CN108509922A - A kind of rotary machinery fault diagnosis method and system - Google Patents
A kind of rotary machinery fault diagnosis method and system Download PDFInfo
- Publication number
- CN108509922A CN108509922A CN201810297692.9A CN201810297692A CN108509922A CN 108509922 A CN108509922 A CN 108509922A CN 201810297692 A CN201810297692 A CN 201810297692A CN 108509922 A CN108509922 A CN 108509922A
- Authority
- CN
- China
- Prior art keywords
- illustrative plates
- collection
- fault
- edge
- vibration
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M7/00—Vibration-testing of structures; Shock-testing of structures
- G01M7/02—Vibration-testing by means of a shake table
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a kind of rotary machinery fault diagnosis method and systems.The method includes：Obtain the fault vibration signal to be diagnosed of rotating machinery and multiple known single fault vibration signals；Fault-signal vibration collection of illustrative plates to be diagnosed is built using S-transformation and multiple known single faults vibrate collection of illustrative plates；The gradient of all vibration collection of illustrative plates is calculated using Sobel edge detection operators；First edge collection of illustrative plates and multiple second edge collection of illustrative plates are obtained according to gradient；First edge collection of illustrative plates is matched respectively with multiple second edge collection of illustrative plates using SURF algorithm, obtains multiple match point groups；The number for screening matching double points is more than the match point group of preset value, obtains match point screening group；It determines in fault vibration signal to be diagnosed comprising the corresponding known single fault vibration signal of match point screening group.Method or system using the present invention, can effectively realize combined failure character separation, improve the accuracy of fault diagnosis.
Description
Technical field
The present invention relates to mechanical fault diagnosis technical fields, more particularly to a kind of rotary machinery fault diagnosis method
And system.
Background technology
In Practical Project, rotating machinery fault tends not to individually occur, but it is multiple intercouple, it is interactional therefore
Barrier exists simultaneously composition combined failure.Since failure generates the difference of root, combined failure of rotating machinery is in vibration signal upper table
It is now the superposition coupling of each single fault, the non-linear mixing of each single fault signal is shown as in time-domain, in frequency domain upper table
It is now the intersection and aliasing of each single fault signal characteristic frequency, each single fault signal is shown as on high-dimensional data space low
Overlapping in manifold is tieed up, shows that two dimensional surface space is rotation, scaling and the single fault signal of each single fault signal
Between overlap, block.
Currently, the rotary machinery fault diagnosis method of generally use is：First, using spectrum analysis, wavelet transformation, experience
The single faults signal processing method such as mode decomposition carries out feature extraction；Then support vector machine, artificial immunity, blind source point are recycled
From etc. artificial intelligence approaches carry out tagsort.This method has good effect to the diagnosis of rotating machinery single failure, still
Due to cannot well be detached to combined failure characteristic frequency aliased portion, for combined failure of rotating machinery
Diagnosis effect is not good enough, and the accuracy of diagnosis is not high.
Invention content
Based on this, it is necessary to a kind of rotary machinery fault diagnosis method and system are provided, with improve rotating machinery it is compound therefore
The diagnosis effect of barrier improves the accuracy of diagnosis.
To achieve the above object, the present invention provides following schemes：
A kind of rotary machinery fault diagnosis method, including：
Obtain the fault vibration signal to be diagnosed of rotating machinery and multiple known single fault vibration signals；
Using S-transformation build described in fault vibration signal to be diagnosed and multiple known single fault vibration signals time-frequency figure,
Obtain fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single fault vibration collection of illustrative plates；
The gradient of all vibration collection of illustrative plates is calculated using Sobel edge detection operators；The gradient includes lateral, longitudinal direction, 45
Spend the gradient on direction and 135 degree of direction four directions；
First edge collection of illustrative plates and multiple second edge collection of illustrative plates are obtained according to the gradient；The first edge collection of illustrative plates is edge
Fault-signal to be diagnosed after detection vibrates collection of illustrative plates, and the second edge collection of illustrative plates is the known single fault vibrorecord after edge detection
Spectrum；
The first edge collection of illustrative plates is matched respectively with multiple second edge collection of illustrative plates using SURF algorithm, is obtained
Multiple match point groups；One match point group is matched by the first edge collection of illustrative plates and a second edge collection of illustrative plates
Obtained multiple matching double points are constituted；
The number for screening matching double points is more than the match point group of preset value, obtains match point screening group；
Include the corresponding known single fault vibration of the match point screening group in fault vibration signal to be diagnosed described in determination
Signal.
Optionally, in the fault vibration signal to be diagnosed for obtaining rotating machinery and multiple known single fault vibration signals
Later, further include：
Fault vibration signal to be diagnosed and multiple known single fault vibration signals to the rotating machinery carry out small
Wave packet decomposes；
The vibration signal after decomposition is filtered using low pass Butterworth filter.
Optionally, the gradient that all vibration collection of illustrative plates are calculated using Sobel edge detection operators, specially：
N=M1=| Sx|+|Sy|+|S45|+|S135| or
N=M∞=max { Sx|,Sy||S45|,S135|}
N indicates gradient, M1Indicate vector [Sx Sy S45 S135] 1 norm, M∞Indicate vector [Sx Sy S45 S135] nothing
Poor norm, SxIndicate the operator of detection vibration collection of illustrative plates transverse direction, SyIndicate the operator of detection vibration collection of illustrative plates longitudinal direction, S45Indicate that detection is shaken
Cardon composes the operator in 45 degree of directions, S135Indicate the operator in 135 degree of directions of detection vibration collection of illustrative plates.
Optionally, it after obtaining first edge collection of illustrative plates and multiple second edge collection of illustrative plates according to the gradient described, also wraps
It includes：
Morphological erosion processing is carried out to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates.
The present invention also provides a kind of Rotary Fault Diagnosis Systems, including：
Signal acquisition module, the fault vibration signal to be diagnosed for obtaining rotating machinery and multiple known single fault vibrations
Signal；
Map construction module, for fault vibration signal to be diagnosed and multiple known single faults described in being built using S-transformation
The time-frequency figure of vibration signal obtains fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single fault vibration collection of illustrative plates；
Gradient computing module, the gradient for calculating all vibration collection of illustrative plates using Sobel edge detection operators；The gradient
Including the gradient on lateral, longitudinal, 45 degree of directions and 135 degree of direction four directions；
Edge collection of illustrative plates acquisition module, for obtaining first edge collection of illustrative plates and multiple second edge collection of illustrative plates according to the gradient；
The first edge collection of illustrative plates is that the fault-signal to be diagnosed after edge detection vibrates collection of illustrative plates, and the second edge collection of illustrative plates is examined for edge
Known single fault after survey vibrates collection of illustrative plates；
Matching module, for being distinguished the first edge collection of illustrative plates and multiple second edge collection of illustrative plates using SURF algorithm
It is matched, obtains multiple match point groups；One match point group is by the first edge collection of illustrative plates and second side
Multiple matching double points that edge collection of illustrative plates is matched are constituted；
Screening module, the number for screening matching double points are more than the match point group of preset value, obtain match point screening group；
Determining module, for including that the match point screening group is corresponding in fault vibration signal to be diagnosed described in determination
Know single fault vibration signal.
Optionally, the system also includes first processing module, the first processing module specifically includes：
Resolving cell, for the rotating machinery fault vibration signal to be diagnosed and multiple known single faults shake
Dynamic signal carries out WAVELET PACKET DECOMPOSITION；
Filter unit, for being filtered to the vibration signal after decomposition using low pass Butterworth filter.
Optionally, the system also includes Second processing module, the Second processing module specifically includes：
Processing unit, for being carried out at morphological erosion to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates
Reason.
Compared with prior art, the beneficial effects of the invention are as follows：
The present invention proposes a kind of rotary machinery fault diagnosis method and system, the method includes：Obtain rotating machinery
Vibration signal；Vibration collection of illustrative plates is built using S-transformation；The gradient of vibration collection of illustrative plates is calculated using Sobel edge detection operators；Gradient
Including the gradient on lateral, longitudinal, 45 degree of directions and 135 degree of direction four directions；According to gradient obtain first edge collection of illustrative plates and
Multiple second edge collection of illustrative plates；First edge collection of illustrative plates is matched respectively with multiple second edge collection of illustrative plates using SURF algorithm, is obtained
To multiple match point groups；Judge whether the number of matching double points in match point group is more than preset value；If so, determining combined failure
It include the corresponding single fault vibration signal of match point group in vibration signal；If not, it is determined that do not wrapped in combined failure vibration signal
The corresponding single fault vibration signal of group containing match point.The present invention ensure that fault message is clear using S-transformation structure vibration collection of illustrative plates
Ground is embodied in vibration collection of illustrative plates, improves the accuracy of fault diagnosis；Vibration collection of illustrative plates is calculated using Sobel edge detection operators to exist
Laterally, the gradient on longitudinal, 45 degree of directions and 135 degree of direction four directions, and SURF algorithm is combined to carry out match cognization, it realizes
Combined failure character separation, further improves the accuracy of fault diagnosis.
Description of the drawings
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the present invention
Example, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is a kind of flow chart of rotary machinery fault diagnosis method of the embodiment of the present invention；
Fig. 2 is the time-frequency gray-scale map of the vibration signal under different situations；
Fig. 3 be misalign the vibration collection of illustrative plates of fault vibration signal with it is uneven-misalign shaking for combined failure vibration signal
The matching result figure of cardon spectrum；
Fig. 4 is that the vibration collection of illustrative plates of imbalance fault vibration signal and imbalance-misalign shaking for combined failure vibration signal
The matching result figure of cardon spectrum；
Fig. 5 is a kind of structural schematic diagram of Rotary Fault Diagnosis System of the embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, below in conjunction with the accompanying drawings and specific real
Applying mode, the present invention is described in further detail.
Fig. 1 is a kind of flow chart of rotary machinery fault diagnosis method of the embodiment of the present invention.
Referring to Fig. 1, the rotary machinery fault diagnosis method of embodiment, including：
Step S1：Obtain the fault vibration signal to be diagnosed of rotating machinery and multiple known single fault vibration signals.
The known single fault vibration signal is by experimental bench, being simulated to single failure.
Step S2：Using S-transformation build described in fault vibration signal to be diagnosed and multiple known single fault vibration signals
Time-frequency figure obtains fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single fault vibration collection of illustrative plates.
Step S3：The gradient of all vibration collection of illustrative plates is calculated using Sobel edge detection operators.
The gradient includes gradient laterally, on longitudinal, 45 degree of directions and 135 degree of direction four directions, gradient it is specific
It calculates as follows：
N=M1=| Sx|+|Sy|+|S45|+|S135| or
N=M∞=max { Sx|,Sy||S45|,S135|}
N indicates gradient, M1Indicate vector [Sx Sy S45S135] 1 norm, M∞Indicate vector [Sx Sy S45 S135] nothing
Poor norm, SxIndicate the operator of detection vibration collection of illustrative plates transverse direction, SyIndicate the operator of detection vibration collection of illustrative plates longitudinal direction, S45Indicate that detection is shaken
Cardon composes the operator in 45 degree of directions, S135Indicate the operator in 135 degree of directions of detection vibration collection of illustrative plates.
Step S4：First edge collection of illustrative plates and multiple second edge collection of illustrative plates are obtained according to the gradient.The first edge figure
It composes and vibrates collection of illustrative plates for the fault-signal to be diagnosed after edge detection, the second edge collection of illustrative plates is known single event after edge detection
Barrier vibration collection of illustrative plates.
Step S3 in the present embodiment and step S4 realizes the detection to vibrating collection of illustrative plates multiple directions, edge detection
Precision it is high, but also carrying out the first edge collection of illustrative plates obtained after Edge Gradient Feature and second edge collection of illustrative plates to vibration collection of illustrative plates
Clarity higher.
Step S5：The first edge collection of illustrative plates and multiple second edge collection of illustrative plates are carried out respectively using SURF algorithm
Match, obtains multiple match point groups.
One match point group is matched to obtain by the first edge collection of illustrative plates and a second edge collection of illustrative plates
Multiple matching double points constitute.
Step S6：The number for screening matching double points is more than the match point group of preset value, obtains match point screening group.
Step S7：In fault vibration signal to be diagnosed described in determination it is corresponding known single comprising the match point screening group therefore
Hinder vibration signal.
In the present embodiment, after step S1, also to the obtained fault vibration signal to be diagnosed of rotating machinery and multiple
Known single fault vibration signal is pre-processed, and to remove extra interference noise, preprocessing process is as follows：
1) the fault vibration signal to be diagnosed of the rotating machinery and multiple known single fault vibration signals are carried out
WAVELET PACKET DECOMPOSITION.Specially：
It selects wavelet function for db2 small echos, WAVELET PACKET DECOMPOSITION, Decomposition order n, in n is carried out to the vibration signal of acquisition
When layer decomposes, n frequency range will be generated, since sample frequency is w Hz, the frequency range representated by the vibration data of acquisition is 0~
W/2Hz, according to the principle of WAVELET PACKET DECOMPOSITION, can deduce the band width shared by each frequency range of n-th layer be (w/2)/
256Hz。
2) vibration signal after decomposition is filtered using low pass Butterworth filter.Specially：
N-th of frequency range of step 1) wavelet decomposition is subjected to low-pass filtering, that is, establishes the low pass Butterworth filter of a n rank
Wave device obtains multiple known single fault vibration letters after diagnosis fault vibration signal and denoising of the rotating machinery after denoising
Number.
In the present embodiment, after step s4, denoising also is carried out to first edge collection of illustrative plates and multiple second edge collection of illustrative plates
Processing, specially：Morphological erosion processing is carried out to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates, to eliminate
Some isolated points and pseudo-edge so that first edge collection of illustrative plates, the edge of second edge collection of illustrative plates are rounder and more smooth and continuous.
Above-mentioned rotary machinery fault diagnosis method is verified below.
Using rotor-support-foundation system as object, on the rotor fault simulated experiment platform of Spectra Quest companies, rotor is being established just
Often, uneven, misalign and imbalance misaligns combined failure etc., with PULSE multichannel data acquisition systems to every group therefore
The vibration signal of barrier is acquired, specially：Acceleration transducer is arranged on rotor two bearings seat, by axial, radial
And three direction placement sensors such as vertical, realization are acquired fault vibration signal.
It is chosen at and turns frequency and studied for the data acquired under 30Hz working conditions, each group of selection 16384 is list
Position (each group can obtain 10 units) carries out WAVELET PACKET DECOMPOSITION to the data of selection, removes noise, low-pass filtering obtains pre-
The data of processing.Data after pretreatment are subjected to S-transformation and obtain time-frequency figure.Fig. 2 is the vibration signal under different situations
Time-frequency gray-scale map, Fig. 2 (a) are the time-frequency gray-scale map of imbalance fault vibration signal, and Fig. 2 (b) is to misalign fault vibration signal
Time-frequency gray-scale map, Fig. 2 (c) is the time-frequency gray-scale map of vibration signal under normal condition, Fig. 2 (d) be it is uneven-misalign it is compound
The time-frequency gray-scale map of fault vibration signal.
All time-frequency gray-scale maps are calculated in lateral, longitudinal, 35 degree of directions and 135 degree of sides using Sobel edge detection operators
Gradient on four direction, and obtain the vibration collection of illustrative plates after edge detection；It will be not right after edge detection using SURF algorithm
The vibration collection of illustrative plates that the vibration collection of illustrative plates of middle fault vibration signal misaligns combined failure vibration signal with imbalance-is matched, and is obtained
To the first match point group, Fig. 3 be misalign the vibration collection of illustrative plates of fault vibration signal with it is uneven-misalign combined failure vibration and believe
Number vibration collection of illustrative plates matching result figure, wherein Fig. 3 (a) be edge detection after the vibration collection of illustrative plates for misaligning fault vibration signal
The matching result figure of the vibration collection of illustrative plates of combined failure vibration signal is misaligned with the imbalance-after edge detection, Fig. 3 (b) is not
The time-frequency gray-scale map of centering fault vibration signal misaligns the matching of the time-frequency gray-scale map of combined failure vibration signal with imbalance-
Therefore result figure, demonstrates imbalance-no from the figure 3, it may be seen that the number of matching double points is more than preset value in the first match point group
It is contained in centering combined failure vibration signal and misaligns fault vibration signal.
The vibration collection of illustrative plates of the imbalance fault vibration signal after edge detection and imbalance-are misaligned using SURF algorithm
The vibration collection of illustrative plates of combined failure vibration signal is matched, and the second match point group is obtained, and Fig. 4 is imbalance fault vibration signal
Vibration collection of illustrative plates misaligns the matching result figure of the vibration collection of illustrative plates of combined failure vibration signal with imbalance-, and wherein Fig. 4 (a) is side
The vibration collection of illustrative plates of imbalance fault vibration signal after edge detection misaligns combined failure vibration with the imbalance-after edge detection
The matching result figure of the vibration collection of illustrative plates of signal, Fig. 4 (b) be imbalance fault vibration signal time-frequency gray-scale map with it is uneven-no
The matching result figure of the time-frequency gray-scale map of centering combined failure vibration signal, as shown in Figure 4, matching double points in the second match point group
Number be more than preset value, therefore, demonstrate it is uneven-misalign and contain imbalance fault in combined failure vibration signal and shake
Dynamic signal.
Rotary machinery fault diagnosis method in the present embodiment builds vibration collection of illustrative plates using S-transformation, ensure that fault message
It is clearly embodied in vibration collection of illustrative plates, improves the accuracy of fault diagnosis；Vibrorecord is calculated using Sobel edge detection operators
The gradient on lateral, longitudinal, 45 degree of directions and 135 degree of direction four directions is composed, and SURF algorithm is combined to carry out match cognization,
Combined failure character separation is realized, the accuracy of fault diagnosis is further improved.
The present invention also provides a kind of Rotary Fault Diagnosis System, Fig. 5 is a kind of rotating machinery of the embodiment of the present invention
The structural schematic diagram of fault diagnosis system.
Rotary Fault Diagnosis System in embodiment, including：
Signal acquisition module 501, the fault vibration signal to be diagnosed for obtaining rotating machinery and multiple known single faults
Vibration signal.
First processing module 502, for fault vibration signal to be diagnosed to the rotating machinery and multiple described known
Single fault vibration signal is handled.
The first processing module 502, specifically includes：
Resolving cell, for the rotating machinery fault vibration signal to be diagnosed and multiple known single faults shake
Dynamic signal carries out WAVELET PACKET DECOMPOSITION.
Filter unit, for being filtered to the vibration signal after decomposition using low pass Butterworth filter.
Map construction module 503, for fault vibration signal to be diagnosed described in being built using S-transformation and it is multiple known single therefore
The time-frequency figure for hindering vibration signal obtains fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single fault vibration collection of illustrative plates.
Gradient computing module 504, the gradient for calculating all vibration collection of illustrative plates using Sobel edge detection operators；It is described
Gradient includes the gradient on lateral, longitudinal, 45 degree of directions and 135 degree of direction four directions.The specific calculating of the gradient is as follows：
N=M1=| Sx|+|Sy|+|S45|+|S135| or
N=M∞=max { Sx|,Sy||S45|,S135|}
N indicates gradient, M1Indicate vector [Sx Sy S45 S135] 1 norm, M∞Indicate vector [Sx Sy S45 S135] nothing
Poor norm, SxIndicate the operator of detection vibration collection of illustrative plates transverse direction, SyIndicate the operator of detection vibration collection of illustrative plates longitudinal direction, S45Indicate that detection is shaken
Cardon composes the operator in 45 degree of directions, S135Indicate the operator in 135 degree of directions of detection vibration collection of illustrative plates.
Edge collection of illustrative plates acquisition module 505, for obtaining first edge collection of illustrative plates and multiple second edge figures according to the gradient
Spectrum；The first edge collection of illustrative plates is that the fault-signal to be diagnosed after edge detection vibrates collection of illustrative plates, and the second edge collection of illustrative plates is side
Known single fault after edge detection vibrates collection of illustrative plates.
Second processing module 506, for handling the first edge collection of illustrative plates and multiple second edge collection of illustrative plates.
The Second processing module 506, specifically includes：
Processing unit, for being carried out at morphological erosion to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates
Reason.
Matching module 507, for utilizing SURF algorithm by the first edge collection of illustrative plates and multiple second edge collection of illustrative plates
It is matched respectively, obtains multiple match point groups；One match point group is by the first edge collection of illustrative plates and one described
Multiple matching double points that two edge collection of illustrative plates are matched are constituted.
Screening module 508, the number for screening matching double points are more than the match point group of preset value, obtain match point screening
Group.
Determining module 509, for including that the match point screening group corresponds in fault vibration signal to be diagnosed described in determination
Known single fault vibration signal.
Rotary Fault Diagnosis System in the present embodiment builds vibration collection of illustrative plates using S-transformation, ensure that fault message
It is clearly embodied in vibration collection of illustrative plates, improves the accuracy of fault diagnosis；Vibrorecord is calculated using Sobel edge detection operators
The gradient on lateral, longitudinal, 45 degree of directions and 135 degree of direction four directions is composed, and SURF algorithm is combined to carry out match cognization,
Combined failure character separation is realized, the accuracy of fault diagnosis is further improved.
In this specification for system disclosed in embodiment, since it is corresponded to the methods disclosed in the examples, institute
With the fairly simple of description, reference may be made to the description of the method.
Principle and implementation of the present invention are described for specific case used herein, and above example is said
The bright method and its core concept for being merely used to help understand the present invention；Meanwhile for those of ordinary skill in the art, foundation
The thought of the present invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (7)
1. a kind of rotary machinery fault diagnosis method, which is characterized in that including：
Obtain the fault vibration signal to be diagnosed of rotating machinery and multiple known single fault vibration signals；
Using S-transformation build described in fault vibration signal to be diagnosed and multiple known single fault vibration signals time-frequency figure, obtain
Fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single faults vibrate collection of illustrative plates；
The gradient of all vibration collection of illustrative plates is calculated using Sobel edge detection operators；The gradient includes lateral, longitudinal, 45 degree of sides
To with the gradient on 135 degree of direction four directions；
First edge collection of illustrative plates and multiple second edge collection of illustrative plates are obtained according to the gradient；The first edge collection of illustrative plates is edge detection
Fault-signal to be diagnosed afterwards vibrates collection of illustrative plates, and the second edge collection of illustrative plates is that the known single fault after edge detection vibrates collection of illustrative plates；
The first edge collection of illustrative plates is matched respectively with multiple second edge collection of illustrative plates using SURF algorithm, is obtained multiple
Match point group；One match point group is matched to obtain by the first edge collection of illustrative plates and a second edge collection of illustrative plates
Multiple matching double points constitute；
The number for screening matching double points is more than the match point group of preset value, obtains match point screening group；
Include the corresponding known single fault vibration signal of the match point screening group in fault vibration signal to be diagnosed described in determination.
2. a kind of rotary machinery fault diagnosis method according to claim 1, which is characterized in that in the acquisition whirler
After the fault vibration signal to be diagnosed of tool and multiple known single fault vibration signals, further include：
Fault vibration signal to be diagnosed and multiple known single fault vibration signals to the rotating machinery carry out wavelet packet
It decomposes；
The vibration signal after decomposition is filtered using low pass Butterworth filter.
3. a kind of rotary machinery fault diagnosis method according to claim 1, which is characterized in that described to utilize the sides Sobel
Edge detective operators calculate the gradient of all vibration collection of illustrative plates, specially：
N=M1=| Sx|+|Sy|+|S45|+|S135| or
N=M∞=max | Sx|,|Sy||S45|,|S135}
N indicates gradient, M1Indicate vector [Sx Sy S45 S135] 1 norm, M∞Indicate vector [Sx Sy S45 S135] infinite model
Number, SxIndicate the operator of detection vibration collection of illustrative plates transverse direction, SyIndicate the operator of detection vibration collection of illustrative plates longitudinal direction, S45Indicate detection vibrorecord
Compose the operator in 45 degree of directions, S135Indicate the operator in 135 degree of directions of detection vibration collection of illustrative plates.
4. a kind of rotary machinery fault diagnosis method according to claim 1, which is characterized in that described according to the ladder
After degree obtains first edge collection of illustrative plates and multiple second edge collection of illustrative plates, further include：
Morphological erosion processing is carried out to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates.
5. a kind of Rotary Fault Diagnosis System, which is characterized in that including：
Signal acquisition module, the fault vibration signal to be diagnosed for obtaining rotating machinery and multiple known single fault vibration letters
Number；
Map construction module, for fault vibration signal to be diagnosed described in being built using S-transformation and multiple known single fault vibrations
The time-frequency figure of signal obtains fault-signal vibration collection of illustrative plates to be diagnosed and multiple known single fault vibration collection of illustrative plates；
Gradient computing module, the gradient for calculating all vibration collection of illustrative plates using Sobel edge detection operators；The gradient includes
Laterally, the gradient on longitudinal, 45 degree of directions and 135 degree of direction four directions；
Edge collection of illustrative plates acquisition module, for obtaining first edge collection of illustrative plates and multiple second edge collection of illustrative plates according to the gradient；It is described
First edge collection of illustrative plates is that the fault-signal to be diagnosed after edge detection vibrates collection of illustrative plates, and the second edge collection of illustrative plates is after edge detection
Known single fault vibrate collection of illustrative plates；
Matching module, for being carried out the first edge collection of illustrative plates and multiple second edge collection of illustrative plates respectively using SURF algorithm
Matching, obtains multiple match point groups；One match point group is by the first edge collection of illustrative plates and a second edge figure
The multiple matching double points matched are composed to constitute；
Screening module, the number for screening matching double points are more than the match point group of preset value, obtain match point screening group；
Determining module, for including that the match point screening group is corresponding known single in fault vibration signal to be diagnosed described in determination
Fault vibration signal.
6. a kind of Rotary Fault Diagnosis System according to claim 5, which is characterized in that further include the first processing mould
Block, the first processing module, specifically includes：
Resolving cell, for the rotating machinery fault vibration signal to be diagnosed and multiple known single fault vibration letters
Number carry out WAVELET PACKET DECOMPOSITION；
Filter unit, for being filtered to the vibration signal after decomposition using low pass Butterworth filter.
7. a kind of Rotary Fault Diagnosis System according to claim 5, which is characterized in that further include second processing mould
Block, the Second processing module, specifically includes：
Processing unit, for carrying out morphological erosion processing to the first edge collection of illustrative plates and multiple second edge collection of illustrative plates.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810297692.9A CN108509922A (en) | 2018-03-30 | 2018-03-30 | A kind of rotary machinery fault diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810297692.9A CN108509922A (en) | 2018-03-30 | 2018-03-30 | A kind of rotary machinery fault diagnosis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108509922A true CN108509922A (en) | 2018-09-07 |
Family
ID=63380487
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810297692.9A Pending CN108509922A (en) | 2018-03-30 | 2018-03-30 | A kind of rotary machinery fault diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108509922A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112163630A (en) * | 2020-10-12 | 2021-01-01 | 清华大学 | Compound fault diagnosis method and device based on unbalanced learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04241040A (en) * | 1991-01-25 | 1992-08-28 | Nec Corp | Fault simulation method |
| CN103471841A (en) * | 2013-09-30 | 2013-12-25 | 国家电网公司 | Method for diagnosing vibration faults of rotary machine |
| CN106257535A (en) * | 2016-08-11 | 2016-12-28 | 河海大学常州校区 | Electrical equipment based on SURF operator is infrared and visible light image registration method |
| CN107784659A (en) * | 2017-10-16 | 2018-03-09 | 华南理工大学 | A kind of method for searching for the similar visible images of electrical equipment infrared image |
-
2018
- 2018-03-30 CN CN201810297692.9A patent/CN108509922A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04241040A (en) * | 1991-01-25 | 1992-08-28 | Nec Corp | Fault simulation method |
| CN103471841A (en) * | 2013-09-30 | 2013-12-25 | 国家电网公司 | Method for diagnosing vibration faults of rotary machine |
| CN106257535A (en) * | 2016-08-11 | 2016-12-28 | 河海大学常州校区 | Electrical equipment based on SURF operator is infrared and visible light image registration method |
| CN107784659A (en) * | 2017-10-16 | 2018-03-09 | 华南理工大学 | A kind of method for searching for the similar visible images of electrical equipment infrared image |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112163630A (en) * | 2020-10-12 | 2021-01-01 | 清华大学 | Compound fault diagnosis method and device based on unbalanced learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sung et al. | Locating defects of a gear system by the technique of wavelet transform | |
| CN101221066B (en) | Engineering nonlinear vibration detecting method | |
| CN101929917B (en) | Fault diagnosis method for rotary machine | |
| CN107505135A (en) | A kind of rolling bearing combined failure extracting method and system | |
| CN106124126B (en) | The separated method for testing and analyzing of whole vehicle state underdrive system unbalance dynamic | |
| El Badaoui et al. | Use of the moving cepstrum integral to detect and localise tooth spalls in gears | |
| CN101907089A (en) | Fault diagnosis method of compressor shafting based on three-dimensional space axle center orbit | |
| US20170067860A1 (en) | Control Area Network Machine Diagostic | |
| CN108362492A (en) | Vibration separation method suitable for fault diagnosis of planetary gear train at low rotating speed | |
| CN108647667B (en) | A kind of implementation method of the frequency domain amplitude spectrum kurtosis figure based on signal Time-frequency Decomposition | |
| Xue et al. | The diagnostic analysis of the planet bearing faults using the torsional vibration signal | |
| CN109443768A (en) | A kind of Helicopter Main Reducer planetary gear vibration signal separation method | |
| CN106812050A (en) | Method for drawing the compaction state of ground | |
| CN114838936A (en) | Rolling bearing fault diagnosis method and system for sample unbalance under variable rotating speed working condition | |
| CN108760310A (en) | Accidental resonance Fault Diagnosis of Roller Bearings based on novel signal-to-noise ratio index | |
| CN100434890C (en) | Flexible rotor holographic dynamic balancing method based on empirical mode decomposition | |
| JP4307591B2 (en) | Method for monitoring vehicles with acceleration sensors, in particular helicopter planetary gearboxes | |
| CN108509922A (en) | A kind of rotary machinery fault diagnosis method and system | |
| CN105425298B (en) | The method and apparatus of numerical solidification during a kind of elimination finite difference forward modeling | |
| CN101368870B (en) | Preparation method of amplitude frequency spectrum used for mechanical rotor single cross section shaft vibration analysis | |
| JP4220019B2 (en) | Method for monitoring a power transmission device of a helicopter with an acceleration sensor | |
| CN104568132B (en) | Reference signal constraint-based mechanical characteristic acoustic signal frequency-domain semi-blind extraction method | |
| Çelebi | Golden gate bridge response: A study with low-amplitude data from three earthquakes | |
| CN101451882B (en) | Short time amplitude frequency spectrum array for single section shaft vibration analysis for mechanical rotor | |
| DE102019219371A1 (en) | Measuring arrangement |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180907 |