CN114136604B - Rotary equipment fault diagnosis method and system based on improved sparse dictionary - Google Patents
Rotary equipment fault diagnosis method and system based on improved sparse dictionary Download PDFInfo
- Publication number
- CN114136604B CN114136604B CN202111442076.6A CN202111442076A CN114136604B CN 114136604 B CN114136604 B CN 114136604B CN 202111442076 A CN202111442076 A CN 202111442076A CN 114136604 B CN114136604 B CN 114136604B
- Authority
- CN
- China
- Prior art keywords
- sparse
- dictionary
- fault
- component signal
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000003745 diagnosis Methods 0.000 title claims abstract description 46
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 12
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 17
- 238000012360 testing method Methods 0.000 claims description 13
- 238000001228 spectrum Methods 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 4
- 238000001514 detection method Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/021—Gearings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2134—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on separation criteria, e.g. independent component analysis
- G06F18/21345—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on separation criteria, e.g. independent component analysis enforcing sparsity or involving a domain transformation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Acoustics & Sound (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a rotating equipment fault diagnosis method and system based on an improved sparse dictionary, and relates to the technical field of nuclear power safety management, wherein the method comprises the following steps: s1: obtaining a vibration signal, and carrying out k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2; s2: constructing an overcomplete atom library to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal; s3: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity; s4: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification. According to the invention, fault signal extraction is carried out by improving a sparse dictionary method, key characteristic signals are reserved, different fault models are established according to fault types, and finally fault diagnosis is realized according to a classification method, so that the equipment fault diagnosis accuracy is effectively improved.
Description
Technical Field
The invention relates to the technical field of nuclear power safety management,
in particular, the invention relates to a rotating equipment fault diagnosis method and system based on an improved sparse dictionary.
Background
Along with the progress of society, the use of green new energy is more and more paid attention to, especially the huge potential of nuclear power is paid great attention to, but the nuclear energy especially needs to pay attention to the safety when in use, and then the operation and maintenance requirements for nuclear power are quite high, and at present, the existing nuclear power equipment is provided with a large number of rotating equipment, and the rotating equipment has a complex structure and is easily broken down due to the influence of environment and noise. Failure of any type of rotating equipment can cause unplanned system downtime, and a great deal of time is required for workers to perform failure location, maintenance, order replacement parts, etc., thus resulting in great economic loss. The rotating equipment internally comprises a motor, a bearing, a pump and other parts, and when the bearing and the gear are in fault, various transient impact response components appear in the vibration signal. Because of manufacturing errors and installation errors of the equipment, frequency components related to shaft rotation frequency exist in vibration signals, accurate separation of fault signals can be achieved under the conditions of strong noise and multiple interferences, and the method has important significance for state monitoring of a rotary mechanical system.
In recent years, along with the deep research of machine learning, a data model is gradually applied to fault diagnosis of rotating equipment, and a common method is to extract characteristics of collected vibration signals, label data of different fault types, and then realize data classification by various classifier methods such as a support vector machine, nearest neighbor, a neural network, a decision tree and the like. The method for extracting the characteristic parameters comprises the steps of extracting time domain characteristic parameters, processing vibration signals by using a signal processing method such as envelope spectrum analysis, wavelet analysis, empirical mode decomposition and the like on the basis of the frequency domain characteristic parameters, and then extracting the time domain characteristic parameters and the frequency domain characteristic parameters on different component signals, wherein the sparse decomposition method realizes sparse expression of the signals by constructing an overcomplete dictionary and has certain anti-noise capability and weak signal extraction capability.
For example, chinese patent invention patent CN111678691a discloses a gear fault detection method based on an improved sparse decomposition algorithm, which relates to the technical field of fault detection, and the method increases pretreatment on signals and optimal design on a parameter dictionary based on traditional parameter dictionary sparse reconstruction, and utilizes dual-tree complex wavelet decomposition to combine with maximum kurtosis principle to realize pretreatment on signals, thereby greatly reducing influence of noise on subsequent treatment, determining target characteristic parameters based on relevant filtering of laplace wavelet to construct an overcomplete dictionary, not only effectively reducing dictionary redundancy, but also enabling the designed dictionary to be more similar to fault characteristics, and finally combining with a matching pursuit algorithm to realize fault detection on extraction of impact characteristics in vibration signals.
However, the above fault detection method still has the following drawbacks: although the optimal overcomplete dictionary is preferred, the same overcomplete dictionary is still used for full-band signals, certain key signals cannot be effectively extracted, the signal reconstruction accuracy is reduced, the key signals are lost, and the equipment state evaluation cannot be effectively realized.
Therefore, how to design a rotating equipment fault diagnosis method or system based on an improved sparse dictionary becomes a current urgent problem to be solved.
Disclosure of Invention
The invention aims to provide the rotating equipment fault diagnosis method based on the improved sparse dictionary, which is convenient to execute, performs fault signal extraction by improving the sparse dictionary method, reserves key characteristic signals, establishes different fault models according to fault types, realizes fault diagnosis according to a classification method, and effectively improves equipment fault diagnosis accuracy.
In order to achieve the above purpose, the invention is realized by adopting the following technical scheme:
a rotating equipment fault diagnosis method based on an improved sparse dictionary, the method comprising the steps of:
s1: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
s2: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
s3: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
s4: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification.
Preferably, the predetermined value k is not less than 2 when step S1 is performed.
As a preferred aspect of the present invention, when step S2 is performed, the method specifically includes the steps of:
s2.1: constructing an initial optimal related atom based on the fault type of the vibration signal, randomly selecting partial data in the component signal to construct an initial dictionary, determining algorithm parameters of dictionary learning, and constructing an overcomplete dictionary, namely an overcomplete atom library;
s2.2: fixing the constructed overcomplete dictionary, and determining an optimal sparse coefficient;
s2.3: fixing sparse coefficients, updating the dictionary column by column to minimize residual errors, and finally determining an optimal dictionary to obtain a reconstructed sparse representation;
s2.4: repeatedly executing the steps S2.1 to S2.3 to obtain an optimal dictionary and a reconstruction sparse representation corresponding to each component signal;
s2.5: and combining the optimal sparse coefficients with the corresponding optimal dictionaries, and performing sparse reconstruction on each component signal.
As a preferred aspect of the present invention, when step S2.2 is performed, after the overcomplete dictionary is fixed, the sparse coding is adjusted so that the objective function error is minimized, and then the optimal sparse coefficient can be determined.
As a preferred aspect of the present invention, when step S3 is performed, the method specifically includes the steps of:
s3.1: calculating the time domain waveform kurtosis index on each scale of the component signal;
s3.2: calculating kurtosis index values of the envelope demodulation spectrum;
s3.3: according to the two kurtosis values, calculating to obtain impact sparsity, and reconstructing a component signal according to the impact sparsity value;
s3.4: steps S3.1 to S3.3 are repeatedly performed, and each component signal is reconstructed.
As a preferable mode of the invention, when the step S3.3 is executed, corresponding weight values of two kurtosis indexes are set, the sum of the two weight values is equal to one, the two weight values are respectively multiplied by the waveform kurtosis index and the kurtosis index value to obtain two kurtosis values, and the impact sparsity is obtained according to the two kurtosis values.
As a preferred aspect of the present invention, when step S4 is performed, the method specifically includes the steps of:
s4.1: extracting characteristic parameters of the reconstructed component signals, wherein the characteristic parameter extraction comprises effective values, peak values, kurtosis indexes, waveform indexes and frequency domain characteristic parameters;
s4.2: establishing a training set and a testing set according to the characteristic parameters and the corresponding fault categories, training the training set, and determining optimal parameters of the model;
s4.3: and testing the test set data, realizing fault classification of the test set data, and carrying out fault recognition.
Preferably, the training set is trained by using a gaussian process classification method when step S4.2 is performed.
In another aspect, the present invention further provides a rotating equipment fault diagnosis system based on an improved sparse dictionary, the system comprising:
and a pretreatment module: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
a first reconstruction module: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
and a second reconstruction module: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
and a diagnosis module: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification.
The rotating equipment fault diagnosis method and system based on the improved sparse dictionary have the beneficial effects that: the method is convenient to execute, fault signal extraction is carried out by improving a sparse dictionary method, key characteristic signals are reserved, different fault models are built according to fault types, and finally fault diagnosis is realized according to a classification method, so that the equipment fault diagnosis accuracy is effectively improved.
Drawings
FIG. 1 is a schematic flow chart of a method for diagnosing faults of rotary equipment based on an improved sparse dictionary;
FIG. 2 is a dictionary parameter of a component signal in one embodiment of a rotary equipment fault diagnosis method based on an improved sparse dictionary of the present invention;
FIG. 3 is a time domain contrast plot of another component signal in one embodiment of a rotating equipment fault diagnosis method based on an improved sparse dictionary of the present invention;
FIG. 4 is a spectral contrast plot of another component signal in one embodiment of a rotating equipment fault diagnosis method based on an improved sparse dictionary of the present invention;
FIG. 5 is a time domain diagram of a vibration signal after reconstruction from sparsity of impact for yet another component signal in one embodiment of a rotating equipment failure diagnosis method based on an improved sparse dictionary of the present invention;
fig. 6 is a schematic diagram of a module configuration of a rotary equipment fault diagnosis system based on an improved sparse dictionary of the present invention.
Detailed Description
The following are specific examples of the present invention, and the technical solutions of the present invention are further described, but the present invention is not limited to these examples.
Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the modules and steps set forth in these embodiments and the steps do not limit the scope of the present invention unless specifically stated otherwise.
Meanwhile, it should be understood that the flow in the drawings is not merely performed alone, but a plurality of steps are performed to cross each other for convenience of description.
The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses.
Techniques, methods, and systems known to those of ordinary skill in the relevant art may not be discussed in detail, but should be considered part of the authorization specification where appropriate.
Example 1
As shown in fig. 1 to 5, which are only one embodiment of the present invention, the present invention provides a rotating equipment fault diagnosis method based on an improved sparse dictionary, the method comprising the steps of:
s1: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
here, when step S1 is performed, the predetermined value k is not smaller than 2, for example, for the driving end acceleration data collected on a kesixi bearing simulation experiment table in a state that the rotation speed is 1797r/min, the sampling frequency is 12kHz, and in this case, the vibration signal is generally subjected to 3-layer wavelet packet decomposition, so as to obtain 8 component signals in equal frequency bands.
S2: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
in this embodiment, when executing step S2, the method specifically includes the following steps:
s2.1: constructing an initial optimal related atom based on the fault type of the vibration signal, randomly selecting partial data in the component signal to construct an initial dictionary, determining algorithm parameters of dictionary learning, and constructing an overcomplete dictionary, namely an overcomplete atom library;
s2.2: fixing the constructed overcomplete dictionary, and determining an optimal sparse coefficient;
s2.3: fixing sparse coefficients, updating the dictionary column by column to minimize residual errors, and finally determining an optimal dictionary to obtain a reconstructed sparse representation;
s2.4: repeatedly executing the steps S2.1 to S2.3 to obtain an optimal dictionary and a reconstruction sparse representation corresponding to each component signal;
s2.5: and combining the optimal sparse coefficients with the corresponding optimal dictionaries, and performing sparse reconstruction on each component signal.
Moreover, after the overcomplete dictionary is fixed while step S2.2 is performed, the sparse coding is adjusted so that the objective function error is minimized to determine the optimal sparse coefficient.
It should be noted that the parameters of the optimal dictionary for each component signal are different, and as shown in fig. 2, are parameters of the optimal dictionary obtained for the first component signal in step S1.
When executing step S2.5, sparse reconstruction is performed on each component signal, so as to obtain a time domain contrast chart and a frequency spectrum contrast chart before and after sparse reconstruction of each component signal, as shown in fig. 3 and fig. 4, so that step S3 is conveniently performed.
S3: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
in this embodiment, when executing step S3, the method specifically includes the following steps:
s3.1: calculating a time domain waveform kurtosis index K on each scale of the component signal, wherein the specific calculation score is as follows:
wherein the method comprises the steps ofIs the time-domain average value of the component signal,/>Is the fourth power of the time domain average value,/>Is the standard deviation of the component signals.
S3.2: the kurtosis index value Ke of the envelope demodulation spectrum is calculated, and the specific calculation score is as follows:
wherein:hilbert transform representing a signal;
s3.3: according to the two kurtosis values, calculating to obtain impact sparsity S, and reconstructing a component signal according to the impact sparsity value;
here, corresponding weight values α and β of two kurtosis indexes are set, the two weight values are added to be equal to one, that is, α+β=1, the two weight values are multiplied by the waveform kurtosis index and the kurtosis index value respectively to obtain two kurtosis values, and the impact sparsity S is calculated according to the two kurtosis values, wherein the calculation formula is as follows:
each may be set to 0.5 and the reconstructed signal as shown in fig. 5;
s3.4: steps S3.1 to S3.3 are repeatedly performed, and each component signal is reconstructed.
S4: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification.
In this embodiment, when executing step S4, the method specifically includes the following steps:
s4.1: extracting characteristic parameters of the reconstructed component signals, wherein the characteristic parameter extraction comprises, but is not limited to, an effective value, a peak value, kurtosis indexes, waveform indexes and frequency domain characteristic parameters, and generally 15 characteristic parameters are extracted in total;
s4.2: establishing a training set and a testing set according to the characteristic parameters and the corresponding fault categories, training the training set, and determining optimal parameters of the model;
s4.3: and testing the test set data, realizing fault classification of the test set data, and carrying out fault recognition.
Also, in performing step S4.2, the method of training the training set includes using a gaussian process classification method.
The number of samples is 1331, wherein the number of the samples is 121, 241, 242 and 727. 931 sets of data are randomly shuffled and selected as training samples, and 400 sets of data are selected as test samples. Training the training set by using a Gaussian process classification method, wherein a kernel function is selected as a Gaussian kernel function, and a parameter is selected as 1.
Finally, 70 inner ring fault samples are detected, the number of diagnosis is 70, and the diagnosis accuracy is 100%; 57 rolling body fault samples are detected, the number of diagnosis accuracy is 57, and the diagnosis accuracy is 100%; 231 outer ring fault samples are detected, the number of diagnosis accuracy is 231, and the diagnosis accuracy is 100%; 42 normal samples are detected, 42 diagnosis accuracy is achieved, and the diagnosis accuracy is 100%. The method provided by the invention can effectively improve the accuracy of identifying the faults of the rotary mechanical equipment.
The rotary equipment fault diagnosis method based on the improved sparse dictionary is convenient to execute, fault signal extraction is carried out through the improved sparse dictionary method, key characteristic signals are reserved, different fault models are built according to fault types, finally fault diagnosis is realized according to a classification method, and equipment fault diagnosis accuracy is effectively improved.
Example two
As shown in fig. 6, which is only one embodiment of the present invention, the present invention also provides a system in which a rotating equipment fault diagnosis method based on an improved sparse dictionary can be implemented in the above embodiment, the system comprising:
and a pretreatment module: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
a first reconstruction module: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
and a second reconstruction module: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
and a diagnosis module: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification.
The rotary equipment fault diagnosis method and system based on the improved sparse dictionary are convenient to execute, fault signal extraction is carried out through the improved sparse dictionary method, key characteristic signals are reserved, different fault models are built according to fault types, finally fault diagnosis is realized according to a classification method, and equipment fault diagnosis accuracy is effectively improved.
While certain specific embodiments of the present invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the foregoing examples are provided for the purpose of illustration only and are not intended to limit the scope of the invention, and that various modifications or additions and substitutions to the described specific embodiments may be made by those skilled in the art without departing from the scope of the invention or exceeding the scope of the invention as defined in the accompanying claims. It should be understood by those skilled in the art that any modification, equivalent substitution, improvement, etc. made to the above embodiments according to the technical substance of the present invention should be included in the scope of protection of the present invention.
Claims (6)
1. The rotary equipment fault diagnosis method based on the improved sparse dictionary is characterized by comprising the following steps of:
s1: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
s2: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
s3: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
s4: extracting features of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification;
when executing the step S2, the method specifically comprises the following steps:
s2.1: constructing an initial optimal related atom based on the fault type of the vibration signal, randomly selecting partial data in the component signal to construct an initial dictionary, determining algorithm parameters of dictionary learning, and constructing an overcomplete dictionary, namely an overcomplete atom library;
s2.2: fixing the constructed overcomplete dictionary, and determining an optimal sparse coefficient;
s2.3: fixing sparse coefficients, updating the dictionary column by column to minimize residual errors, and finally determining an optimal dictionary to obtain a reconstructed sparse representation;
s2.4: repeatedly executing the steps S2.1 to S2.3 to obtain an optimal dictionary and a reconstruction sparse representation corresponding to each component signal;
s2.5: combining the optimal sparse coefficients with the corresponding optimal dictionaries, and performing sparse reconstruction on each component signal respectively;
when executing the step S3, the method specifically comprises the following steps:
s3.1: calculating the time domain waveform kurtosis index on each scale of the component signal;
s3.2: calculating kurtosis index of the envelope demodulation spectrum;
s3.3: according to the kurtosis index of the time domain waveform in the step S3.1 and the kurtosis index of the envelope demodulation spectrum in the step S3.2, calculating to obtain impact sparsity, and reconstructing a component signal according to the impact sparsity value;
s3.4: repeatedly executing the steps S3.1 to S3.3, and reconstructing each component signal;
when executing the step S3.3, setting corresponding weight values of two kurtosis indexes, adding the two weight values to be equal to one, multiplying the two weight values by the kurtosis index of the time domain waveform and the kurtosis index of the envelope demodulation spectrum respectively to obtain two kurtosis values, and calculating according to the two kurtosis values to obtain the impact sparsity.
2. The rotary equipment fault diagnosis method based on the improved sparse dictionary of claim 1, wherein:
when step S1 is performed, the predetermined value k is not less than 2.
3. The rotary equipment fault diagnosis method based on the improved sparse dictionary of claim 1, wherein:
and when the step S2.2 is executed, after the overcomplete dictionary is fixed, the sparse coding is adjusted to minimize the error of the objective function, and then the optimal sparse coefficient can be determined.
4. The rotary equipment fault diagnosis method based on the improved sparse dictionary of claim 1, wherein:
when executing the step S4, the method specifically comprises the following steps: s4.1: extracting characteristic parameters of the reconstructed component signals, wherein the characteristic parameter extraction comprises an effective value, a peak value, a kurtosis index, a waveform index and a frequency domain characteristic parameter;
s4.2: establishing a training set and a testing set according to the characteristic parameters and the corresponding fault categories, training the training set, and determining optimal parameters of the model;
s4.3: and testing the test set data, realizing fault classification of the test set data, and carrying out fault recognition.
5. The rotary equipment fault diagnosis method based on the improved sparse dictionary of claim 4, wherein:
in performing step S4.2, the method of training the training set includes using a gaussian process classification method.
6. A rotating equipment fault diagnosis system based on an improved sparse dictionary, comprising:
and a pretreatment module: obtaining a vibration signal of a fault of rotating equipment, and carrying out a preset value k-layer wavelet packet decomposition on the vibration signal to obtain component signals of k equal frequency bands of 2;
a first reconstruction module: constructing an overcomplete atom library of the vibration signal to obtain a sparse coefficient in the overcomplete atom library, obtaining a reconstructed sparse representation, and carrying out single reconstruction on each component signal;
and a second reconstruction module: calculating the impact sparsity of each component signal, and realizing component signal reconstruction according to the impact sparsity;
and a diagnosis module: and extracting the characteristics of each reconstructed component signal, establishing a model according to the fault type of the vibration signal, and carrying out fault identification.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111442076.6A CN114136604B (en) | 2021-11-30 | 2021-11-30 | Rotary equipment fault diagnosis method and system based on improved sparse dictionary |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111442076.6A CN114136604B (en) | 2021-11-30 | 2021-11-30 | Rotary equipment fault diagnosis method and system based on improved sparse dictionary |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114136604A CN114136604A (en) | 2022-03-04 |
CN114136604B true CN114136604B (en) | 2023-11-28 |
Family
ID=80389781
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111442076.6A Active CN114136604B (en) | 2021-11-30 | 2021-11-30 | Rotary equipment fault diagnosis method and system based on improved sparse dictionary |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114136604B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115112061B (en) * | 2022-06-28 | 2023-07-25 | 苏州大学 | Rail wave grinding detection method and system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109765055A (en) * | 2019-01-31 | 2019-05-17 | 杭州安脉盛智能技术有限公司 | Rolling bearing fault testing method and system based on EWT, spectrum virtual value and KNN |
CN109827081A (en) * | 2019-02-28 | 2019-05-31 | 昆明理工大学 | A kind of buried drain pipe road plugging fault and branch pipe tee connection part diagnostic method based on acoustics active detecting |
CN109993105A (en) * | 2019-03-29 | 2019-07-09 | 北京化工大学 | A kind of improved adaptive sparse sampling Fault Classification |
CN111678691A (en) * | 2020-07-27 | 2020-09-18 | 江南大学 | Gear fault detection method based on improved sparse decomposition algorithm |
-
2021
- 2021-11-30 CN CN202111442076.6A patent/CN114136604B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109765055A (en) * | 2019-01-31 | 2019-05-17 | 杭州安脉盛智能技术有限公司 | Rolling bearing fault testing method and system based on EWT, spectrum virtual value and KNN |
CN109827081A (en) * | 2019-02-28 | 2019-05-31 | 昆明理工大学 | A kind of buried drain pipe road plugging fault and branch pipe tee connection part diagnostic method based on acoustics active detecting |
CN109993105A (en) * | 2019-03-29 | 2019-07-09 | 北京化工大学 | A kind of improved adaptive sparse sampling Fault Classification |
CN111678691A (en) * | 2020-07-27 | 2020-09-18 | 江南大学 | Gear fault detection method based on improved sparse decomposition algorithm |
Non-Patent Citations (2)
Title |
---|
基于稀疏表示的轴承早期故障特征提取;余发军;周凤星;严保康;;北京理工大学学报(第04期);48-53 * |
机械故障特征提取的拓扑集分形稀疏字典;曹新城;陈彬强;姚斌;贺王鹏;振动与冲击;第39卷(第06期);210-218 * |
Also Published As
Publication number | Publication date |
---|---|
CN114136604A (en) | 2022-03-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | Fault diagnosis of a rolling bearing using wavelet packet denoising and random forests | |
CN105275833B (en) | CEEMD (Complementary Empirical Mode Decomposition)-STFT (Short-Time Fourier Transform) time-frequency information entropy and multi-SVM (Support Vector Machine) based fault diagnosis method for centrifugal pump | |
Yu et al. | Weak fault feature extraction of rolling bearings using local mean decomposition-based multilayer hybrid denoising | |
CN110135492B (en) | Equipment fault diagnosis and abnormality detection method and system based on multiple Gaussian models | |
CN105760839A (en) | Bearing fault diagnosis method based on multi-feature manifold learning and support vector machine | |
CN113420691A (en) | Mixed domain characteristic bearing fault diagnosis method based on Pearson correlation coefficient | |
CN104748961A (en) | Gear fault diagnosis method based on SVD decomposition and noise reduction and correlation EEMD entropy features | |
CN110795843B (en) | Method and device for identifying faults of rolling bearing | |
CN115688018B (en) | Method for monitoring state and diagnosing faults of bearing under multiple working conditions | |
CN111413089A (en) | Gear fault diagnosis method based on combination of VMD entropy method and VPMCD | |
CN113176092B (en) | Motor bearing fault diagnosis method based on data fusion and improved experience wavelet transform | |
CN105258940B (en) | The standardization m ultiwavelet that mechanical breakdown is quantitatively extracted and multi-wavelet packets transform method | |
CN116226646B (en) | Method, system, equipment and medium for predicting health state and residual life of bearing | |
CN107543722A (en) | The Rolling Bearing Fault Character extracting method of dictionary learning is stacked based on depth | |
CN111238843A (en) | Fan health evaluation method based on rapid spectrum kurtosis analysis | |
CN109060347B (en) | Planetary gear fault identification method based on stacking noise-eliminating automatic encoder and gated cyclic unit neural network | |
CN106771598A (en) | A kind of Adaptive spectra kurtosis signal processing method | |
CN114136604B (en) | Rotary equipment fault diagnosis method and system based on improved sparse dictionary | |
CN108869145A (en) | Pumping plant unit diagnostic method based on compound characteristics index and depth limit learning machine | |
CN113390631A (en) | Fault diagnosis method for gearbox of diesel engine | |
CN114263621A (en) | Test method and system for diagnosing and simulating cavitation fault of centrifugal pump | |
CN113281047A (en) | Bearing inner and outer ring fault quantitative trend diagnosis method based on variable-scale Lempel-Ziv | |
Zhao et al. | A novel feature extraction approach for mechanical fault diagnosis based on ESAX and BoW model | |
CN114897016A (en) | Fan bearing fault intelligent diagnosis method based on multi-source frequency spectrum characteristics | |
CN114912480A (en) | Fault diagnosis method and system based on depth residual variation self-encoder |
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 |