Skip to main navigation Skip to search Skip to main content

近似直驱双足机器人跳跃运动非线性优化及试验验证

Translated title of the contribution: Nonlinear Optimization and Experimental Validation of a Quasi-direct-drive Bipedal Robot's Jumping Motion
  • Kaizheng Shan
  • , Haitao Yu*
  • , Liangliang Han
  • , Shengjun Wang
  • , Jun Li
  • , Haibo Gao
  • , Zongquan Deng
  • *Corresponding author for this work
  • Harbin Institute of Technology
  • China Aerospace Science and Technology Corporation

Research output: Contribution to journalArticlepeer-review

Abstract

Jumping is one of the most common gait patterns for cursorial mammals. With rapid and impulsive output, animals can cross obstacles and escape from enemies. However, subject to the limitations of actuator capabilities and the difficulty of planning and controlling highly dynamic movements, it is extremely challenging for legged robots to achieve movements such as jumping and running. In view of this, a bipedal robot prototype is devised with the co-axial quasi-direct-drive structure. Based on the simplified three-bar model, the constraint conditions of stance phase and flight phase are confirmed. With the goal of optimal jumping height, the motion planning issue of the robot is transferred into a nonlinear optimization problem with finite dimensions, and the variable parameter PID is used for the joint torque tracking control. The effectiveness of the motion planning algorithm is verified on physical robot platform. The results show that the biped robot can achieve jumping with ground clearance of 0.34 m (equivalent to 50% of the leg length). The research results can provide theoretical support and technical reference for future research on legged robots with remarkable obstacle negotiation ability.

Translated title of the contributionNonlinear Optimization and Experimental Validation of a Quasi-direct-drive Bipedal Robot's Jumping Motion
Original languageChinese (Traditional)
Pages (from-to)153-162 and 171
JournalJixie Gongcheng Xuebao/Chinese Journal of Mechanical Engineering
Volume57
Issue number13
DOIs
StatePublished - 5 Jul 2021

Fingerprint

Dive into the research topics of 'Nonlinear Optimization and Experimental Validation of a Quasi-direct-drive Bipedal Robot's Jumping Motion'. Together they form a unique fingerprint.

Cite this