Skip to main navigation Skip to search Skip to main content

Kinematics analysis and performance optimization of a novel asymmetric parallel biped robot

  • Harbin Institute of Technology

Research output: Contribution to journalArticlepeer-review

Abstract

The asymmetric parallel mechanism offers advantages such as strong bearing capacity, flexible kinematics, and a large workspace, which conventional mechanisms cannot simultaneously achieve. However, few studies have applied it to the structural design of biped robots, resulting in a lack of a comprehensive theoretical system for analyzing the kinematic characteristics of such biped robots. This paper addresses this gap by analyzing a novel type of asymmetric parallel biped robot with common hinges and optimizing its dimensions. A method for degrees of freedom analysis based on the screw system is introduced. By combining the screw method with Lie algebra, a decoupled kinematic model is established, and the Jacobian matrix is derived. The workspace is determined using the Monte Carlo method. The distribution of singularity, the dexterity over time, and the stiffness mapping in two-dimensional space are obtained and analyzed based on the Jacobian matrix. Particle swarm optimization is employed to optimize the linkage length. Finally, the improvement of the optimized performance and the correctness of the kinematic derivation are validated by the gait experiment. This study offers valuable insights into exploring the kinematic characteristics of asymmetrical parallel mechanisms with common hinges, not only within the realm of biped robots but also in the broader field of multi-legged robots.

Original languageEnglish
Article number105003
JournalRobotics and Autonomous Systems
Volume191
DOIs
StatePublished - Sep 2025

Keywords

  • Asymmetric parallel mechanism
  • Biped robot
  • Kinematics characteristics
  • Monte-Carlo method
  • Multi-objective optimization
  • Stiffness mapping

Fingerprint

Dive into the research topics of 'Kinematics analysis and performance optimization of a novel asymmetric parallel biped robot'. Together they form a unique fingerprint.

Cite this