中国机械工程 ›› 2026, Vol. 37 ›› Issue (1): 51-59.DOI: 10.3969/j.issn.1004-132X.2026.01.006

• 机械基础工程 • 上一篇    下一篇

复杂工况下磨齿机主轴运行模态的分析方法

李国龙1(), 赵晓亮1, 王玉1, 陶一杰2   

  1. 1.重庆大学高端装备机械传动全国重点实验室, 重庆, 40044
    2.东风汽车集团有限公司研发总院, 武汉, 430058
  • 收稿日期:2024-06-28 修回日期:2025-12-25 出版日期:2026-01-25 发布日期:2026-02-05
  • 通讯作者: 李国龙
  • 作者简介:李国龙*(通信作者),男,1968年生,教授、博士研究生导师。研究方向为智能制造技术与系统、复杂零件数字化制系统与装备、精密/超精密加工技术。发表论文144篇。E-mail:glli@cqu.edu.cn
    第一联系人:吴健鹏*(通信作者),男,1991年生,副研究员。研究方向为车辆传动系统智能运维、数字孪生仿真建模方法。发表论文32篇。E-mail: 15811319103@163.com.
  • 基金资助:
    国家自然科学基金(52275475);重庆英才计划(cstc2022ycjh-bgzxm0060)

Operation Modal Analysis Method of Gear Grinding Machine Spindle Operations under Complex Working Conditions

LI Guolong1(), ZHAO Xiaoliang1, WANG Yu1, TAO Yijie2   

  1. 1.State Key Laboratory of Mechanical Transmission for Advanced Equipment,Chongqing University,Chongqing,400044
    2.Research & Development Institute,Dongfeng Motor Corporation,Wuhan,430058
  • Received:2024-06-28 Revised:2025-12-25 Online:2026-01-25 Published:2026-02-05
  • Contact: LI Guolong

摘要:

针对磨齿机主轴服役状态下振动形式复杂、模态特征难以有效识别的问题,提出一种基于自适应噪声完备集合经验模态分解与相关性分析的方法。采用有限元模态分析方法定义频带范围,采用小波阈值分级法保留模态特征信息。采用倒频谱法编辑信号,以识别并剔除转子产生的谐波响应。不同降噪方法与二自由度算例的验证结果表明,所提方法处理后的模态识别误差减小至1.3%,极点稳定时的拟合阶次降低76.7%,可准确识别服役状态下机床旋转部件的模态特征。

关键词: 工作模态分析, 自适应噪声完备集合经验模态分解, 小波阈值分级准则, 倒频谱编辑, 磨齿机, 参数识别

Abstract:

Aiming at the problems that the spindle vibrations in grinding machines were complex and the modal characteristics were difficult to effectively identified under the service states. Based on adaptive noise complete ensemble empirical mode decomposition and correlation analysis, a method was proposed. The finite element modal analysis was used to define the frequency band range, and the wavelet threshold classification method was used to retain the modal feature information. In order to identify and eliminate the harmonic response generated by rotor, a method was used in signal cepstrum editing. Different noise reduction methods and 2-DOF examples show that the modal identification errors are reduced to 1.3% after processing by the proposed method, the fitting order is reduced 76.7% as the poles are stable, and the modal characteristics of the rotating parts are accurately identified for the machine tool in service.

Key words: operational modal analysis, complete ensemble empirical mode decomposition with adaptive noise, wavelet threshold grading criterion, cepstrum editing, grinding machine, parameter identification

中图分类号: