CN109948485B - Rotary machine fault feature extraction method based on vibration signal correlation analysis - Google Patents
Rotary machine fault feature extraction method based on vibration signal correlation analysis Download PDFInfo
- Publication number
- CN109948485B CN109948485B CN201910174733.XA CN201910174733A CN109948485B CN 109948485 B CN109948485 B CN 109948485B CN 201910174733 A CN201910174733 A CN 201910174733A CN 109948485 B CN109948485 B CN 109948485B
- Authority
- CN
- China
- Prior art keywords
- signal
- sub
- vibration
- correlation coefficient
- reference sub
- 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.)
- Active
Links
Images
Abstract
The method comprises the steps of firstly, collecting vibration signals on rotary mechanical equipment, and selecting a section of reference sub-signals from the vibration signals; then, enabling the reference sub-signal to start from the initial time of the vibration signal, translating point by point along the direction of the time axis, continuously covering a sub-signal from the vibration signal, and simultaneously calculating the correlation coefficient between the reference sub-signal and the covering sub-signal; obtaining a correlation coefficient sequence after translation is finished; and finally, extracting fault characteristic frequency from the frequency spectrum of the correlation coefficient sequence. The method is simple and easy to implement, the characteristic of the original vibration signal is self-adaptive in the process of acquiring the correlation coefficient sequence, the change rule of the fault characteristic of the rotary mechanical equipment can be intuitively displayed, and the fault characteristic frequency can be effectively and conveniently extracted from the frequency spectrum of the correlation coefficient sequence and is used for fault diagnosis of the rotary mechanical equipment.
Description
Technical Field
The invention relates to a rotary machine fault feature extraction method based on vibration signal correlation analysis.
Background
Regular impact vibration is a typical feature of failure damage to components within a rotating machine, the frequency of such impact features being directly related to a particular failure state of the rotating machine. However, since the components of the device are generally more, there is some attenuation in the transmission of the fault impact vibration from the fault damaged vibration source to the surface of the device along a complex path. In addition, other moving elements in the device can excite corresponding vibration, and in addition, interference of working environment noise, fault impact characteristics which can be measured by the vibration sensor on the device are usually weak, especially in early damage of the device. Therefore, it is not easy to detect or extract the regular impact characteristics related to faults from the vibration signals of the equipment, and further identify the frequency of the fault characteristics, and the research on the frequency is also an important content in the fault diagnosis research of the rotating machinery equipment.
Common methods for extracting fault impact characteristics from vibration signals of rotary mechanical equipment include wavelet threshold noise reduction, singular value decomposition noise reduction, empirical mode decomposition, stochastic resonance and the like. However, wavelet threshold noise reduction has the defect that a threshold is difficult to select, singular value decomposition noise reduction has the defect that an effective singular value threshold is difficult to determine, empirical mode decomposition has the problems of end-point effect, mode aliasing and the like, and stochastic resonance is difficult to adjust parameters in application, so that the methods have certain limitations in fault feature extraction of rotating mechanical equipment.
Disclosure of Invention
The invention provides a rotary mechanical fault feature extraction method based on vibration signal correlation analysis, which aims to solve the problem that the fault feature extraction of rotary mechanical equipment is difficult in the prior art and overcome the defects of the prior method.
The invention relates to a rotary machine fault feature extraction method based on vibration signal correlation analysis, which comprises the following steps:
(1) Collecting a set of vibration signals y (k) of length L on a rotating machine, wherein k=1, 2, …, L;
(2) At any given time point k=k in the vibration signal y (k) 0 Starting at this point, a section of length L is cut s Reference sub-signal y of (2) s (k s ) Wherein L is s <<L，k s ＝1,2,…,L s And refers to the sub-signal y s (k s )＝y(k d ) Wherein k is d ＝k 0 ,k 0 +1,…,k 0 +L s -1；
(3) Let the reference sub-signal y s (k s ) From an initial time point k=1 of the vibration signal y (k), shifting point by point in a time axis direction in which k increases;
(4) At time point k=k t Where 1.ltoreq.k t ≤L-L s Reference sub-signal y s (k s ) Masking a sub-signal y of equal length in the vibration signal y (k) m (k s ) And mask the sub-signal y m (k s )＝y(k m ) Wherein k is m ＝k t ,k t +1,…,k t +L s -1；
(5) Calculating the reference sub-signal y s (k s ) And mask sub-signal y m (k s ) Correlation coefficient ρ between sm (k t ) The calculation formula is as follows:
wherein the method comprises the steps of
For reference sub-signal y s (k s ) The mean value of (2) is calculated as:
(6) Reference sub-signal y s (k s ) After translation in the vibration signal y (k) to the time point k=l', where L s ≤L′≤L-L s The translation is finished, and a reference sub-signal y is obtained s (k s ) Correlation coefficient sequence ρ between L' mask sub-signals sm (k '), wherein k ' =1, 2, …, L ';
(7) For the correlation coefficient sequence ρ sm And (k') performing Fourier transformation, and extracting the interval frequency of the side band spectral lines from the frequency spectrum, thereby obtaining the fault characteristic frequency.
The invention has the following positive effects: according to the invention, the reference sub-signal is directly selected from the vibration signals sampled on the rotary mechanical equipment, so that the reference sub-signal translates in the sampled vibration signals, and a section of sub-signal is continuously covered in the sampled vibration signals, so that a correlation coefficient sequence of the reference sub-signal and the covered sub-signal is obtained, and the method is simple and easy to implement; the acquisition process of the correlation coefficient sequence is self-adaptive to the characteristics of the original vibration signal, so that useful information can be extracted to the maximum extent; the correlation coefficient sequence curve can intuitively display the change rule of the fault characteristics of the rotary mechanical equipment, can effectively and conveniently extract the fault characteristic frequency and is used for fault diagnosis of the rotary mechanical equipment.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention;
FIG. 2 is a set of time domain waveforms of vibration acceleration signals of a wind gearbox, along with a reference sub-signal, a mask sub-signal at time point k=1, and a mask sub-signal at time point k=30000 taken from the vibration signals, in an embodiment of the invention;
FIG. 3 is a time domain waveform diagram of a reference sub-signal according to an embodiment of the present invention;
FIG. 4 is a graph of a correlation coefficient sequence in an embodiment of the present invention;
FIG. 5 shows the correlation coefficient sequence spectrum in the embodiment of the present invention, the frequency range is 400 Hz-700 Hz.
Detailed Description
The invention is further described below with reference to the drawings and examples.
Referring to fig. 1, the rotary machine fault feature extraction method based on vibration signal correlation analysis includes the steps of:
(1) Collecting a set of vibration signals y (k) of length L on a rotating machine, wherein k=1, 2, …, L;
(2) At any given time point k=k in the vibration signal y (k) 0 Starting at this point, a section of length L is cut s Reference sub-signal y of (2) s (k s ) Wherein L is s <<L，k s ＝1,2,…,L s And refers to the sub-signal y s (k s )＝y(k d ) Wherein k is d ＝k 0 ,k 0 +1,…,k 0 +L s -1；
(3) Let the reference sub-signal y s (k s ) From an initial time point k=1 of the vibration signal y (k), shifting point by point in a time axis direction in which k increases;
(4) At time point k=k t Where 1.ltoreq.k t ≤L-L s Reference sub-signal y s (k s ) Masking a sub-signal y of equal length in the vibration signal y (k) m (k s ) And mask the sub-signal y m (k s )＝y(k m ) Wherein k is m ＝k t ,k t +1,…,k t +L s -1；
(5) Calculating the reference sub-signal y s (k s ) And mask sub-signal y m (k s ) Correlation coefficient ρ between sm (k t ) The calculation formula is as follows:
wherein the method comprises the steps of
For reference sub-signal y s (k s ) The mean value of (2) is calculated as:
(6) Reference sub-signal y s (k s ) After translation in the vibration signal y (k) to the time point k=l', where L s ≤L′≤L-L s The translation is finished, and a reference sub-signal y is obtained s (k s ) Correlation coefficient sequence ρ between L' mask sub-signals sm (k '), wherein k ' =1, 2, …, L ';
(7) For the correlation coefficient sequence ρ sm And (k') performing Fourier transformation, and extracting the interval frequency of the side band spectral lines from the frequency spectrum, thereby obtaining the fault characteristic frequency.
The invention is applied to the fault vibration signal processing of the wind power gear box. The gear tooth breakage fault damage occurs on one pinion of the selected wind power gear box, the number of teeth of the pinion is 32, the rated rotation speed is 1800rpm, the rotation frequency is 30Hz, the fault characteristic frequency is consistent with the rotation frequency, the same is 30Hz, the meshing frequency is 960Hz, the sampling frequency of a vibration signal of the wind power gear box is 97656Hz, and the time domain synchronous average noise reduction is carried out on the vibration signal by the method.
(1) On a wind power gearbox, a set of vibration acceleration signals y (k) of length l=100000 are acquired with a vibration sensor, where k=1, 2, …,100000, as shown in fig. 2.
(2) At any given time point k=72467 in the vibration signal y (k)Starting at this point, a reference sub-signal y of length 12568 is truncated s (k s ) Wherein k is s =1, 2, …,12568, as shown by the solid box-select area in fig. 2; and reference sub-signal y s (k s )＝y(k d ) Wherein k is d = 72467,72468, …,85034; reference sub-signal y s (k s ) The time domain waveform of (2) is shown in figure 3.
(3) Let the reference sub-signal y s (k s ) From an initial time point k=1 of the vibration signal y (k), the shift is made point by point in the time axis direction in which k increases.
(4) At time point k=k t (1≤k t 87432), reference sub-signal y s (k s ) Masking a sub-signal of equal length in the vibration signal y (k), the masking sub-signal being defined as y m (k s )＝y(k m )，k m ＝k t ,k t +1,…,k t +12567. At time point l=1, the mask sub-signal is shown as the left dashed box select area in fig. 2.
(5) Calculating the reference sub-signal y s (k s ) And mask sub-signal y m (k s ) Correlation coefficient ρ between sm (k t ) The calculation formula is as follows:
wherein the method comprises the steps of
And->
Respectively the reference sub-signals y s (k s ) And mask sub-signal y m (k s ) The mean value of (2) is calculated as follows:
(6) Reference sub-signal y s (k s ) After shifting to time point k=30000 in vibration signal y (k), shifting is completed, and the masking sub-signal is shown as the right-side dashed box selection area in fig. 2, and finally the reference sub-signal y is obtained s (k s ) Correlation coefficient sequence rho between 30000 mask sub-signals sm (k '), where k' =1, 2, …,30000, as shown in fig. 4, where very pronounced periodic features occur.
(7) For the correlation coefficient sequence ρ sm (k') performing fast Fourier transform, wherein the frequency spectrum of 400 Hz-700 Hz is shown in figure 5, the interval frequency of the side band spectral lines in the frequency spectrum chart is 29.3Hz, and the interval frequency is basically consistent with the fault characteristic frequency of a pinion in the wind power gear box, which indicates that the pinion is damaged by local faults.
The embodiments described in the present specification are merely examples of implementation forms of the inventive concept, and the scope of protection of the present invention should not be construed as being limited to the specific forms set forth in the embodiments, and the scope of protection of the present invention and equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.
Claims (1)
1. The rotary machine fault feature extraction method based on vibration signal correlation analysis comprises the following steps:
(1) Collecting a set of vibration signals y (k) of length L on a rotating machine, wherein k=1, 2, …, L;
(2) At any given time point k=k in the vibration signal y (k) 0 Starting at this point, a section of length L is cut s Reference sub-signal y of (2) s (k s ) Wherein L is s <<L，k s ＝1,2,…,L s And refers to the sub-signal y s (k s )＝y(k d ) Wherein k is d ＝k 0 ,k 0 +1,…,k 0 +L s -1；
(3) Let the reference sub-signal y s (k s ) From an initial time point k=1 of the vibration signal y (k), shifting point by point in a time axis direction in which k increases;
(4) At time point k=k t Where 1.ltoreq.k t ≤L-L s Reference sub-signal y s (k s ) Masking a sub-signal y of equal length in the vibration signal y (k) m (k s ) And mask the sub-signal y m (k s )＝y(k m ) Wherein k is m ＝k t ,k t +1,…,k t +L s -1；
(5) Calculating the reference sub-signal y s (k s ) And mask sub-signal y m (k s ) Correlation coefficient ρ between sm (k t ) The calculation formula is as follows:
wherein the method comprises the steps of
For reference sub-signal y s (k s ) The mean value of (2) is calculated as:
(6) Reference sub-signal y s (k s ) After translation in the vibration signal y (k) to the time point k=l', where L s ≤L′≤L-L s Flat, flatAfter the shift, a reference sub-signal y is obtained s (k s ) Correlation coefficient sequence ρ between L' mask sub-signals sm (k '), wherein k ' =1, 2, …, L ';
(7) For the correlation coefficient sequence ρ sm And (k') performing Fourier transformation, and extracting the interval frequency of the side band spectral lines from the frequency spectrum, thereby obtaining the fault characteristic frequency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910174733.XA CN109948485B (en) | 2019-03-08 | 2019-03-08 | Rotary machine fault feature extraction method based on vibration signal correlation analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910174733.XA CN109948485B (en) | 2019-03-08 | 2019-03-08 | Rotary machine fault feature extraction method based on vibration signal correlation analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109948485A CN109948485A (en) | 2019-06-28 |
| CN109948485B true CN109948485B (en) | 2023-05-12 |
Family
ID=67009445
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910174733.XA Active CN109948485B (en) | 2019-03-08 | 2019-03-08 | Rotary machine fault feature extraction method based on vibration signal correlation analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109948485B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007033258A2 (en) * | 2005-09-14 | 2007-03-22 | University Of Massachusetts | Multi-scale enveloping spectrogram signal processing for condition monitoring and the like |
| CN105241666A (en) * | 2015-09-21 | 2016-01-13 | 华南理工大学 | Rolling bearing fault feature extraction method based on signal sparse representation theory |
| CN105303181A (en) * | 2015-11-04 | 2016-02-03 | 燕山大学 | Stochastic resonance weak impact feature enhancement extraction method on the basis of sliding window |
| CN108168891A (en) * | 2018-02-26 | 2018-06-15 | 成都昊铭科技有限公司 | The extracting method and equipment of rolling bearing Weak fault signal characteristic |
| CN108444704A (en) * | 2018-03-30 | 2018-08-24 | 华中科技大学 | A kind of Fault Diagnosis of Rolling Element Bearings method |
| CN108875279A (en) * | 2018-07-27 | 2018-11-23 | 中国计量大学 | Bearing sound emission signal characteristic extracting method based on EMD and shape filtering |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10359473B2 (en) * | 2012-05-29 | 2019-07-23 | Nutech Ventures | Detecting faults in turbine generators |
| US10018358B2 (en) * | 2015-09-29 | 2018-07-10 | Siemens Energy, Inc. | Method and system for igniter health monitoring in a gas turbine engine |
-
2019
- 2019-03-08 CN CN201910174733.XA patent/CN109948485B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007033258A2 (en) * | 2005-09-14 | 2007-03-22 | University Of Massachusetts | Multi-scale enveloping spectrogram signal processing for condition monitoring and the like |
| CN105241666A (en) * | 2015-09-21 | 2016-01-13 | 华南理工大学 | Rolling bearing fault feature extraction method based on signal sparse representation theory |
| CN105303181A (en) * | 2015-11-04 | 2016-02-03 | 燕山大学 | Stochastic resonance weak impact feature enhancement extraction method on the basis of sliding window |
| CN108168891A (en) * | 2018-02-26 | 2018-06-15 | 成都昊铭科技有限公司 | The extracting method and equipment of rolling bearing Weak fault signal characteristic |
| CN108444704A (en) * | 2018-03-30 | 2018-08-24 | 华中科技大学 | A kind of Fault Diagnosis of Rolling Element Bearings method |
| CN108875279A (en) * | 2018-07-27 | 2018-11-23 | 中国计量大学 | Bearing sound emission signal characteristic extracting method based on EMD and shape filtering |
Non-Patent Citations (5)
| Title |
|---|
| Chen Chu等.The research of machinery fault feature extraction methods based on vibration signal.ScienceDirect.2018,第51卷(第17期),第346-352页. * |
| 张强等.基于改进EMD和滑动峰态算法的滚棒轴承声发射信号故障特征提取.沈阳航空航天大学学报.2015,第32卷(第2期),第43-47+53页. * |
| 郭艳平 ; 龙涛元 ; .振动信号模型在滚动轴承故障诊断中的应用.机械设计与制造.2018,(第1期),第270-272页. * |
| 郭远晶等.频谱密度函数相似性比较的齿轮箱故障诊断.振动工程学报.2018,第31卷(第1期),第157-164页. * |
| 马兴伟.基于VPMCD的旋转机械故障诊断方法研究.中国优秀硕士学位论文全文数据库工程科技Ⅱ辑.2015,(第3期),第C029-145页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109948485A (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107505135B (en) | Rolling bearing composite fault extraction method and system | |
| Li et al. | Application of Hermitian wavelet to crack fault detection in gearbox | |
| Li et al. | An enhanced rolling bearing fault detection method combining sparse code shrinkage denoising with fast spectral correlation | |
| CN103499445B (en) | A kind of Fault Diagnosis of Roller Bearings based on time-frequency slice analysis | |
| CN109883706B (en) | Method for extracting local damage weak fault features of rolling bearing | |
| Bozchalooi et al. | Teager energy operator for multi-modulation extraction and its application for gearbox fault detection | |
| CN108151869B (en) | Mechanical vibration characteristic index extraction method, system and device | |
| CN111769810B (en) | Fluid mechanical modulation frequency extraction method based on energy kurtosis spectrum | |
| Kocur et al. | Order bispectrum: a new tool for reciprocated machine condition monitoring | |
| Buzzoni et al. | Diagnosis of localized faults in multistage gearboxes: a vibrational approach by means of automatic EMD-based algorithm | |
| CN108398260B (en) | Method for quickly evaluating instantaneous angular speed of gearbox based on mixed probability method | |
| Xin et al. | Extraction of second-order cyclostationary sources by matching instantaneous power spectrum with stochastic model–application to wind turbine gearbox | |
| KR102093929B1 (en) | Apparatus and Method for Diagnosing Mechanical System Health based on CIM | |
| CN109948487B (en) | Rotating machinery fault feature extraction method based on time-frequency spectrum correlation analysis | |
| CN109948485B (en) | Rotary machine fault feature extraction method based on vibration signal correlation analysis | |
| CN112597969B (en) | Rolling bearing fault diagnosis method, system and medium | |
| CN111582248B (en) | SVD-based gearbox signal noise reduction method | |
| CN109948486B (en) | Time domain synchronous average noise reduction method for mechanical vibration signal | |
| CN113074941B (en) | Variable-speed gear fault signal extraction method based on self-adaptive time-varying comb filtering | |
| CN116361733A (en) | Fault diagnosis method, device, system and storage medium | |
| Huang et al. | Short-time kurtogram for bearing fault feature extraction under time-varying speed conditions | |
| El-morsy et al. | Gearbox damage diagnosis using wavelet transform technique | |
| Saidi et al. | The use of spectral kurtosis as a trend parameter for bearing faults diagnosis | |
| Kumar et al. | Denoising of vibration signal used in fault diagnosis of machine | |
| CN102798462B (en) | No-time-mark order tracking method based on self-demodulation transform |
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 |