• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2017, Vol. 53 ›› Issue (9): 83-91.doi: 10.3901/JME.2017.09.083

• • 上一篇    下一篇

基于故障特征趋势线模板的滚动轴承故障诊断*

刘东东, 程卫东, 万广通   

  1. 北京交通大学机械与电子控制工程学院 北京 100044
  • 出版日期:2017-05-05 发布日期:2017-05-05
  • 作者简介:

    刘东东,男,1990年出生,主要研究方向为机械故障诊断和信号处理。

    E-mail:14121237@bjtu.edu.cn

    程卫东(通讯作者),男,1967年出生,博士,教授,博士研究生导师。主要研究方向为制造装备智能测控与故障诊断。

    E-mail: wdcheng@bjtu.edu.cn

  • 基金资助:
    * 国家自然科学基金资助项目(51275030); 20161008收到初稿,20170220收到修改稿;

Bearing Fault Diagnosis Based on Fault Characteristic Trend Template

LIU Dongdong, CHENG Weidong, WAN Guangtong   

  1. School of Mechanical Electronic and Control Engineering, Beijing Jiaotong University, Beijing 100044
  • Online:2017-05-05 Published:2017-05-05

摘要:

齿轮噪声和变转速工况干扰下的滚动轴承故障诊断存在两个问题,一是转速波动限制了齿轮噪声滤除算法的使用,二是常用于克服变转速条件的阶比跟踪技术存在计算效率低及包络畸变等问题。为避免这两个问题,提出基于故障特征趋势线模板的滚动轴承故障特征提取方法。使用线调频小波路径追踪算法分别在降采样的时域信号和经Hilbert变换得到包络信号中提取齿轮峰值啮合倍频趋势线和瞬时故障特征频率趋势线;计算齿轮峰值啮合倍频趋势线与瞬时故障特征频率趋势线对应时间点的比值,并连接各个时间点的比值得到故障特征趋势线;将故障特征趋势线与理论计算的故障趋势线模板进行匹配,观察匹配结果完成故障诊断。该算法的创新点是将较难提取的包含轴承故障信息频带的获取直接转换为瞬时频率趋势线的 提取,定义了故障特征趋势线的概念,根据故障特征趋势线寻找轴承故障特征。仿真算例和应用实例证明了该方法的有 效性。

关键词: 变转速, 峰值啮合倍频, 故障趋势线模板, 瞬时故障特征频率, 轴承故障诊断, 齿轮噪声

Abstract:

:The removal of gear vibration noise is limited because of the rotational speed fluctuation and there are some defects in order tracking commonly used in fault diagnosis of rolling bearing under a variable rotational speed, so the method of rolling bearing fault diagnosis based on fault characteristic trend template is proposed. The instantaneous dominant meshing multiply (IDMM) trend is extracted in down sampled time domain signal and the instantaneous fault characteristic frequency (IFCF) is extracted in envelope signal obtained by Hilbert transform of down sampled signal by using the Chirplet path pursuit. The fault characteristic trend is obtained by connecting the ratio points of corresponding IDMM and IFCF. The fault diagnosis is completed by matching the fault characteristic trend and fault characteristic trend template. The innovative method does not need to obtain the frequency band containing fault information which is presently a hard process. The paper defined the concept of fault characteristic trend, and tried to find the bearing fault feature in the fault characteristic trend. The effectiveness of the proposed method has been proved by both simulated and experimental bearing vibration signals.

Key words: bearing fault diagnosis, fault characteristic trend template, IDMM, IFCF, variable rotational speed, gear noise