CN111666877B - Rotating machinery weak fault detection method based on variable parameter wavelet manifold - Google Patents
Rotating machinery weak fault detection method based on variable parameter wavelet manifold Download PDFInfo
- Publication number
- CN111666877B CN111666877B CN202010507311.2A CN202010507311A CN111666877B CN 111666877 B CN111666877 B CN 111666877B CN 202010507311 A CN202010507311 A CN 202010507311A CN 111666877 B CN111666877 B CN 111666877B
- Authority
- CN
- China
- Prior art keywords
- wavelet
- fault
- manifold
- envelope
- envelopes
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
-
- 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
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The application discloses a rotating machinery weak fault detection method based on variable parameter wavelet manifold. The application discloses a rotating machinery weak fault detection method based on variable parameter wavelet manifold, which comprises the following steps: step (1), selecting a mother wavelet, and determining a scale range of signal analysis by giving 1 value combination of wavelet parameters; step (2), performing CWT processing on the signals, and calculating wavelet envelopes under all scales; and (3) selecting 1 wavelet envelope containing the most fault information from the whole scale band according to a given index, namely, a fault wavelet envelope. The application has the beneficial effects that: according to the application, fault wavelet envelopes of signals under different wavelet parameters are extracted, and pulse envelopes with stable structures are extracted from Gao Weibian parameter wavelet envelopes by utilizing excellent feature mining capability of manifold learning, so that in-band noise is removed, and effective detection of weak fault pulse envelopes is realized.
Description
Technical Field
The application relates to the field of fault diagnosis of rotary machinery, in particular to a weak fault detection method of rotary machinery based on variable-parameter wavelet manifold.
Background
The working condition of the rotary machine is directly related to the running performance and safety of the whole mechanical equipment, so that the rapid and accurate fault detection of the rotary machine has very important practical significance. The processing of vibration signals and the extraction of state characteristics are key technologies for the state evaluation and fault diagnosis of the rotary machine. Therefore, research into advanced signal processing methods to extract sensitive state features has become one of the most active fields of rotating machinery fault diagnosis disciplines. Among them, a common means of fault diagnosis of rotary machines is to successfully detect transient impulse response components present in vibration signals. However, due to the changing working environment and the possible mechanical faults, the vibration signal of the equipment tends to appear non-stationary and contain various frequency components, resulting in very weak mechanical faults reflected in the signal, which causes inconvenience to the fault diagnosis.
Continuous Wavelet Transform (CWT) is a method of processing non-stationary signals that enables local analysis of the time-frequency domain, i.e., multi-scale analysis of the signals by means of operations such as telescoping and panning. The idea of CWT is to inner-product transform the measurement signal and the wavelet function, thereby decomposing the measurement signal into wavelet coefficients. Wherein the scale of the wavelet function can be determined from the specific frequency components that need to be extracted from the measurement signal: the small scale can extract high-frequency quick change features in the signal, and the large scale can extract low-frequency slow change features, so that more multi-scale features reflecting the state and essence of the system can be obtained from the signal through scale transformation. Therefore, the wavelet analysis technology is widely applied to a plurality of nonlinear scientific fields such as signal processing, image processing, voice analysis, pattern recognition, fault diagnosis and the like. In rotating machinery fault detection, the result of analysis of the vibration signal by the CWT is a two-dimensional Time Scale Distribution (TSD), and how to extract the weak fault features from the analysis result quickly is a key to fault detection.
The wavelet coefficient modulus value corresponding to each scale in the TSD is the envelope of the signal under the scale, and is called wavelet envelope. The wavelet envelopes under different scales can reflect the envelope information of the signals on the corresponding scales, and meanwhile, the information under other scales is filtered, so that the method has the functions of signal demodulation and band-pass filtering. Therefore, in the application of CWT in fault detection of rotating machinery, in the prior art, a wavelet envelope containing the most fault information is selected for spectrum analysis to identify a fault characteristic frequency, which comprises the following specific steps: 1) Selecting a mother wavelet and parameters thereof, and implementing CWT on the signal in the signal frequency band range; 2) Calculating wavelet envelope under each scale; 3) Selecting fault envelope selection indexes such as kurtosis, smoothness factors and the like, calculating index values of each wavelet envelope, and selecting the wavelet envelope containing the most fault information, namely the fault wavelet envelope; 4) And carrying out frequency spectrum analysis on the fault wavelet envelope, identifying whether the fault characteristic frequency is included, and judging whether and what kind of faults exist.
The principle of CWT is that the information similar to wavelet is mined, however, the fault signals of different rotary machines have different forms, wavelet parameters need to be optimized to mine out the essential characteristics of the fault, whether the wavelet parameters are optimal or not is limited by the selection of optimization indexes, and different optimization indexes possibly obtain different optimal wavelet parameters. Therefore, the wavelet analysis result under a single parameter is not comprehensive enough, and useful fault information under other parameters can be ignored. In addition, the measured signal in a complex environment often contains severe noise, and the selected faulty wavelet envelope, while filtering out noise outside the scale, cannot filter out noise within the scale. Thus, the prior art has mainly two problems: a) The wavelet parameters need to be optimized, and useful fault information under other parameters is ignored by the optimized results; b) The extracted fault wavelet envelope is still affected by in-band noise at the scale where it is located.
Disclosure of Invention
The application aims to solve the technical problems of single wavelet parameters and difficult noise elimination in the existing fault wavelet envelope extraction by providing a rotating machinery weak fault detection method based on variable parameter wavelet manifold, and adopts a manifold learning method to carry out nonlinear fusion on fault wavelet envelopes under different wavelet parameters so as to comprehensively consider fault information under different wavelet parameters, reveal the essential structure of the fault envelopes and improve the accuracy of rotating machinery fault detection.
In order to solve the technical problems, the application provides a rotating machinery weak fault detection method based on variable parameter wavelet manifold, which comprises the following steps:
step (1), selecting a mother wavelet, and determining a scale range of signal analysis by giving 1 value combination of wavelet parameters;
step (2), carrying out continuous wavelet transformation processing on the signals, and calculating wavelet envelopes under all scales;
step (3), 1 wavelet envelope containing the most fault information is selected from the whole scale band according to a given index, namely the fault wavelet envelope;
step (4), changing the value of the wavelet parameters, repeating the above two steps for N times to obtain fault wavelet envelopes under N different wavelet parameters, and forming an N-dimensional variable parameter wavelet envelope;
step (5), fusing the N-dimensional variable parameter wavelet envelopes according to a given manifold learning method to obtain an internal envelope structure of the fault pulse, namely a variable parameter wavelet manifold;
and (6) performing spectrum analysis on the variable parameter wavelet manifold, and identifying fault characteristic frequency.
In one embodiment, the mother wavelet comprises a wavelet basis function that is amenable to continuous complex wavelet transforms.
In one embodiment, the wavelet basis function is specifically a complex Morlet wavelet, complex Gaussian wavelet, complex aromatic dense wavelet, frequency B spline curve wavelet, or the like.
In one embodiment, the given indicator comprises an indicator that is capable of selecting the faulty wavelet envelope from wavelet envelopes obtained from a continuous wavelet transform.
In one embodiment, the given index is specifically kurtosis, a smoothness factor, a root mean square, a base index, a sparsity index, a correlation coefficient, and combinations thereof.
In one embodiment, in the step (5), the given manifold learning method includes a method having a dimension reduction function.
In one embodiment, the given manifold learning method is specifically a local tangent space arrangement algorithm, an equidistant mapping algorithm, a local linear embedding algorithm, a laplace feature mapping algorithm, a local preserving projection algorithm, or the like.
Based on the same inventive concept, the present application also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, said processor implementing the steps of any one of the methods when executing said program.
Based on the same inventive concept, the present application also provides a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, implements the steps of any of the methods.
Based on the same inventive concept, the present application also provides a processor for running a program, wherein the program runs to execute the method of any one of the above.
The application has the beneficial effects that:
the application comprehensively considers the fault wavelet envelope of the signal under different wavelet parameters, extracts the pulse envelope with a stable structure from the Gao Weibian parameter wavelet envelope by utilizing the excellent characteristic mining capability of manifold learning, removes in-band noise and realizes the effective detection of weak fault pulse envelope. The technical method has at least the following advantages: without optimizing wavelet parameters, in-band noise can be removed, higher signal-to-noise ratios can be obtained, etc.
Drawings
Fig. 1 is a flowchart of a rotating machine weak fault detection method based on variable-parameter wavelet manifold, which is disclosed in the embodiment of the application.
Fig. 2 is a time domain waveform diagram of a gear failure vibration signal according to an embodiment of the present application.
Fig. 3 is a result of processing the signal shown in fig. 2 by using a conventional fault wavelet envelope extraction method.
Fig. 4 is a graph of the results obtained by processing the signal of fig. 2 using the techniques disclosed herein.
Detailed Description
The present application will be further described with reference to the accompanying drawings and specific examples, which are not intended to be limiting, so that those skilled in the art will better understand the application and practice it.
As known from the background art, in the existing CWT method, the wavelet envelope containing the most fault information under a certain parameter is selected for spectrum analysis, but the wavelet analysis result under a single parameter is not comprehensive enough, the useful fault information under other parameters can be ignored, and the in-band noise cannot be removed.
Therefore, the application discloses a rotating machinery weak fault detection method based on variable parameter wavelet manifold. The method adopts manifold learning method to carry out nonlinear fusion on fault wavelet envelopes under different wavelet parameters, and reveals the essential structure of the fault envelopes. Because manifold learning has excellent feature mining capability, the manifold learning can extract a pulse envelope with a stable structure from Gao Weibian parameter wavelet envelopes, remove in-band noise and realize effective detection of weak fault pulse envelopes.
According to the above summary and fig. 1, a method for detecting weak faults of a rotating machine based on a variable-parameter wavelet manifold specifically includes:
step 101: selecting a mother wavelet, and determining a scale range of signal analysis by giving 1 value combination of wavelet parameters;
the mother wavelet includes, but is not limited to, complex Morlet wavelet, complex Gaussian wavelet, complex aromatic dense wavelet, frequency B spline curve wavelet, etc. wavelet basis functions capable of continuous complex wavelet transformation. Different mother wavelets have different shapes, and the same mother wavelet is controlled by wavelet parameters. Wavelet parameters typically include center frequency and bandwidth.
Step 102: performing CWT processing on the signals, and calculating wavelet envelopes under various scales;
after CWT processing, wavelet envelopes of all scales corresponding to the whole frequency spectrum of the signal can be obtained, and envelope information of fault pulses is only distributed in a resonance scale band of a time scale plane, so that the method further processes the time scale plane.
Step 103: selecting 1 wavelet envelope containing the most fault information from the whole scale band according to a given index, namely, fault wavelet envelope;
the given index includes, but is not limited to, kurtosis, smoothing factor, root mean square, base index, sparse index, correlation coefficient, combinations thereof, and the like, which can select the index of the faulty wavelet envelope from the wavelet envelopes obtained from CWT. The fault wavelet envelope is located on the central scale of the resonance scale band with the highest index value.
Step 104: changing the value of the wavelet parameters, repeating the two steps for N times to obtain fault wavelet envelopes under N different wavelet parameters, and forming an N-dimensional variable parameter wavelet envelope;
the wavelet parameters of different mother wavelets have different values, and the values of the wavelet parameters are required to be standard by being capable of extracting fault wavelet envelopes. Because manifold learning does not require input data to have a very high dimension, the value of N is not too large.
Step 105: fusing the N-dimensional parametric wavelet envelopes according to a given manifold learning method to obtain an internal envelope structure of the fault pulse, namely the parametric wavelet manifold;
manifold learning is a nonlinear dimension reduction method that can be used to extract the inherent low-dimensional manifold structure embedded in high-dimensional data. In the N-dimensional parametric wavelet envelope, each dimension of data includes a fault pulse envelope and noise. Because wavelet parameters are different, the amplitudes of different fault wavelet envelopes at the same time point are different, but the fault pulse envelopes exist in each fault wavelet envelope, have a stable structure and can be regarded as a fault pulse envelope manifold with N-dimension data, the fault pulse envelope manifold can be reserved in manifold learning results, the noise forms in each fault wavelet envelope are different, have no stable structure and can be removed in the manifold learning results. Therefore, after the N-dimensional parametric wavelet envelopes are fused by a given manifold learning method, the obtained parametric wavelet manifold is a fault pulse envelope with high signal-to-noise ratio.
The given manifold learning method includes, but is not limited to, a local tangent space arrangement algorithm, an equidistant mapping algorithm, a local linear embedding algorithm, a Laplace feature mapping algorithm, a local preserving projection algorithm and other methods with dimension reduction functions.
Step 106: and carrying out frequency spectrum analysis on the variable parameter wavelet manifold, and identifying fault characteristic frequency.
Different faults of the rotating machine have different fault characteristic frequencies, and if a certain fault characteristic frequency exists in the frequency spectrum of the variable-parameter wavelet manifold, the rotating machine is detected to have corresponding faults.
In order to more clearly understand the technical scheme and the effect of the present application, the following detailed description is provided with reference to a specific embodiment.
Taking the case of gearbox vibration signal fault detection as an example, the experimental data is obtained from an automotive transmission gearbox having 5 forward gears and 1 reverse gear. The meshing frequency of the gearbox was 500Hz, and the rotational frequencies of the driving gear and the driven gear being tested were 20Hz and 18.5Hz, respectively. The accelerometer is arranged on the shell of the gear box and receives the signal with the sampling frequency of 2kHzVibration signals are collected. When the tested driving gear has abrasion fault, the characteristic frequency of the fault gear is f d ＝20Hz.
Referring to fig. 2, fig. 2 is a waveform diagram and a power spectrum diagram of a gearbox provided by an embodiment of the present application. The fault pulse of the gear in the waveform diagram is destroyed by noise, the power spectrogram can see that the frequency is highest at 500Hz, but the frequency is not modulated yet, and the resonance frequency band is in the range of 200 Hz-360 Hz.
The signal shown in fig. 2 is processed by adopting a traditional fault wavelet envelope extraction method, the mother wavelet is a complex Morlet wavelet, the central frequency and bandwidth parameter distribution of the mother wavelet is 2 and 1, and the root mean square index is adopted to select the fault wavelet envelope from the whole scale band. Fig. 3 shows the selected fault wavelet envelope and its spectrogram. From the waveform diagram, the fault pulse envelope can be observed, but is greatly affected by noise, which is unfavorable for the identification of fault features. The fault characteristic frequency f can be observed in the spectrogram d And its double harmonics.
The signal shown in figure 2 is processed by adopting the technology disclosed by the application, the mother wavelet is selected from complex Morlet wavelet, and the selection ranges of the center frequency and the bandwidth are respectively 0.5 and 3.5]And [1,3 ]]The combination of the wavelet parameters is 50 groups, the given fault wavelet envelope selection index is root mean square, and the given manifold learning method is a local tangent space arrangement algorithm. The results of the processing are given in fig. 4. The waveform graph in the graph is smoother, each pulse is obviously distinguished, the noise is smaller, and the fault characteristic frequency f can be observed in the spectrogram d And its double and triple harmonics. Therefore, compared with the traditional fault wavelet envelope extraction method, the technology disclosed by the application can extract the fault wavelet envelope with higher signal-to-noise ratio and richer harmonic wave, so that the weak fault of the gear can be detected from the vibration signal of the fault gearbox more accurately.
In summary, through carrying out continuous wavelet transformation under different wavelet parameters on the vibration signals of the gear box, then respectively selecting fault wavelet envelopes to form Gao Weibian parameter wavelet envelopes, and finally adopting manifold learning to fuse the obtained parameter-changing fault characteristics, certain noise components in the fault characteristics can be removed, and therefore weak fault of the gear box can be effectively detected. The method solves the problems that wavelet parameters need to be optimized, wavelet analysis results under single parameters are not comprehensive enough and in-band noise is difficult to remove in the prior art, can comprehensively consider fault characteristics under different wavelet parameters, has the advantages of enhancing fault wavelet envelope and removing in-band noise, and has important significance for effectively detecting weak faults of rotating machinery.
The above-described embodiments are merely preferred embodiments for fully explaining the present application, and the scope of the present application is not limited thereto. Equivalent substitutions and modifications will occur to those skilled in the art based on the present application, and are intended to be within the scope of the present application. The protection scope of the application is subject to the claims.
Claims (3)
1. The weak fault detection method of the rotary machine based on the variable-parameter wavelet manifold is characterized by comprising the following steps of:
step (1), selecting a mother wavelet, and determining a scale range of signal analysis by giving 1 value combination of wavelet parameters;
step (2), carrying out continuous wavelet transformation processing on the signals, and calculating wavelet envelopes under all scales;
step (3), 1 wavelet envelope containing the most fault information is selected from the whole scale band according to a given index, namely the fault wavelet envelope;
step (4), changing the value of the wavelet parameter, and repeating the two stepsNNext, obtainNFault wavelet envelope under different wavelet parameters to formNA vitamin-variable parametric wavelet envelope;
step (5), according to a given manifold learning methodNFusing the vitamin variable parameter wavelet envelopes to obtain an internal envelope structure of the fault pulse, namely a variable parameter wavelet manifold;
step (6), carrying out spectrum analysis on the variable parameter wavelet manifold, and identifying fault characteristic frequency;
the mother wavelet comprises a wavelet basis function capable of continuous complex wavelet transformation, the wavelet basis function is specifically a complex Morlet wavelet, a complex Gaussian wavelet, a complex aromatic dense wavelet or a frequency B spline curve wavelet, the given index comprises an index capable of selecting the fault wavelet envelope from wavelet envelopes obtained by continuous wavelet transformation, and the given index is specifically kurtosis, a smoothing factor, a root mean square, a base index, a sparse index, a correlation coefficient and a combination thereof;
in the step (5), the given manifold learning method includes a method with a dimension reduction function, and the given manifold learning method is specifically a local tangent space arrangement algorithm, an equidistant mapping algorithm, a local linear embedding algorithm, a laplace feature mapping algorithm or a local preserving projection algorithm.
2. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of claim 1 when executing the program.
3. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010507311.2A CN111666877B (en) | 2020-06-05 | 2020-06-05 | Rotating machinery weak fault detection method based on variable parameter wavelet manifold |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010507311.2A CN111666877B (en) | 2020-06-05 | 2020-06-05 | Rotating machinery weak fault detection method based on variable parameter wavelet manifold |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111666877A CN111666877A (en) | 2020-09-15 |
| CN111666877B true CN111666877B (en) | 2023-08-29 |
Family
ID=72386983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010507311.2A Active CN111666877B (en) | 2020-06-05 | 2020-06-05 | Rotating machinery weak fault detection method based on variable parameter wavelet manifold |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111666877B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102539150A (en) * | 2012-01-17 | 2012-07-04 | 电子科技大学 | Self-adaptive failure diagnosis method of rotary mechanical component based on continuous wavelet transformation |
| CN103018043A (en) * | 2012-11-16 | 2013-04-03 | 东南大学 | Fault diagnosis method of variable-speed bearing |
| CN108168891A (en) * | 2018-02-26 | 2018-06-15 | 成都昊铭科技有限公司 | The extracting method and equipment of rolling bearing Weak fault signal characteristic |
| CN108760316A (en) * | 2018-08-16 | 2018-11-06 | 苏州大学 | Information fusion method is joined in the change of variation mode decomposition |
-
2020
- 2020-06-05 CN CN202010507311.2A patent/CN111666877B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102539150A (en) * | 2012-01-17 | 2012-07-04 | 电子科技大学 | Self-adaptive failure diagnosis method of rotary mechanical component based on continuous wavelet transformation |
| CN103018043A (en) * | 2012-11-16 | 2013-04-03 | 东南大学 | Fault diagnosis method of variable-speed bearing |
| CN108168891A (en) * | 2018-02-26 | 2018-06-15 | 成都昊铭科技有限公司 | The extracting method and equipment of rolling bearing Weak fault signal characteristic |
| CN108760316A (en) * | 2018-08-16 | 2018-11-06 | 苏州大学 | Information fusion method is joined in the change of variation mode decomposition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111666877A (en) | 2020-09-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Periodic impulses extraction based on improved adaptive VMD and sparse code shrinkage denoising and its application in rotating machinery fault diagnosis | |
| Wang et al. | Fault diagnosis of diesel engine based on adaptive wavelet packets and EEMD-fractal dimension | |
| Rai et al. | A review on signal processing techniques utilized in the fault diagnosis of rolling element bearings | |
| Lei et al. | Fault diagnosis of rotating machinery using an improved HHT based on EEMD and sensitive IMFs | |
| CN111089726A (en) | Rolling bearing fault diagnosis method based on optimal dimension singular spectrum decomposition | |
| Sheng et al. | Applications in bearing fault diagnosis of an improved Kurtogram algorithm based on flexible frequency slice wavelet transform filter bank | |
| CN103792000A (en) | Method and device for detecting transient components in signal based on sparse representation | |
| CN111881594B (en) | Non-stationary signal state monitoring method and system for nuclear power equipment | |
| Jiang et al. | Study on nature of crossover phenomena with application to gearbox fault diagnosis | |
| CN105865794A (en) | Engine misfire fault diagnosis method based on short-time Fourier transformation and principal component analysis | |
| Zhang et al. | Improved local cepstrum and its applications for gearbox and rolling bearing fault detection | |
| Jiang et al. | A novel method for self-adaptive feature extraction using scaling crossover characteristics of signals and combining with LS-SVM for multi-fault diagnosis of gearbox | |
| CN107966287B (en) | Weak fault feature extraction method for self-adaptive electromechanical equipment | |
| Shi et al. | The VMD-scale space based hoyergram and its application in rolling bearing fault diagnosis | |
| CN114021616A (en) | Method, device and storage medium for diagnosing hydraulic motor fault | |
| CN113033304A (en) | Multi-resonance-band amplitude demodulation analysis method for overcoming frequency domain overlapping interference | |
| CN112945546A (en) | Accurate diagnosis method for complex fault of gear box | |
| Wang et al. | Information interval spectrum: a novel methodology for rolling-element bearing diagnosis | |
| Jiang et al. | Differential spectral amplitude modulation and its applications in rolling bearing fault diagnosis | |
| Li et al. | A novel method for diagnosing rolling bearing faults based on the frequency spectrum distribution of the modulation signal | |
| Alsalaet et al. | Bearing fault diagnosis using normalized diagnostic feature-gram and convolutional neural network | |
| CN112985809A (en) | Rolling bearing fault diagnosis method based on signal multi-dimensional fine image | |
| Jiang et al. | Bearing failure impulse enhancement method using multiple resonance band centre positioning and envelope integration | |
| CN111666877B (en) | Rotating machinery weak fault detection method based on variable parameter wavelet manifold | |
| Wang et al. | Research on gearbox composite fault diagnosis based on improved local mean decomposition |
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 |