CN110146152A - Aero-engine vibrates source separation method - Google Patents
Aero-engine vibrates source separation method Download PDFInfo
- Publication number
- CN110146152A CN110146152A CN201910519782.2A CN201910519782A CN110146152A CN 110146152 A CN110146152 A CN 110146152A CN 201910519782 A CN201910519782 A CN 201910519782A CN 110146152 A CN110146152 A CN 110146152A
- Authority
- CN
- China
- Prior art keywords
- formula
- signal
- vibration source
- source signal
- frequency
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H1/00—Measuring characteristics of vibrations in solids by using direct conduction to the detector
- G01H1/003—Measuring characteristics of vibrations in solids by using direct conduction to the detector of rotating machines
Abstract
The invention discloses a kind of aero-engines to vibrate source separation method, first decomposes the observation signal that aero-engine difference is observed on channel by continuous wavelet transform, is analyzed according to spectrum peak and determine primary oscillation source and corresponding fundamental frequency；Then by the steady method of time synchronization, the harmonic wave and subharmonic ingredient of each vibration source is extracted from each observation channel respectively, has extracted source signal from each observation channel；For same source signal, a basic image can be extracted from each observation channel, finally by two norms for comparing each image signal, determines the optimal estimation of each source signal.This method can estimate the number of the primary oscillation source of engine interior in the case where having certain priori knowledge to engine speed, and can extract the main component of source signal.
Description
Technical field
The present invention relates to fault diagnosis fields, specifically provide a kind of aero-engine vibration source separation method.
Background technique
When carrying out status monitoring to aero-engine, sensor is typically located on the casing of engine, this results in passing
The vibration signal that sensor measures is usually the mixed signal of different components generate in machinery vibration source and noise, if directly right
This observation signal, which carries out diagnosis, can have biggish error, even result in the result of mistake.Existing algorithm is based on source mostly
This basic assumption of independence is difficult effectively to estimate each source letter when separating more complex rotating machinery signal
Number, because there may be couplings between source signal each in this complicated machinery, and there are also very noisy interference etc. to influence.
Therefore, a kind of new aero-engine vibration source separation method is researched and developed, by the analysis to mixed signal, with accurate
Ground estimates primary oscillation source and the corresponding fundamental component of each source signal, and extracts the harmonic wave in each source and some time humorous
Wave component becomes people's urgent problem to be solved.
Summary of the invention
In consideration of it, the purpose of the present invention is to provide a kind of aero-engines to vibrate source separation method, to solve existing skill
The problem of each source signal is effectively estimated out is difficult in art.
Present invention provide the technical scheme that a kind of aero-engine vibrates source separation method, include the following steps:
S1: arranging multiple vibrating sensors in engine environment, for detecting the vibration signal of engine, and establishes each
The observation signal model of vibrating sensor, wherein each vibrating sensor represents a channel, and the observation signal model is such as
Shown in formula (1)
In formula, xi(t) i-th of vibrating sensor is indicated in the observation signal of moment t, N indicates the number of vibration source signal,
M indicates the number of vibrating sensor, sj(t) j-th of vibration source signal in moment t, a are indicatedijIndicate vibration source signal sj(t)
In observation signal xi(t) shared weight in；
S2: wavelet scale range [L is determined using formula (2)*, L]
In formula: finterestTo vibrate source signal fundamental frequency, L*For the lower limit value and L of wavelet scale*=1, L are wavelet scale
Upper limit value,Pseudo frequency when for wavelet scale being a, fcFor the centre frequency of the wavelet basis function of selection, FsFor vibration
The sample frequency of dynamic source signal, wherein pseudo frequency must cover all vibration source signal fundamental frequencies, i.e.,
S3: the observation signal that each vibrating sensor obtains decomposed according to formula (3) determined by S2 it is different small
On wave scale, wavelet coefficient W is obtainedi(a,b)
In formula: L is the upper limit value of wavelet scale；Indicate the complex conjugate of wavelet basis function ω ()；
Wherein, shown in wavelet basis function such as formula (4) used in formula (3):
In formula, a is scale coefficient；B is time delay coefficient；Coefficient 1/ | a |1/2It can guarantee that energy is constant；
S4: the main frequency number in the observation signal of each vibrating sensor is estimated according to formula (5)
In formula:For the main frequency in j-th of channel that i-th of sensor detects；PiTo be estimated by i-th of sensor
The vibration source number counted out；
S5: estimating the main frequency number P of tested aero-engine according to formula (6), and determines that each main frequency is corresponding
Fundamental frequency
In formula: fkFor the corresponding main frequency of vibration source signal each in system；M is the number of vibrating sensor；PiIt is
The vibration source number that i observation channel estimates；The main of source signal is vibrated for j-th that i-th of observation channel estimates
Frequency, n are positive integer；
S6: vibration source signal is divided into S sections, it is the Related Component of the signal of T with extracting cycle that each section, which sets Q point,；
According to formula (7), vibration source signal is extracted from each vibrating sensor
In formula: xi(n Δ t) is the discrete form of the observation signal on i-th of channel, Δ t=1/Fs, FsFor vibration source letter
Number sample frequency；
S7: the optimal estimation of each vibration source signal is determined according to formula (8)
In formula: Si,2Indicate two norms of i-th of vibration source signal；Sij,2Indicate extract from j-th of sensor i-th
Two norms of a vibration source signal；P is main frequency number；M is the number of vibrating sensor.
Aero-engine provided by the invention vibrates source separation method, firstly, estimating source in system according to wavelet transformation
The number of signal and its corresponding fundamental frequency；Secondly, extracting each source signal harmonic wave and subharmonic ingredient according to the steady method of time synchronization；
Finally, determining the optimal estimation of each source signal according to two norms.
Aero-engine provided by the invention vibrates source separation method, can accurately estimate primary oscillation source and each
The corresponding fundamental component of source signal, and extract the harmonic wave in each source signal and some subharmonic ingredients.
Specific embodiment
The present invention is further explained below in conjunction with specific embodiment, but the not limitation present invention.
The present invention provides a kind of aero-engines to vibrate source separation method, includes the following steps:
S1: arranging multiple vibrating sensors in engine environment, for detecting the vibration signal of engine, and establishes each
The observation signal model of vibrating sensor, wherein each vibrating sensor represents a channel, and the observation signal model is such as
Shown in formula (1)
In formula, xi(t) i-th of vibrating sensor is indicated in the observation signal of moment t, N indicates the number of vibration source signal,
M indicates the number of vibrating sensor, sj(t) j-th of vibration source signal in moment t, a are indicatedijIndicate vibration source signal sj(t)
In observation signal xi(t) shared weight in；
S2: wavelet scale range [L is determined using formula (2)*, L]
In formula: finterestTo vibrate source signal fundamental frequency, L*For the lower limit value and L of wavelet scale*=1, L are wavelet scale
Upper limit value,Pseudo frequency when for wavelet scale being a, fcFor the centre frequency of the wavelet basis function of selection, FsFor vibration
The sample frequency of dynamic source signal, wherein pseudo frequency must cover all vibration source signal fundamental frequencies, i.e.,
S3: the observation signal that each vibrating sensor obtains decomposed according to formula (3) determined by S2 it is different small
On wave scale, wavelet coefficient W is obtainedi(a,b)
In formula: L is the upper limit value of wavelet scale；Indicate the complex conjugate of wavelet basis function ω ()；
Wherein, shown in wavelet basis function such as formula (4) used in formula (3):
In formula, a is scale coefficient；B is time delay coefficient；Coefficient 1/ | a |1/2It can guarantee that energy is constant；
S4: the main frequency number in the observation signal of each vibrating sensor is estimated according to formula (5)
In formula:For the main frequency in j-th of channel that i-th of sensor detects；PiTo be estimated by i-th of sensor
The vibration source number counted out；
S5: estimating the main frequency number P of tested aero-engine according to formula (6), and determines that each main frequency is corresponding
Fundamental frequency
In formula: fkFor the corresponding main frequency of vibration source signal each in system；M is the number of vibrating sensor；PiIt is
The vibration source number that i observation channel estimates；The main of source signal is vibrated for j-th that i-th of observation channel estimates
Frequency, n are positive integer；
S6: vibration source signal is divided into S sections, it is the Related Component of the signal of T with extracting cycle that each section, which sets Q point,；
According to formula (7), vibration source signal is extracted from each vibrating sensor
In formula: xi(n Δ t) is the discrete form of the observation signal on i-th of channel, Δ t=1/Fs, FsFor vibration source letter
Number sample frequency；
S7: the optimal estimation of each vibration source signal is determined according to formula (8)
In formula: Si,2Indicate two norms of i-th of vibration source signal；Sij,2Indicate extract from j-th of sensor i-th
Two norms of a vibration source signal；P is main frequency number；M is the number of vibrating sensor.
The aero-engine vibrates source separation method, based on the rotating machineries shaft fault vibration signal such as aero-engine
Frequency spectrum characteristic, the vibrational spectra of fault-signal generally comprise fundamental frequency, harmonic components and subharmonic ingredient, such as misalign, touch mill, split
Line etc..The main method that the application uses is continuous wavelet transform and the steady method of time synchronization, key step are as follows: passes through company first
Continuous wavelet transformation decomposes the observation signal that aero-engine difference is observed on channel, is analyzed according to spectrum peak and determines principal vibration
Source and corresponding fundamental frequency；Then by the steady method of time synchronization, the humorous of each vibration source is extracted from each observation channel respectively
Wave and subharmonic ingredient have extracted source signal from each observation channel；For same source signal, from each observation channel
A basic image can be extracted, finally by two norms for comparing each image signal, determines the optimal of each source signal
Estimation.
Embodiments of the present invention are elaborated above, but present invention is not limited to the embodiments described above,
Those of ordinary skill in the art within the scope of knowledge, can also make various without departing from the purpose of the present invention
Variation.
Claims (1)
1. aero-engine vibrates source separation method, which comprises the steps of:
S1: multiple vibrating sensors are arranged in engine environment, for detecting the vibration signal of engine, and establish each vibration
The observation signal model of sensor, wherein each vibrating sensor represents a channel, the observation signal model such as formula
(1) shown in
In formula, xi(t) i-th of vibrating sensor is indicated in the observation signal of moment t, and N indicates the number of vibration source signal, and M is indicated
The number of vibrating sensor, sj(t) j-th of vibration source signal in moment t, a are indicatedijIndicate vibration source signal sj(t) it is seeing
Survey signal xi(t) shared weight in；
S2: wavelet scale range [L is determined using formula (2)*, L]
In formula: finterestTo vibrate source signal fundamental frequency, L*For the lower limit value and L of wavelet scale*=1, L are the upper limit of wavelet scale
Value,Pseudo frequency when for wavelet scale being a, fcFor the centre frequency of the wavelet basis function of selection, FsFor vibration source
The sample frequency of signal, wherein pseudo frequency must cover all vibration source signal fundamental frequencies, i.e.,
S3: the observation signal that each vibrating sensor obtains is decomposed to the different small echo rulers determined by S2 according to formula (3)
On degree, wavelet coefficient W is obtainedi(a,b)
In formula: L is the upper limit value of wavelet scale；Indicate the complex conjugate of wavelet basis function ω ()；
Wherein, shown in wavelet basis function such as formula (4) used in formula (3):
In formula, a is scale coefficient；B is time delay coefficient；Coefficient 1/ | a |1/2It can guarantee that energy is constant；
S4: the main frequency number in the observation signal of each vibrating sensor is estimated according to formula (5)
In formula:For the main frequency in j-th of channel that i-th of sensor detects；PiIt is estimated by i-th of sensor
Vibration source number；
S5: estimating the main frequency number P of tested aero-engine according to formula (6), and determines the corresponding base of each main frequency
Frequently
In formula: fkFor the corresponding main frequency of vibration source signal each in system；M is the number of vibrating sensor；PiIt is i-th
The vibration source number that observation channel estimates；The main frequency for j-th of vibration source signal that channel estimates is observed for i-th
Rate, n are positive integer；
S6: vibration source signal is divided into S sections, it is the Related Component of the signal of T with extracting cycle that each section, which sets Q point,；
According to formula (7), vibration source signal is extracted from each vibrating sensor
In formula: xi(n Δ t) is the discrete form of the observation signal on i-th of channel, Δ t=1/Fs, FsFor vibration source signal
Sample frequency；
S7: the optimal estimation of each vibration source signal is determined according to formula (8)
In formula: Si,2Indicate two norms of i-th of vibration source signal；Sij,2Indicate i-th extracted from j-th of sensor vibration
Two norms of source signal；P is main frequency number；M is the number of vibrating sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910519782.2A CN110146152A (en) | 2019-06-17 | 2019-06-17 | Aero-engine vibrates source separation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910519782.2A CN110146152A (en) | 2019-06-17 | 2019-06-17 | Aero-engine vibrates source separation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110146152A true CN110146152A (en) | 2019-08-20 |
Family
ID=67591687
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910519782.2A Pending CN110146152A (en) | 2019-06-17 | 2019-06-17 | Aero-engine vibrates source separation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110146152A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102519582A (en) * | 2011-12-22 | 2012-06-27 | 南京航空航天大学 | Blind source separation method of aeroengine vibration signal |
| CN109684898A (en) * | 2017-10-18 | 2019-04-26 | 中国航发商用航空发动机有限责任公司 | Aero-engine and its vibration signal blind separating method and device |
-
2019
- 2019-06-17 CN CN201910519782.2A patent/CN110146152A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102519582A (en) * | 2011-12-22 | 2012-06-27 | 南京航空航天大学 | Blind source separation method of aeroengine vibration signal |
| CN109684898A (en) * | 2017-10-18 | 2019-04-26 | 中国航发商用航空发动机有限责任公司 | Aero-engine and its vibration signal blind separating method and device |
Non-Patent Citations (1)
| Title |
|---|
| 杨广振 等: "利用航空发动机信号特征的振动源盲分离算法", 《西安交通大学学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Delgado-Arredondo et al. | Methodology for fault detection in induction motors via sound and vibration signals | |
| CN107702908A (en) | GIS mechanical oscillation signal Time-Frequency Analysis Methods based on VMD self adapting morphologies | |
| CN110763462B (en) | Time-varying vibration signal fault diagnosis method based on synchronous compression operator | |
| CN105929201B (en) | A method of the accelerometer dynamic model structure parameter identification based on refinement spectrum analysis | |
| CN109684898A (en) | Aero-engine and its vibration signal blind separating method and device | |
| CN107843740B (en) | A kind of rotating speed measurement method of fusion vibration and voice signal spectrum signature | |
| CN107607065A (en) | A kind of impact echo signal analysis method based on variation mode decomposition | |
| CN109668733A (en) | Variation nonlinear model decomposes variable speed Method for Bearing Fault Diagnosis | |
| Antoni et al. | Cyclostationarity in condition monitoring: 10 years after | |
| Coats et al. | Extraction of tacho information from a vibration signal for improved synchronous averaging | |
| Zachar et al. | Utilizing spin-down transients for vibration-based diagnostics of resiliently mounted machines | |
| Miao et al. | Identification of characteristic components in frequency domain from signal singularities | |
| CN108398260A (en) | The fast evaluation method of gear-box instantaneous angular velocity based on mixing probabilistic method | |
| CN205786484U (en) | A kind of test device of composite sound wave attenuation coefficient | |
| US10948376B2 (en) | Apparatus and method of detecting leak sound in plant equipment using time-frequency transformation | |
| CN110146152A (en) | Aero-engine vibrates source separation method | |
| CN108398183B (en) | Multi-line spectrum Duffing oscillator detection and frequency identification method with unknown parameters under strong noise background | |
| JP6430234B2 (en) | Vibration analysis apparatus and program for rotating machine | |
| Niu et al. | Mode separation with one hydrophone in shallow water: A sparse Bayesian learning approach based on phase speed | |
| CN108680247A (en) | Based on the modified vibration signal conversion method of vibration severity low frequency filtering | |
| CN108646248B (en) | Passive acoustic speed and distance measuring method for low-speed moving sound source | |
| CN105954353A (en) | Test method and device of comprehensive acoustic attenuation coefficient | |
| CN100498229C (en) | Method for processing periodic error in inertial components | |
| CN114297579A (en) | Blind source separation method based on cyclostationary measure | |
| Mokhtari et al. | Wavelet spectral techniques for error mitigation in the superconductive angular accelerometer output of a gravity gradiometer system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190820 |
|
| RJ01 | Rejection of invention patent application after publication |