基于人体生物力学阻抗辨识的月面建造远程变阻抗控制

Driven by the continuous progress in lunar exploration, teleoperated robotic arms require highly safe, accurate, and transparent control strategies to handle uncertain and unstructured environments during lunar base construction. Humanoid variable impedance control ensures both safe environmental interaction and high-precision tracking, providing a robust solution for human-robot collaboration. This study investigates a teleoperation strategy that maps human impedance parameters onto a remote robotic arm to meet the interactive demands of lunar tasks. By integrating four-channel surface ElectroMyoGraphy (sEMG) signals with an upper limb mechanics model (built upon Hill’s model and kinematics), a real-time identification system for human end-effector stiffness is established. Unlike conventional methods, this strategy incorporates personalized physical parameters to enhance the generalization of humanoid impedance control. Furthermore, force and visual feedback are utilized to improve information transparency and leverage natural neural reflexes for adaptive impedance adjustment. Finally, experimental results on a lunar truss assembly platform demonstrate that the proposed humanoid variable impedance control significantly outperforms traditional teleoperation schemes.