面向遥操作机器人逆运动学求解的自适应阻尼零空间方法

打开文本图片集

中图分类号：TP242 文献标志码：A

DOI： 10.7652/xjtuxb202511020 文章编号：0253-987X（2025）11-0209-09

Abstract： To address the issues of insufficient solution accuracy in non-singular regions and joint motion uncertainty in singular regions with traditional damped least squares methods，this study proposes an inverse kinematics solving method for teleoperated robots based on adaptive damping and null-space projection. By dynamically weighting manipulability and condition number，the method monitors singular states in real time and adaptively adjusts the damping factor，achieving high-precision trajectory tracking in non-singular regions and motion stability in singular regions. Simultaneously，a joint configuration constraint term is introduced via nul-space projection to suppress the randomness of joint motions when traversing singular regions. Experiments are conducted using the gForcePro myoelectric armband as the teleoperation device to control a SIASUN GCR7-910 collaborative robot in performing a series of tasks. The results demonstrate that the proposed method achieves lower average trajectory tracking errors （ 9m in translation and （204 8×10-5 rad in rotation） in non-singular regions compared to the damped least squares method with various damping factors （0.1，0.3，0.5） . Additionally，the average maximum joint velocity across all tasks （49（∘）/s） ） is lower than that of the damped least squares method，indicating improved teleoperation trajectory tracking accuracy while ensuring motion stability and joint controllability.

Keywords： teleoperated robot； inverse kinematics； adaptive damping；null space projection

人工智能技术的发展促进了机器人在工业、农业和医疗等领域的广泛应用[1-3]。（剩余13386字）