Scholarlink

<p>In this paper, we propose a 4-4R rigid-flexible coupled mechanism characterized by a high load-to-weight ratio and low inertia for demanding applications such as high-precision camera gimbals, robotic joints, and motion compensation systems. High precision and dynamic response control rely on methods derived from an accurate dynamic model. Firstly, we developed a computationally efficient nonlinear dynamic model based on the product of exponential kinematic formulation. The minimum inertial parameters set is determined through QR decomposition to address the strongly coupled and nonlinear dynamic characteristics. Secondly, a stepwise identification strategy is proposed to identify precise dynamic parameters. An Iteratively Reweighted Least Squares algorithm with regularization term (IRRLS) is adopted to identify the fundamental dynamic parameters. Subsequently, an improved Stribeck friction model based on generalized coordinates is proposed to compensate the complex nonlinear friction effects. Finally, a gradient-free and parameter-adaptive covariance matrix adaptation evolution strategy (CMA-ES) is adopted to identify the nonlinear friction parameters. The orientation of the moving platform is obtained by the inertial measurement unit, and its generalized coordinates are obtained by the Newton-Raphson method. The designed excitation trajectory experiment demonstrated the effectiveness and efficiency of the proposed IRRLS-CMAES identification strategy, which is further validated through an additional experiment.</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 stepwise parameter identification of a 4-4R rigid-flexible coupled mechanism with inertial measurement unit

  • 期刊论文

  • 发表日期：

    2026年5月13日

查看全文
  • Guohua Song,
  • Sen Qian,
  • Tao Zhang,
  • Yong Liu,
  • Changqi Li,
  • Bin Zi

摘要

In this paper, we propose a 4-4R rigid-flexible coupled mechanism characterized by a high load-to-weight ratio and low inertia for demanding applications such as high-precision camera gimbals, robotic joints, and motion compensation systems. High precision and dynamic response control rely on methods derived from an accurate dynamic model. Firstly, we developed a computationally efficient nonlinear dynamic model based on the product of exponential kinematic formulation. The minimum inertial parameters set is determined through QR decomposition to address the strongly coupled and nonlinear dynamic characteristics. Secondly, a stepwise identification strategy is proposed to identify precise dynamic parameters. An Iteratively Reweighted Least Squares algorithm with regularization term (IRRLS) is adopted to identify the fundamental dynamic parameters. Subsequently, an improved Stribeck friction model based on generalized coordinates is proposed to compensate the complex nonlinear friction effects. Finally, a gradient-free and parameter-adaptive covariance matrix adaptation evolution strategy (CMA-ES) is adopted to identify the nonlinear friction parameters. The orientation of the moving platform is obtained by the inertial measurement unit, and its generalized coordinates are obtained by the Newton-Raphson method. The designed excitation trajectory experiment demonstrated the effectiveness and efficiency of the proposed IRRLS-CMAES identification strategy, which is further validated through an additional experiment.