CN115219199A - 一种基于深度相关熵谱密度的轴承微弱故障提取方法 - Google Patents

一种基于深度相关熵谱密度的轴承微弱故障提取方法 Download PDF

Info

Publication number
CN115219199A
CN115219199A CN202210834504.8A CN202210834504A CN115219199A CN 115219199 A CN115219199 A CN 115219199A CN 202210834504 A CN202210834504 A CN 202210834504A CN 115219199 A CN115219199 A CN 115219199A
Authority
CN
China
Prior art keywords
depth
entropy
spectral density
bearing
signal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202210834504.8A
Other languages
English (en)
Other versions
CN115219199B (zh
Inventor
李辉
邓三鹏
张春林
王震生
周明龙
戴琨
周旺发
祁宇明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
TANGSHAN INDUSTRIAL VOCATIONAL TECHNICAL COLLEGE
Tianjin Bonuo Intelligent Creative Robotics Technology Co ltd
Tianjin University of Technology
Anhui Technical College of Mechanical and Electrical Engineering
Original Assignee
TANGSHAN INDUSTRIAL VOCATIONAL TECHNICAL COLLEGE
Tianjin Bonuo Intelligent Creative Robotics Technology Co ltd
Tianjin University of Technology
Anhui Technical College of Mechanical and Electrical Engineering
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by TANGSHAN INDUSTRIAL VOCATIONAL TECHNICAL COLLEGE, Tianjin Bonuo Intelligent Creative Robotics Technology Co ltd, Tianjin University of Technology, Anhui Technical College of Mechanical and Electrical Engineering filed Critical TANGSHAN INDUSTRIAL VOCATIONAL TECHNICAL COLLEGE
Priority to CN202210834504.8A priority Critical patent/CN115219199B/zh
Publication of CN115219199A publication Critical patent/CN115219199A/zh
Application granted granted Critical
Publication of CN115219199B publication Critical patent/CN115219199B/zh
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01MTESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00Testing of machine parts
    • G01M13/04Bearings
    • G01M13/045Acoustic or vibration analysis

Landscapes

  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • General Physics & Mathematics (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

本发明提供了一种基于深度相关熵谱密度的轴承微弱故障提取方法,其特征在于所述方法包括以下步骤:采集振动信号
Figure DDA0003747060350000011
计算信号
Figure DDA0003747060350000012
的深度相关熵Vx(n),计算深度相关熵Vx(n)的功率谱密度Px(f),画出深度相关熵Vx(n)的功率谱密度Px(f)图,由频谱尖峰可识别轴承的故障特征信息;本发明对轴承早期微弱故障特征具有增强作用,对干扰噪声具有很强的抑制作用,显著提高了信噪比。

Description

一种基于深度相关熵谱密度的轴承微弱故障提取方法
技术领域
本发明涉及现代信号处理技术领域,尤其涉及一种基于深度相关熵谱密度的轴承微弱故障提取方法。
背景技术
功率谱密度是一种常用的信号处理方法,传统功率谱密度基于信号二阶统计量,只能用于处理平稳信号,当信号不满足平稳条件或信号中含有非高斯噪声时,传统功率谱密度的性能会减退,甚至失效;旋转机械轴承、齿轮等零部件的早期故障振动信号,普遍存在调制现象,早期故障特征往往被强背景噪声和设备固有振动信号淹没,使得测量的振动信号具有较低的信噪比,加大了轴承微弱故障特征的提取难度,直接影响轴承故障诊断的准确性和可靠性。
发明内容
根据以上技术问题,本发明提供一种基于深度相关熵谱密度的轴承微弱故障提取方法。
本发明提供一种基于深度相关熵谱密度的轴承微弱故障提取方法,其具体步骤如下:
步骤1,采集振动信号
Figure BDA0003747060330000011
采样点数是N,采样频率是fs,信号
Figure BDA0003747060330000012
是1×N的行向量;
步骤2,计算信号
Figure BDA0003747060330000013
的深度相关熵Vx(n)(Vx(n)是1×N的行向量),
Figure BDA0003747060330000014
其中:j=0,1,2,…,N-1,n=0,1,2,…,N-1,E(·)是期望均值算子,U(·)是信号x(i)的相关熵,
Figure BDA0003747060330000015
Figure BDA0003747060330000021
Figure BDA0003747060330000022
是核长等于σ1的核函数,
Figure BDA0003747060330000023
Figure BDA0003747060330000024
其中:(·)×(·)是数乘运算,
Figure BDA0003747060330000025
是核长等于σ2的核函数,||·||是范数算子,e(·)是自然指数函数;
步骤3,计算深度相关熵Vx(n)的功率谱密度Px(f),
Figure BDA0003747060330000026
Px(f)是1×N的行向量,其中:j是虚数单位,f是频率,
Figure BDA0003747060330000027
单位Hz;
步骤4,画出深度相关熵的功率谱密度Px(f)图,由频谱尖峰可识别轴承故障特征频率。
本发明的有益效果为:
1.本发明提出的深度相关熵的功率谱密度具有微弱故障特征增强功能,能凸显轴承微弱故障特征,能有效解决轴承早期微弱故障特征难以提取的问题。
2.本发明提出的深度相关熵同时反映了信号的时间特征和统计特征,既包含了信号的二阶统计量信息,又包含了信号的高阶统计量信息,能有效提取淹没在噪声背景中的轴承微弱故障特征,因而基于深度相关熵的功率谱密度,对噪声具有较好的抑制作用,显著提高了信噪比。
附图说明
图1为本发明所述方法的流程图;
图2为实施例2轴承外圈故障振动信号
Figure BDA0003747060330000031
的时域波形;
图3为实施例2轴承外圈故障振动信号
Figure BDA0003747060330000032
的快速傅里叶变换图;
图4为实施例2轴承外圈故障振动信号
Figure BDA0003747060330000033
当核长σ1=0.17、σ2=0.0019时深度相关熵Vx(n)图;
图5为实施例2轴承外圈故障振动信号
Figure BDA0003747060330000034
当核长σ1=0.17、σ2=0.0019时深度相关熵Vx(n)的功率谱密度Px(f)图;
图6为对比例1轴承外圈故障振动信号
Figure BDA0003747060330000035
的传统功率谱密度图;
图7为对比例2轴承外圈故障振动信号
Figure BDA0003747060330000036
的传统包络谱图。
具体实施方式
下面将结合本发明的附图,对本发明的技术方案进行清楚完整地描述。
实施例1
如图1所示,本发明公开了一种基于深度相关熵谱密度的轴承微弱故障提取方法,包括如下步骤:
步骤S1,采集振动信号
Figure BDA0003747060330000037
如图2所示,采样频率是fs,采样点数N,信号
Figure BDA0003747060330000038
是1×N的行向量;
步骤S2,计算信号
Figure BDA0003747060330000039
的深度相关熵Vx(n)(Vx(n)是1×N的行向量),
Figure BDA00037470603300000310
其中:j=0,1,2,…,N-1,n=0,1,2,…,N-1,E(·)是期望均值算子,U(·)是信号x(i)的相关熵,
Figure BDA00037470603300000311
Figure BDA00037470603300000312
Figure BDA00037470603300000313
是核长等于σ1的核函
数,
Figure BDA0003747060330000041
Figure BDA0003747060330000042
其中:(·)×(·)是数乘运算,
Figure BDA0003747060330000043
是核长等于σ2的核函数,||·||是范数算子,e( · )是自然指数函数;
步骤S3,计算深度相关熵Vx(n)的功率谱密度Px(f),
Figure BDA0003747060330000044
Px(f)是1×N的行向量,其中:j是虚数单位,f是频率,
Figure BDA0003747060330000045
单位Hz。
步骤S4,画出深度相关熵的功率谱密度Px(f)图,由频谱尖峰可识别轴承故障特征频率。
实施例2
本实施例是对实施例1给出方法的验证,本实施例采集的轴承外圈故障振动信号为
Figure BDA0003747060330000049
采样频率fs=12kHz,采样点数N=2048,采样时间T=0.17s。本实施例轴承型号为球轴承6205,轴的转动频率fr=29.95Hz,轴承的几何尺寸为:大径D=52.0mm;滚珠直径d=7.94mm;滚珠数量z=9;压力角α=0°。通过计算得到轴承外圈故障特征频率fouter=107.4Hz。
实施例2的振动信号
Figure BDA0003747060330000046
的时域图形,如图2所示;振动信号
Figure BDA0003747060330000047
的快速傅里叶变换,如图3所示,根据实施例1的步骤S2,计算信号
Figure BDA0003747060330000048
的深度相关熵Vx(n),核长σ1=0.17、σ2=0.0019时的深度相关熵Vx(n),如图4所示。根据实施例1的步骤S3,计算深度相关熵Vx(n)的功率谱密度Px(f),如图5所示;从图5可以看出,在谱图的低频段,在轴承外圈故障特征频率fouter=107.4Hz及其二倍频2fouter、三倍频3fouter处存在明显的谱峰,刻画了轴承外圈故障特征信息。
对比例1
为对比深度相关熵功率谱密度Px(f)的轴承外圈故障诊断效果,本对比例采用传统功率谱密度方法,对实施例2中的轴承外圈故障振动信号
Figure BDA0003747060330000051
(图2)进行了分析,图6为传统功率谱密度图,在传统功率谱密度图的低频段,不存在轴承外圈故障特征信息,由于传统功率谱密度方法基于信号二阶统计量且易受噪声干扰的影响,基于传统功率谱密度方法难以识别轴承外圈早期微弱故障特征。
对比例2
为进一步对比深度相关熵功率谱密度Px(f)的轴承外圈故障诊断效果,本对比例利用传统包络谱方法,对实施例2中的轴承外圈故障振动信号
Figure BDA0003747060330000052
(图2)进行了分析,图7为基于传统共振解调技术的包络谱图,根据图3选择信号带通滤波区间为[3000,4000]Hz,对信号
Figure BDA0003747060330000053
进行带通滤波后,计算其包络谱。与对比例1类似,由于受噪声干扰的影响,虽然在轴承外圈故障特征频率fouter=107.4Hz及其二倍频2fouter、三倍频3fouter处存在谱峰,但信噪比很低,容易得出错误的诊断结果。

Claims (3)

1.一种基于深度相关熵谱密度的轴承微弱故障提取方法,其特征在于包括以下步骤:
步骤1,采集振动信号
Figure FDA0003747060320000011
采样点数是N,采样频率是fs,信号
Figure FDA0003747060320000012
是1×N的行向量;
步骤2,计算信号
Figure FDA0003747060320000013
的深度相关熵Vx(n)(Vx(n)是1×N的行向量),
Figure FDA0003747060320000014
其中:j=0,1,2,…,N-1,,n=0,1,2,…,N-1,E(·)是期望均值算子,U(·)是信号x(i)的相关熵,
Figure FDA0003747060320000015
i=0,1,2,…,N-1,
Figure FDA0003747060320000016
是核长等于σ1的核函数,
Figure FDA0003747060320000017
其中:(·)×(·)是数乘运算,
Figure FDA0003747060320000018
是核长等于σ2的核函数,||·||是范数算子,e(·)是自然指数函数;
步骤3,计算深度相关熵Vx(n)的功率谱密度Px(f),
Figure FDA0003747060320000019
Px(f)是1×N的行向量,其中:j是虚数单位,f是频率,
Figure FDA00037470603200000110
单位Hz;
步骤4,画出深度相关熵的功率谱密度Px(f)图,由频谱尖峰可识别轴承故障特征频率。
2.如权利要求1所述的一种基于深度相关熵谱密度的轴承微弱故障提取方法,其特征在于所述步骤2中,信号
Figure FDA00037470603200000111
的深度相关熵
Figure FDA00037470603200000112
Figure FDA00037470603200000113
其中:
Figure FDA00037470603200000114
Figure FDA0003747060320000021
分别是核长等于σ1、σ2的核函数。
3.如权利要求1所述的一种基于深度相关熵谱密度的轴承微弱故障提取方法,其特征在于所述步骤2中,核函数
Figure FDA0003747060320000022
CN202210834504.8A 2022-07-14 2022-07-14 一种基于深度相关熵谱密度的轴承微弱故障提取方法 Active CN115219199B (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210834504.8A CN115219199B (zh) 2022-07-14 2022-07-14 一种基于深度相关熵谱密度的轴承微弱故障提取方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210834504.8A CN115219199B (zh) 2022-07-14 2022-07-14 一种基于深度相关熵谱密度的轴承微弱故障提取方法

Publications (2)

Publication Number Publication Date
CN115219199A true CN115219199A (zh) 2022-10-21
CN115219199B CN115219199B (zh) 2024-07-16

Family

ID=83612093

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210834504.8A Active CN115219199B (zh) 2022-07-14 2022-07-14 一种基于深度相关熵谱密度的轴承微弱故障提取方法

Country Status (1)

Country Link
CN (1) CN115219199B (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190058428A1 (en) * 2017-08-17 2019-02-21 Alstom Transport Technologies Method for determining a state of a bearing, module for determining a state of a bearing, railway vehicle and system
CN110059564A (zh) * 2019-03-19 2019-07-26 西安交通大学 基于功率谱密度和互相关熵谱密度融合的特征提取方法
US20200200648A1 (en) * 2018-02-12 2020-06-25 Dalian University Of Technology Method for Fault Diagnosis of an Aero-engine Rolling Bearing Based on Random Forest of Power Spectrum Entropy
CN111898086A (zh) * 2020-03-03 2020-11-06 天津职业技术师范大学(中国职业培训指导教师进修中心) 基于时域平滑周期图的循环平稳相关熵谱密度计算方法
AU2020103681A4 (en) * 2020-11-26 2021-02-04 Anhui University Of Technology Rolling Bearing Fault Diagnosis Method Based on Fourier Decomposition and Multi-scale Arrangement Entropy Partial Mean Value
CN113869358A (zh) * 2021-08-17 2021-12-31 天津职业技术师范大学(中国职业培训指导教师进修中心) 基于循环相关熵和一维浅卷积神经网络的轴承故障诊断方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190058428A1 (en) * 2017-08-17 2019-02-21 Alstom Transport Technologies Method for determining a state of a bearing, module for determining a state of a bearing, railway vehicle and system
US20200200648A1 (en) * 2018-02-12 2020-06-25 Dalian University Of Technology Method for Fault Diagnosis of an Aero-engine Rolling Bearing Based on Random Forest of Power Spectrum Entropy
CN110059564A (zh) * 2019-03-19 2019-07-26 西安交通大学 基于功率谱密度和互相关熵谱密度融合的特征提取方法
CN111898086A (zh) * 2020-03-03 2020-11-06 天津职业技术师范大学(中国职业培训指导教师进修中心) 基于时域平滑周期图的循环平稳相关熵谱密度计算方法
AU2020103681A4 (en) * 2020-11-26 2021-02-04 Anhui University Of Technology Rolling Bearing Fault Diagnosis Method Based on Fourier Decomposition and Multi-scale Arrangement Entropy Partial Mean Value
CN113869358A (zh) * 2021-08-17 2021-12-31 天津职业技术师范大学(中国职业培训指导教师进修中心) 基于循环相关熵和一维浅卷积神经网络的轴承故障诊断方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
李辉: "相关熵的降噪机理及其在轴承故障诊断中的应用", 《轴承》, no. 3, 31 December 2021 (2021-12-31), pages 36 - 44 *

Also Published As

Publication number Publication date
CN115219199B (zh) 2024-07-16

Similar Documents

Publication Publication Date Title
McCormick et al. Cyclostationarity in rotating machine vibrations
CN102840907B (zh) 早期故障状态下滚动轴承振动信号特征提取和分析方法
CN109635334A (zh) 基于粒子群优化的滚动轴承故障诊断方法、系统及介质
CN109668733B (zh) 变分非线性模式分解变转速轴承故障诊断方法
CN109883706B (zh) 一种滚动轴承局部损伤微弱故障特征提取方法
CN108388908B (zh) 基于k-svd和滑窗降噪的滚动轴承冲击故障诊断方法
CN106908241A (zh) 一种基于lmd与小波去噪相结合的轴承故障判别方法
CN111238812B (zh) 一种滚动轴承保持架故障的自诊断方法
CN104502099A (zh) 齿轮箱瞬变工况特征分量循频提取方法
CN110647871A (zh) 一种基于时域特量增强的滚动轴承故障诊断方法及系统
Zhao et al. Rolling element bearing instantaneous rotational frequency estimation based on EMD soft-thresholding denoising and instantaneous fault characteristic frequency
CN113607415A (zh) 一种变转速下基于短时随机共振的轴承故障诊断方法
CN116361733A (zh) 一种故障诊断方法、装置、系统以及存储介质
Bastami et al. Estimating the size of naturally generated defects in the outer ring and roller of a tapered roller bearing based on autoregressive model combined with envelope analysis and discrete wavelet transform
CN114486263A (zh) 一种旋转机械滚动轴承振动信号降噪解调方法
Hu et al. Frequency loss and recovery in rolling bearing fault detection
CN115219199A (zh) 一种基于深度相关熵谱密度的轴承微弱故障提取方法
CN114897016A (zh) 一种基于多源频谱特征的风机轴承故障智能诊断方法
CN114781466B (zh) 基于旋转机械振动信号谐波基频的故障诊断方法及系统
CN110907174B (zh) 基于信号冲击结构特性增强的旋转机械微弱故障诊断方法
CN115014765B (zh) 一种通过声信号提取滚动轴承保持架故障特征的方法
CN115931353A (zh) 一种基于相关熵和短时傅里叶变换的轴承故障诊断方法
CN116383629A (zh) 一种变转速滚动轴承故障诊断方法
CN112345247A (zh) 一种滚动轴承的故障诊断方法及装置
CN112747925B (zh) 一种基于复合形态学滤波的滚动轴承故障诊断方法

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant