中国机械工程 ›› 2025, Vol. 36 ›› Issue (9): 2022-2031.DOI: 10.3969/j.issn.1004-132X.2025.09.014

• 智能制造 • 上一篇    

基于共振解调新方法的滚动轴承故障诊断

冯思茜1(), 王家序1,2, 张新1,3(), 黄欣玥1   

  1. 1.西南交通大学机械工程学院, 成都, 610031
    2.重庆大学高端装备机械传动全国重点实验室, 重庆, 400044
    3.西南交通大学轨道交通运维技术与装备四川省重点实验室, 成都, 610031
  • 收稿日期:2024-05-17 出版日期:2025-09-25 发布日期:2025-10-15
  • 通讯作者: 张新
  • 作者简介:冯思茜,女,2000年生,硕士研究生。研究方向为故障诊断。E-mail:siqianfeng_lu@163.com
    张 新*(通信作者),男,1989年生,博士、副教授、博士研究生导师。研究方向为高性能机电传动系统智能测控。发表论文40余篇。E-mail:xylon.zhang@swjtu.edu.cn
  • 基金资助:
    国家自然科学基金(52075456);国家自然科学基金(52175122);四川省自然科学基金(2023NSFSC0362);中国博士后科学基金(2023M732917);四川省博士后创新人才支持项目(BX202214)

Rolling Bearing Fault Diagnosis Using a New Resonance Demodulation Method

Siqian FENG1(), Jiaxu WANG1,2, Xin ZHANG1,3(), Xinyue HUANG1   

  1. 1.School of Mechanical Engineering,Southwest Jiaotong University,Chengdu,610031
    2.State Key Laboratory of Mechanical Transmission for Advanced Equipment,Chongqing University,Chongqing,400044
    3.Technology and Equipment of Rail Transit Operation and Maintenance Key Laboratory of Sichuan Province,Southwest Jiaotong University,Chengdu,610031
  • Received:2024-05-17 Online:2025-09-25 Published:2025-10-15
  • Contact: Xin ZHANG

摘要:

为实现滚动轴承微弱特征提取与故障诊断,提出了一种基于子带重构重排-双树复小波包变换(SRR-DTCWPT)与峰值频率提取的共振解调新方法。基于SRR-DTCWPT的频带划分方法较为精细,并且在保持DTCWPT近似平移不变性和谱能量泄漏少的优点的同时解决了频带错乱的问题。基于SRR-DTCWPT与峰值频率提取的共振解调方法不需要任何指标参与,能提取任意位置的频带,避免了强冲击干扰的影响,且计算过程自动化。将所提方法与Fast Kurtogram和Autogram算法进行比较,验证了该方法在滚动轴承故障诊断中的有效性与高效性。

关键词: 轴承故障诊断, 共振解调, 双树复小波包变换, 子带重构重排, 峰值频率提取

Abstract:

To extract weak features and diagnose rolling bearing faults, a new resonance demodulation method was proposed based on subbands-reconstructed and-rearranged dual-tree complex wavelet packet transform(SRR-DTCWPT) and peak frequency extraction. The SRR-DTCWPT-based frequency band division method was fine, maintained the advantages of DTCWPT in approximate translation invariance and less spectral energy leakage, and solved the problem of band misalignment. The resonance demodulation method based on SRR-DTCWPT and peak frequency extraction did not require the participation of any indicator, might extract frequency bands at any position, avoided the influences of strong impact interference, and automated the calculation processes. The proposed method was compared with Fast Kurtogram and Autogram in simulation and case studies, and the results demonstrate the effectiveness and efficiency of the proposed method.

Key words: bearing fault diagnosis, resonance demodulation, dual-tree complex wavelet packet transform(DTCWPT), subbands reconstruction and rearrangement(SRR), peak frequency extraction

中图分类号: