CN114912481A - Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion - Google Patents
Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion Download PDFInfo
- Publication number
- CN114912481A CN114912481A CN202210418871.XA CN202210418871A CN114912481A CN 114912481 A CN114912481 A CN 114912481A CN 202210418871 A CN202210418871 A CN 202210418871A CN 114912481 A CN114912481 A CN 114912481A
- Authority
- CN
- China
- Prior art keywords
- fault
- time
- motor bearing
- frequency
- dimensional 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
Images
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
- 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
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- 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)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Signal Processing (AREA)
- Probability & Statistics with Applications (AREA)
- Acoustics & Sound (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion, which comprises the following steps of: s1: acquiring one-dimensional vibration signals of the motor bearing under various fault conditions to construct a motor bearing fault data set; s2: performing data processing on one-dimensional vibration signals of the motor bearing fault data set to obtain a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram corresponding to each one-dimensional vibration signal; s3: constructing a fault identification network, and training the fault identification network based on a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram of a one-dimensional vibration signal; s4: and acquiring a one-dimensional vibration signal of the motor bearing to be diagnosed, sending the signal into a trained fault recognition network, and acquiring the fault condition type of the motor bearing to be diagnosed. Compared with the prior art, the method has the advantages of high fault diagnosis accuracy, high efficiency and the like.
Description
Technical Field
The invention relates to the field of fault diagnosis, in particular to a motor bearing fault diagnosis method based on self-adaptive fusion of multiple time-frequency analysis.
Background
With the development of society, the requirements on the motor are higher and higher, and as an important part of the motor, the health state of the bearing greatly influences the normal operation of the motor. And traditional intelligent fault diagnosis technologies such as a Support Vector Machine (SVM), a PCA and the like. However, these methods have a limited accuracy for identifying faults in rolling bearings and require manual intervention. In addition, the prior method has the following problems: the rolling bearing fault diagnosis method based on the traditional convolution has the defects of gradient diffusion, parameter explosion, long training time and the like; 2) the problem of difficulty in wavelet base selection is faced when the wavelet transform is simply used as the input of the neural network; 3) traditional machine learning methods do not perform well when faced with large data sets.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a motor bearing fault diagnosis method based on self-adaptive fusion of various time-frequency analyses.
The purpose of the invention can be realized by the following technical scheme:
a motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion comprises the following steps:
s1: acquiring one-dimensional vibration signals of the motor bearing under various fault conditions to construct a motor bearing fault data set;
s2: performing data processing on one-dimensional vibration signals of the motor bearing fault data set to obtain a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram corresponding to each one-dimensional vibration signal;
s3: constructing a fault identification network, and training the fault identification network based on a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram of a one-dimensional vibration signal;
s4: and acquiring a one-dimensional vibration signal of the motor bearing to be diagnosed, sending the one-dimensional vibration signal into a trained fault recognition network, and acquiring the fault condition type of the motor bearing to be diagnosed.
Preferably, the fault identification network comprises a first processing module, a second processing module, a third processing module and a classification output module, wherein the first processing module is used for performing feature extraction on a wavelet time-frequency diagram, the second processing module is used for performing feature extraction on a HHT marginal spectrum, the third processing module is used for performing feature extraction on an STFT time-frequency diagram, and the output module is used for processing extraction results of the first processing module, the second processing module and the third processing module to obtain the fault condition type of the motor bearing.
Preferably, the first processing module, the second processing module and the third processing module are all multiple Convolution modules, and the multiple Convolution modules include a contribution Convolution layer, an inversion Convolution layer and a contribution Convolution layer that are connected in sequence.
Preferably, the classification output module comprises a splicing layer, a CBAM attention module, an FC layer, and a Softmax layer, which are connected in sequence.
Preferably, the step S2 specifically includes:
s21: performing wavelet transformation on the one-dimensional vibration signals of the motor bearings to obtain a wavelet time-frequency diagram;
s22: performing Hilbert-Huang transformation on the one-dimensional vibration signals of the motor bearings to obtain HHT marginal spectrums;
s23: and carrying out short-time Fourier transform on the one-dimensional vibration signals of the motor bearings to obtain an STFT time-frequency diagram.
Preferably, the wave basis of the wavelet transformation is cmor3-3, the scale of the wavelet transformation is 256, the one-dimensional vibration signal and the wave basis of cmor3-3 are sent to a cwt function to calculate a wavelet coefficient coefs, the scale sequence of the wavelet transformation is converted into an actual frequency sequence, and finally, a wavelet time-frequency graph of the one-dimensional vibration signal is obtained by combining the time sequence of the wavelet transformation.
Preferably, in the hilbert-yellow transform, the HHT marginal spectrum is obtained by performing empirical mode decomposition on the one-dimensional vibration signal and processing the one-dimensional vibration signal by using a natural mode function.
Preferably, the analysis function is used for processing the one-dimensional vibration signal in the short-time Fourier transform, and the one-dimensional vibration signal of the bearing is segmented and intercepted to obtain an STFT time-frequency diagram.
Preferably, the analysis function is a pectrogram function.
Preferably, the fault conditions include normal, inner ring fault, outer ring fault and rolling body fault.
Compared with the prior art, the invention has the following advantages:
1) the multiple Convolution module of the present invention replaces the conventional Convolution Convolution with Involution. Under the condition of ensuring high calculation efficiency, the defects of parameter explosion, long training time and the like caused by Convolation Convolution are overcome, self-adaptive modeling of a long-distance relation is realized in a neural network layer, the recognition accuracy of the neural network is increased, and the recognition accuracy and the recognition efficiency of faults are improved;
2) according to the invention, the acquired one-dimensional vibration signals are converted into a two-dimensional time-frequency graph through wavelet transformation, Hilbert-Huang transformation and short-time Fourier transformation respectively, the wavelet time-frequency graph, the STFT time-frequency graph and the HHT marginal spectrum are used as input together, self-adaptive feature fusion is carried out through the CBAM module, the advantages of three time-frequency analysis are retained, and the network model has higher fault recognition rate and anti-interference capability;
3) the wavelet time-frequency graph, the HHT marginal spectrum and the STFT time-frequency graph are respectively sent to a classification output module, a CBAM module is added into the classification output module, the model performance is greatly improved through a space attention mechanism and a channel attention mechanism, and the robustness of a network is enhanced;
4) the fault diagnosis method is based on the fault recognition network, can effectively diagnose the states of normal, inner ring fault, outer ring fault, rolling body fault and the like under the 0 load and 25% load of the motor bearing, meets the requirements of industrial production, has important significance for realizing the automatic diagnosis of the motor bearing fault, can greatly improve the fault recognition accuracy rate and reduce the model training time, has good generalization capability, and has better practical usefulness in the actual fault diagnosis of the rolling bearing.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic diagram of a fault identification network according to the present invention;
FIG. 3 is a schematic diagram of a wavelet time-frequency diagram of the present invention;
FIG. 4 is a schematic representation of the HHT margin spectrum of the present invention;
FIG. 5 is a schematic STFT time-frequency diagram of the present invention;
FIG. 6 is a diagram of an obfuscation matrix in an embodiment of the invention;
FIG. 7 is a t-sne classification visualization in an embodiment of the invention;
FIG. 8 is a t-sne classification visualization in an embodiment of the invention;
FIG. 9 is a graph of accuracy in an embodiment of the present invention;
fig. 10 is a graph of loss curves in an embodiment of the present invention.
The system comprises a first processing module, a second processing module, a third processing module and a classification output module, wherein the first processing module 1, the second processing module 2, the second processing module 3, the third processing module 4 and the classification output module.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. Note that the following description of the embodiments is merely a substantial example, and the present invention is not intended to be limited to the application or the use thereof, and is not limited to the following embodiments.
Examples
A motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion comprises the following steps as shown in figure 1:
s1: and acquiring one-dimensional vibration signals of the motor bearing under various fault conditions to construct a motor bearing fault data set.
In this embodiment, the fault conditions in this embodiment are classified into four types, i.e., an outer ring fault, an inner ring fault, a rolling element fault, and a normal condition. The bearing model number is 6206-2RS in the embodiment. 500 samples were taken for each class, for a total of 2000 samples. According to the following steps: 2: 1 into a training set, a validation set and a test set. The sampling point is 864, which contains two failure periods. Under the actual working condition, a one-dimensional vibration signal is acquired through a sensor arranged at the drive end of the motor.
S2: and performing data processing on the one-dimensional vibration signals of the motor bearing fault data set to obtain a corresponding wavelet time-frequency graph, HHT marginal spectrum and STFT time-frequency graph of each one-dimensional vibration signal, as shown in figures 3-5.
The step S2 specifically includes:
s21: performing wavelet transformation on the one-dimensional vibration signals of the motor bearings to obtain a wavelet time-frequency diagram;
s22: performing Hilbert-Huang transformation on the one-dimensional vibration signals of the motor bearings to obtain HHT marginal spectrums;
s23: and carrying out short-time Fourier transform on the one-dimensional vibration signals of the motor bearings to obtain an STFT time-frequency diagram.
Specifically, the wave basis of the wavelet transform in this embodiment is cmor3-3, the scale of the wavelet transform is 256, the one-dimensional vibration signal and the cmor3-3 wave basis are sent to the cwt function to calculate the wavelet coefficient coefs, the scale sequence of the wavelet transform is converted into an actual frequency sequence, and finally, the wavelet time-frequency diagram of the one-dimensional vibration signal is obtained by combining the time sequence of the wavelet transform, the scale sequence is obtained by calculation according to the preset scale size, and the time sequence is obtained by calculation according to the preset sampling frequency. And then the original vibration signals are respectively subjected to empirical mode decomposition and an inherent mode function to obtain HHT marginal spectrum. In short-time Fourier transform, the window length is 512, a spectral function is used as an analysis function, one-dimensional vibration signals of the bearing are segmented and intercepted, and finally an STFT time-frequency diagram is obtained.
S3: and constructing a fault recognition network, and training the fault recognition network based on a wavelet time-frequency graph, a HHT marginal spectrum and an STFT time-frequency graph of the one-dimensional vibration signal.
As shown in fig. 2, in this embodiment, the fault identification network includes a first processing module 1, a second processing module 2, a third processing module 3, and a classification output module 4, where the first processing module 1 is configured to perform feature extraction on a wavelet time-frequency graph, the second processing module 2 is configured to perform feature extraction on a HHT marginal spectrum, the third processing module 3 is configured to perform feature extraction on an STFT time-frequency graph, and the output module is configured to process extraction results of the first processing module 1, the second processing module 2, and the third processing module 3, and obtain a fault condition type of a motor bearing.
The first processing module 1, the second processing module 2 and the third processing module 3 are all multiple Convolution modules, and the multiple Convolution modules comprise a Convolation convolutional layer, an Involution convolutional layer and a Convolation convolutional layer which are sequentially connected. In this embodiment, specifically, the input of the multiple Convolution module is subjected to multiple different Convolution kernels to extract the local area features of the original image in different ways, so as to expand the number of channels of the input image. And performing Concat operation after further feature extraction on each pixel point of the feature map through inversion convolution, performing adaptive feature fusion through a CBAM module, and finally outputting a classification result through a Softmax layer. The classification output module 4 comprises a splicing layer, a CBAM attention module, an FC layer and a Softmax layer which are connected in sequence.
In this embodiment, the Maxpool maximum pooling layer, the batchnormalization layer, and the Relu activation function layer are all set after the convergence Convolution layer to prevent overfitting, and specifically, the parameters and the output characteristic of each layer of the fault identification network are as follows:
Layer | Output Shape |
Conv2d(5*5*9) | (60*60*9) |
Maxpool2d | (30*30*9) |
BatchNormlization | / |
Relu | / |
Involution(3*3*1) | (30*30*9) |
Conv2d(3*3*12) | (28*28*12) |
Maxpool2d | (14*14*12) |
BatchNormlization | / |
Relu | / |
CBAM | (14*14*36) |
Linear | (14*14*36)>84>8 |
|
8 |
in the embodiment of the method, the wavelet time-frequency diagram, the HHT marginal spectrum and the STFT time-frequency diagram of the one-dimensional vibration signals of the motor bearings in the data set are respectively sent to different corresponding processing modules of the fault recognition network, and the fault recognition network is trained based on a training set, a verification set and a test set. In this embodiment, after the training is completed, the calculation of the classification accuracy curve and the loss curve is shown in fig. 9 and 10. Finally, a confusion matrix and a t-sne visual map are obtained by using the test set. As shown in FIGS. 6 to 8.
In order to verify the capacity of the fault identification network facing to the actual working condition, a sensor is arranged at the driving end of a motor under the conditions of 1kHZ driving, 10kHz sampling frequency, 1420, 0 load and 25% load, and a one-dimensional vibration signal of a vibration bearing is acquired.
S4: and acquiring a one-dimensional vibration signal of the motor bearing to be diagnosed, sending the signal into a trained fault recognition network, and acquiring the fault condition type of the motor bearing to be diagnosed.
The above embodiments are merely examples and do not limit the scope of the present invention. These embodiments may be implemented in other various manners, and various omissions, substitutions, and changes may be made without departing from the technical spirit of the present invention.
Claims (10)
1. A motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion is characterized by comprising the following steps:
s1: acquiring one-dimensional vibration signals of the motor bearing under various fault conditions to construct a motor bearing fault data set;
s2: performing data processing on one-dimensional vibration signals of the motor bearing fault data set to obtain a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram corresponding to each one-dimensional vibration signal;
s3: constructing a fault identification network, and training the fault identification network based on a wavelet time-frequency diagram, a HHT marginal spectrum and an STFT time-frequency diagram of a one-dimensional vibration signal;
s4: and acquiring a one-dimensional vibration signal of the motor bearing to be diagnosed, sending the signal into a trained fault recognition network, and acquiring the fault condition type of the motor bearing to be diagnosed.
2. The method for diagnosing the fault of the motor bearing based on the multiple time-frequency analysis self-adaptive fusion is characterized in that the fault identification network comprises a first processing module, a second processing module, a third processing module and a classification output module, wherein the first processing module is used for carrying out feature extraction on a wavelet time-frequency graph, the second processing module is used for carrying out feature extraction on a HHT marginal spectrum, the third processing module is used for carrying out feature extraction on an STFT time-frequency graph, and the output module is used for processing extraction results of the first processing module, the second processing module and the third processing module to obtain the fault condition type of the motor bearing.
3. The method for diagnosing the fault of the motor bearing based on the multiple time-frequency analysis self-adaptive fusion of the claim 2 is characterized in that the first processing module, the second processing module and the third processing module are all multiple Convolution modules, and the multiple Convolution modules comprise a Convolution Convolution layer, an Involution Convolution layer and a Convolution Convolution layer which are sequentially connected.
4. The method for diagnosing the fault of the motor bearing based on the self-adaptive fusion of the time-frequency analysis of the multiple types of the motor bearings according to claim 2, wherein the classification output module comprises a splicing layer, a CBAM attention module, an FC layer and a Softmax layer which are sequentially connected.
5. The method for diagnosing the fault of the motor bearing based on the adaptive fusion of the time-frequency analyses as claimed in claim 1, wherein the step S2 specifically comprises:
s21: performing wavelet transformation on the one-dimensional vibration signals of the motor bearings to obtain a wavelet time-frequency diagram;
s22: performing Hilbert-Huang transformation on the one-dimensional vibration signals of the motor bearings to obtain HHT marginal spectrums;
s23: and carrying out short-time Fourier transform on the one-dimensional vibration signals of the motor bearings to obtain an STFT time-frequency diagram.
6. The method for diagnosing the fault of the motor bearing based on the multiple time-frequency analysis self-adaptive fusion as claimed in claim 5, wherein a cmor3-3 is selected as a wave base of the wavelet transformation, the scale of the wavelet transformation is 256, the one-dimensional vibration signal and the cmor3-3 wave base are sent to a cwt function to calculate a wavelet coefficient coefs, the scale sequence of the wavelet transformation is converted into an actual frequency sequence, and finally, a wavelet time-frequency graph of the one-dimensional vibration signal is obtained by combining the time sequence of the wavelet transformation.
7. The method for diagnosing the fault of the motor bearing based on the self-adaptive fusion of the time-frequency analyses as claimed in claim 5, wherein the Hilbert-Huang transform is used for decomposing a one-dimensional vibration signal in an empirical mode and processing the signal by a natural mode function to obtain a HHT marginal spectrum.
8. The method for diagnosing the fault of the motor bearing based on the self-adaptive fusion of the time-frequency analyses as claimed in claim 5, wherein the analysis function is utilized to process the one-dimensional vibration signals in the short-time Fourier transform, and the one-dimensional vibration signals of the bearing are segmented and intercepted to obtain an STFT time-frequency diagram.
9. The method for diagnosing the fault of the motor bearing based on the self-adaptive fusion of the time-frequency analyses of the multiple types according to claim 8, wherein the analysis function is a pectrogram function.
10. The method for diagnosing the faults of the motor bearing based on the self-adaptive fusion of the time-frequency analyses as claimed in claim 1, wherein the fault conditions comprise normal faults, inner ring faults, outer ring faults and rolling body faults.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210418871.XA CN114912481A (en) | 2022-04-20 | 2022-04-20 | Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210418871.XA CN114912481A (en) | 2022-04-20 | 2022-04-20 | Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114912481A true CN114912481A (en) | 2022-08-16 |
Family
ID=82765513
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210418871.XA Pending CN114912481A (en) | 2022-04-20 | 2022-04-20 | Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114912481A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115905853A (en) * | 2022-09-05 | 2023-04-04 | 同济大学 | Aero-engine rotor system fault diagnosis method and device based on deep learning |
-
2022
- 2022-04-20 CN CN202210418871.XA patent/CN114912481A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115905853A (en) * | 2022-09-05 | 2023-04-04 | 同济大学 | Aero-engine rotor system fault diagnosis method and device based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110186680B (en) | Intelligent fault diagnosis method for confrontation discriminant domain adaptive one-dimensional convolutional neural network | |
CN111914883B (en) | Spindle bearing state evaluation method and device based on deep fusion network | |
CN112417954A (en) | Bearing fault mode diagnosis method and system for small sample data set | |
CN112036301A (en) | Driving motor fault diagnosis model construction method based on intra-class feature transfer learning and multi-source information fusion | |
CN113780412A (en) | Fault diagnosis model training method and system and fault diagnosis model training method and system | |
CN111753891A (en) | Rolling bearing fault diagnosis method based on unsupervised feature learning | |
Vong et al. | Case-based expert system using wavelet packet transform and kernel-based feature manipulation for engine ignition system diagnosis | |
CN114912481A (en) | Motor bearing fault diagnosis method based on multiple time-frequency analysis self-adaptive fusion | |
CN112347910A (en) | Signal fingerprint identification method based on multi-mode deep learning | |
CN114441173B (en) | Rolling bearing fault diagnosis method based on improved depth residual error shrinkage network | |
CN115577237A (en) | Elevator fault diagnosis method based on one-dimensional convolutional neural network and meta-learning | |
CN114065809A (en) | Method and device for identifying abnormal sound of passenger car, electronic equipment and storage medium | |
CN116484176A (en) | Bearing fault diagnosis method, system and storage medium based on ultra-wavelet | |
CN111860246A (en) | Deep convolutional neural network-oriented data expansion method for heart sound signal classification | |
CN110222390B (en) | Gear crack identification method based on wavelet neural network | |
CN112345251B (en) | Mechanical intelligent fault diagnosis method based on signal resolution enhancement | |
CN110631827B (en) | Gearbox fault diagnosis method based on improved collaborative representation classification | |
CN113505654A (en) | Bearing health state identification method based on multi-view attention network | |
Chen et al. | Bearing fault diagnosis based on pooling weight multi-scale convolutional neural networks | |
CN113409213A (en) | Plunger pump fault signal time-frequency graph noise reduction enhancement method and system | |
CN113158769A (en) | CEEMDAN and FastICA-based electromechanical device bearing vibration signal denoising method | |
CN117556317A (en) | Rotary bearing fault diagnosis method based on time-frequency image | |
CN118051830B (en) | Perimeter security intrusion event identification method | |
CN113988138A (en) | Signal extraction and classification method based on deep learning | |
CN114048682B (en) | Rolling bearing acoustic emission intelligent diagnosis method based on fusion of optimized wavelet basis and multidimensional depth characteristics |
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 |