• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2025, Vol. 61 ›› Issue (17): 15-26.doi: 10.3901/JME.2025.17.015

• 机器人及机构学 • 上一篇    

扫码分享

机器人关节误差分析及参数辨识

严文俊, 蔡月日   

  1. 北京航空航天大学机器人研究所 北京 100191
  • 收稿日期:2024-05-24 修回日期:2024-09-05 发布日期:2025-10-24
  • 作者简介:严文俊,男,1999年出生。主要研究方向为机器人高精度控制技术。E-mail:wen_jun_yan@163.com;蔡月日(通信作者),男,1983年出生,副教授,博士研究生导师。主要研究方向为仿生机器人、特种机器人。E-mail:caiyueri@buaa.edu.cn
  • 基金资助:
    基础产品创新科研资助项目。

Error Theoretical Analysis of Robot Joint and Research on Parameter Identification Methods

YAN Wenjun, CAI Yueri   

  1. Robotics Institute, Beihang University, Beijing 100191
  • Received:2024-05-24 Revised:2024-09-05 Published:2025-10-24

摘要: 机器人关节往往具有很强的非线性刚度与迟滞、非线性摩擦以及运动学传递误差等复杂的动力学特性,这显著影响了关节的定位精度以及运动平稳性,因此分析机器人关节的误差并研究其参数辨识方法至关重要。基于PSO-LSSVM算法得到机器人关节精确的Preisach模型,用以描述非线性刚度与迟滞特性;基于FFT算法得到运动学传动误差模型中高幅值分量;基于Levenberg-Marquardt算法精确获得Stribeck模型以描述机器人关节的非线性摩擦特性。最后,根据对误差模型以及辨识算法的分析结果,通过设计实验验证了模型及辨识算法的可靠性。研究可为后续设计高精度的机器人关节控制器提供参考。

关键词: 机器人, 关节, 误差模型, 参数辨识

Abstract: The joints of robots often exhibit complex dynamic characteristics such as strong nonlinear stiffness and hysteresis, nonlinear friction, and kinematic transmission errors. These factors significantly impact the positioning accuracy and smoothness of joint movements. Therefore, analyzing joint errors and studying parameter identification methods are crucial. This research uses the PSO-LSSVM algorithm to obtain an accurate Preisach model for robot joints, describing their nonlinear stiffness and hysteresis characteristics. The FFT algorithm is employed to identify the high amplitude components in the kinematic transmission error model. The Levenberg-Marquardt algorithm is used to accurately obtain the Stribeck model, which describes the nonlinear friction characteristics of robot joints. Finally, based on the analysis results of the error models and identification algorithms, experiments were designed to verify the reliability of the models and identification methods. This research can provide references for the subsequent design of high-precision robot joint controllers.

Key words: robotic, joints, error model, parameter identification

中图分类号: