical basis for the motion control of the whole robot prototype.

Original languageEnglish
Article number103704
JournalRobotics and Autonomous Systems
Volume139
DOIs
StatePublished - May 2021

Keywords

  • Dynamics compensation
  • Impedance based motion control
  • Leg hydraulic drive system (LHDS)
  • Legged robot

Fingerprint

Dive into the research topics of 'Dynamics compensation of impedance-based motion control for LHDS of legged robot'. Together they form a unique fingerprint.

Cite this