Scholarlink

<p>This paper investigates the precise dynamic modeling and base dynamic parameter identification of a 6-UPRU parallel manipulator with branch force sensors. First, the kinematic model is established, and an accurate dynamic model considering the effects of passive rotational joints is developed using the Newton–Euler method. To address the difficulty in parameter extraction caused by high coupling and strong nonlinearity, a symbolic computation rule is proposed to extract dynamic parameters, enabling a linear representation of the dynamic model with respect to the dynamic parameters. Second, the base dynamic parameters and corresponding closed-form reduced dynamic model are derived via QR decomposition of the observation matrix, reducing the number of parameters from 29 to 17. Furthermore, based on the optimized fifth-order Fourier series excitation trajectory, physically feasible solutions for the base dynamic parameters are identified using the iteratively reweighted least-squares (IRLS) algorithm with physical constraints, resolving the issue of physical infeasibility in traditional identification methods. Finally, the correctness of the model is validated through SimMechanics simulations, and identification experiments are conducted to verify the accuracy of the identified parameters.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dynamic modeling and base dynamic parameter identification of the 6-UPRU parallel manipulator with branch force sensors

  • 期刊论文

  • 发表日期：

    2025年10月2日

查看全文
  • Tao Ni,
  • Yahui Zhao,
  • Panhong Zhang,
  • Qingchuan Ning,
  • Zeren Zhao,
  • Jingxin Liang

摘要

This paper investigates the precise dynamic modeling and base dynamic parameter identification of a 6-UPRU parallel manipulator with branch force sensors. First, the kinematic model is established, and an accurate dynamic model considering the effects of passive rotational joints is developed using the Newton–Euler method. To address the difficulty in parameter extraction caused by high coupling and strong nonlinearity, a symbolic computation rule is proposed to extract dynamic parameters, enabling a linear representation of the dynamic model with respect to the dynamic parameters. Second, the base dynamic parameters and corresponding closed-form reduced dynamic model are derived via QR decomposition of the observation matrix, reducing the number of parameters from 29 to 17. Furthermore, based on the optimized fifth-order Fourier series excitation trajectory, physically feasible solutions for the base dynamic parameters are identified using the iteratively reweighted least-squares (IRLS) algorithm with physical constraints, resolving the issue of physical infeasibility in traditional identification methods. Finally, the correctness of the model is validated through SimMechanics simulations, and identification experiments are conducted to verify the accuracy of the identified parameters.