• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2022, Vol. 58 ›› Issue (9): 98-106.doi: 10.3901/JME.2022.09.098

• 机械动力学 • 上一篇    下一篇

扫码分享

滚珠丝杠传动系统线性变参数动力学模型试验建模方法

黄涛1,2, 杜双江1,2, 张钎1,2, 曹华军1,2   

  1. 1. 重庆大学机械与运载工程学院 重庆 400044;
    2. 重庆大学机械传动国家重点实验室 重庆 400044
  • 收稿日期:2021-05-18 修回日期:2022-01-08 出版日期:2022-05-05 发布日期:2022-06-23
  • 通讯作者: 曹华军(通信作者),男,1978年出生,博士,教授,博士研究生导师。主要研究方向为绿色低碳制造基础理论,高速干切工艺及智能装备,高性能传动系统。E-mail:hjcao@cqu.edu.cn E-mail:hjcao@cqu.edu.cn
  • 作者简介:黄涛,男,1986年出生,博士,副教授,博士研究生导师。主要研究方向为精密/超精密传动系统动力学与控制。E-mail:thuang@cqu.edu.cn
  • 基金资助:
    国家重点研发计划(2018YFB1701203)、国家自然科学基金(51805053)和中央高校基本科研业务费专项资金(2020CDJGFJX010)资助项目

Experimental Modeling of Linear Parameter Varying Dynamics Model of Ball Screw Drive System

HUANG Tao1,2, DU Shuangjiang1,2, ZHANG Qian1,2, CAO Huajun1,2   

  1. 1. College of Mechanical and Vehicle Engineering, Chongqing University, Chongqing 400044;
    2. State Key Laboratory of Mechanical Transmission, Chongqing University, Chongqing 400044
  • Received:2021-05-18 Revised:2022-01-08 Online:2022-05-05 Published:2022-06-23

摘要: 准确建立滚珠丝杠传动系统动力学模型是实现高速、高加速度和高精度运动控制的基础。滚珠丝杠传动系统的动力学特性具有不确定性,实际是随位置、负载质量等变量变化的,数学上应描述为变参数模型。因此,一种滚珠丝杠传动系统线性变参数(Linear parameter varying, LPV)动力学模型实验建模方法被提出。通过局部频域辨识方法获得非参数模型,并采用非线性最小二乘方法得到模态叠加形式的传递函数模型。采用傅里叶级数形式的正交基函数拟合方法,对不同位置下的局部传递函数模型参数进行拟合,获得与位置相关的参数模型。基于拟合结果,选择合适的工况点进行三次样条插值,得到与位置、负载质量相关的LPV模型。所提出方法应用于滚珠丝杠试验台,获得了一个与全局位置、负载质量相关性高的LPV模型,为实现基于LPV模型的高速、高精度运动控制提供了基础。

关键词: 滚珠丝杠, 线性变参数, 系统辨识, 参数拟合, 插值

Abstract: Accurate modeling of the dynamics of ball screw drive system is the foundation for high speed, high acceleration and high precision motion control. The dynamics of the ball screw drive system has uncertainties and actually changes with position, load and other variables, which should be described as a variable parameter model in mathematics. Therefore, an experimental modeling method of linear parameter varying (LPV) dynamic model of ball screw drive system is proposed. Specially, the nonparametric model is obtained by the local frequency domain identification method, and the transfer function model in the form of modal superposition is obtained by the nonlinear least squares method. Then, an orthogonal basis function fitting method in the form of Fourier series is utilized to fit the local transfer function model parameters at different positions to obtain the position-dependent parameter model. Based on the fitting results, the position- and load-dependent LPV model is obtained by selecting suitable working points for third spline interpolation. The proposed method is applied to a developed ball screw experimental system to obtain a LPV model with high correlation to the full working position and load, which provides a basis for realizing high-speed and high-precision motion control based on the LPV model.

Key words: ball screw, linear parameter varying, system identification, parameter fitting, interpolation

中图分类号: