CN116428529A - Heating pipeline leakage detection method based on AlexNet convolutional neural network - Google Patents
Heating pipeline leakage detection method based on AlexNet convolutional neural network Download PDFInfo
- Publication number
- CN116428529A CN116428529A CN202310222531.4A CN202310222531A CN116428529A CN 116428529 A CN116428529 A CN 116428529A CN 202310222531 A CN202310222531 A CN 202310222531A CN 116428529 A CN116428529 A CN 116428529A
- Authority
- CN
- China
- Prior art keywords
- neural network
- convolutional neural
- leakage
- data
- model
- 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
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 35
- 238000001514 detection method Methods 0.000 title claims abstract description 28
- 238000010438 heat treatment Methods 0.000 title claims abstract description 20
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000004458 analytical method Methods 0.000 claims abstract description 12
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 238000001914 filtration Methods 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 23
- 238000013528 artificial neural network Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 12
- 230000006872 improvement Effects 0.000 claims description 6
- 238000013519 translation Methods 0.000 claims description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 239000006185 dispersion Substances 0.000 claims description 3
- 210000005036 nerve Anatomy 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000012512 characterization method Methods 0.000 claims 1
- 230000002265 prevention Effects 0.000 abstract description 2
- 238000011160 research Methods 0.000 abstract description 2
- 238000003745 diagnosis Methods 0.000 description 11
- 238000012360 testing method Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D5/00—Protection or supervision of installations
- F17D5/02—Preventing, monitoring, or locating loss
- F17D5/06—Preventing, monitoring, or locating loss using electric or acoustic means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- 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
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2123/00—Data types
- G06F2123/02—Data types in the time domain, e.g. time-series data
-
- 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
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mechanical Engineering (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Examining Or Testing Airtightness (AREA)
Abstract
A heating pipeline leakage detection method based on an AlexNet convolutional neural network relates to a heating pipeline leakage detection method, which comprises the following steps: collecting normal working condition data, leakage working condition data and valve adjusting working condition data of a pipe network; step 2: filtering the acquired data through wavelet analysis; step 3: constructing a leakage signal AlexNet convolutional neural network identification model according to the extracted characteristic parameters; step 4: improving a one-dimensional leakage signal AlexNet convolutional neural network model; step 5: training and parameter optimization are carried out on the improved model of the leakage signal AlexNet convolutional neural network. The invention improves the AlexNet convolutional neural network, can improve the diagnostic capability of the convolutional neural network on leakage signals, reduces the false alarm rate, provides a basis for the prevention and remedy of heat supply pipeline leakage, and has great theoretical research value and wide application prospect.
Description
Technical Field
The invention relates to a heating pipeline leakage detection method, in particular to a heating pipeline leakage detection method based on an AlexNet convolutional neural network.
Background
With the expansion of the scale of the heating pipeline and the increase of the service life, the problem of leakage inevitably occurs, and pipeline corrosion and mechanical impact are important causes of leakage. In order to detect pipe leakage problems, numerous pipe leakage detection techniques have been developed and used, with increasing leakage failure detection efficiency using artificial intelligence techniques having been of great interest.
Existing techniques for actual pipe leak detection can be divided into two categories, hardware-based methods and software-based methods. The hardware detection-based methods such as an optical fiber detection method, an acoustic detection method, an infrared imaging method, a magnetic leakage detection method and the like are mature and stable in the use process, but the methods have high input cost or strong artificial dependency, and limit the use scale of the technology. Software-based methods are data analysis using collected data, with negative pressure wave detection and neural network methods being very representative. The negative pressure wave detection method has the characteristics of low cost, high sensitivity, accurate positioning precision and the like, but has the defect of false alarm caused by the interference of the adjusting working condition. The neural network method has a certain adaptability due to its strong data learning ability, and is therefore popular with students in recent years.
Identification by neural networks in combination with different types of signals has proven to be effective in leak diagnosis. The early BP neural network method can judge the leakage pipe section by combining a pipe network leakage working condition hydraulic calculation model of a spatial structure. However, the Bp neural network has the problems of slow convergence and easy occurrence of local extremum, although the recognition capability can be improved through an optimization algorithm. However, in the Bp neural network method, because the spatial topological structure of the pipe network is complex and the target pipe network trained by the neural network is fixed, the prediction generalization capability of a variable pipe network system is required to be improved. Deep learning algorithms are yet another milestone of neural network development, and various methods are continuously tried with the enhancement of algorithm model functionality. The LSTM model trained by using the physical principle as a guide shows excellent precision in testing, but the model trained based on a specific physical process cannot adapt to complex flow state changes in a pipe network. The deep confidence network can reach 96.87% of accuracy in pipe network fault diagnosis, but the deep confidence network is not easy to expand to different pipe networks due to the fact that the learning parameters are more. The acoustic signal images are classified by adopting a convolutional neural network method, and the diagnosis result shows that the convolutional network model is effective for diagnosing low leakage, but the acoustic signal is more easily interfered by the outside compared with the pressure signal. In general, neural networks exhibit good diagnostic capabilities in the classification and identification of signals, and the selection of appropriate neural networks and signal types has an important impact on the diagnosis of pipeline faults.
Disclosure of Invention
The invention aims to provide a heating pipeline leakage detection method based on an AlexNet convolutional neural network, and provides a diagnosis method for identifying pipeline leakage faults by combining wavelet denoising and improving the AlexNet convolutional neural network. The invention utilizes the obtained pressure data of the normal working condition, the leakage working condition and the valve adjusting working condition as learning samples to construct a leakage fault diagnosis model suitable for the heating pipe network, and identifies the negative pressure wave when the pipeline leaks, thereby judging whether the pipeline leaks or not and reducing the false alarm rate.
The invention aims at realizing the following technical scheme:
a heating pipe leak detection method based on an improved AlexNet convolutional neural network, the method comprising the steps of:
1) Collecting normal working condition data, leakage working condition data and valve adjusting working condition data of a pipe network;
2) The acquired data is subjected to filtering processing through wavelet analysis, interference of pipeline noise is reduced, and various characteristic parameters which can be used for representing leakage signals are extracted according to each section of data;
3) Constructing a leakage signal AlexNet convolutional neural network identification model according to the extracted characteristic parameters;
4) Aiming at the gradient dispersion problem of the model, the one-dimensional leakage signal AlexNet convolutional neural network model is improved to relieve the overfitting phenomenon, and the generalization capability of the model is improved;
5) Training and parameter optimization are carried out on the improved model of the leakage signal AlexNet convolutional neural network.
According to the heating pipeline leakage detection method based on the improved AlexNet convolutional neural network, the pipeline leakage data acquired in the step 1 are 5KHz in sampling rate, and the fixed duration of each section of acoustic signal is 5s.
The heating pipeline leakage detection method based on the improved AlexNet convolutional neural network comprises the following steps of:
step 2.1: the data denoising of wavelet analysis is an analysis method of time domain and frequency domain, the wavelet coefficient can be obtained by convolution operation of a wavelet basis function and an original signal, which is a function cluster generated by translating and stretching operation of a mother wavelet, and the mathematical expression is as follows:
wherein m is a scale factor, and the function of m is to stretch the mother wavelet function; n is a parameter of the translation factor reflecting translation; psi (x) is the mother wavelet;
step 2.2: the mathematical expression of discretizing the wavelet is as follows:
step 2.3: extracting time domain features of the leakage signal, including average value, variance, energy, average amplitude, root mean square, square root amplitude, effective value and effective value entropy;
step 2.4: extracting the shape characteristics of the leakage signal, including peak value coefficient, shape parameter, skewness parameter, pulse factor, valley factor, kurtosis and kurtosis factor;
step 2.5: extracting frequency domain characteristics of the leakage signal, including peak frequency, bandwidth and center frequency of the leakage signal.
The step 4 comprises the step of utilizing the AlexNet convolutional neural network to improve a heating pipeline leakage detection model, wherein the improvement of the heating pipeline leakage detection model comprises the following steps of:
step 4.1: the convolution kernel adopts a one-dimensional convolution kernel, and the original convolution kernel 11 multiplied by 11 in the first convolution layer is adjusted to be 1 multiplied by 7; not only is the dimension adjusted, but also the width of the convolution kernel is reduced, and finer features can be extracted through the filtering of the small convolution kernel;
step 4.2: the LRN layer is deleted, and because the parameters of the LRN layer need to be cross-validated, the improvement of the model is also very limited in practice, and training time is increased;
step 4.3: the number of neurons is reduced on the full-connection layer according to the complexity of the data, the first full-connection layer is adjusted to 1000 from 4096 nerve units of the original structure, and a large number of redundant weights are reduced;
step 4.4: the weight parameter optimization is carried out by adopting Adam, the algorithm adopts the principle of self-adaption of learning rate, compared with a random gradient descent algorithm with fixed learning rate, the weight of the neural network can be iteratively updated based on training data, the convergence speed of the neural network is accelerated, and the updating of the super-parameters is not influenced by gradient telescopic transformation.
The invention has the advantages and effects that:
the invention provides a diagnosis method capable of identifying pipeline leakage faults by combining wavelet analysis denoising and improving AlexNet convolutional neural network technology in the process of detecting the pipeline leakage, eliminates a large amount of noise signals, ensures that the wavelet analysis denoised signals have better fitting degree with the original signals, improves the AlexNet convolutional neural network, can improve the diagnosis capability of the convolutional neural network on the leakage signals, reduces false alarm rate, provides a basis for the prevention and remedy of the pipeline leakage, and has great theoretical research value and wide application prospect.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, while it is apparent that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a diagnostic flow chart of the present invention;
FIG. 2 is a block diagram of a conventional AlexNet convolutional neural network of the present invention;
FIG. 3 is a graph of the loss function of the present invention;
FIG. 4 is a graph of accuracy of model identification in accordance with the present invention.
Detailed Description
The following will clearly and completely describe the technical solutions in the embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The specific diagnosis flow of the invention is shown in fig. 1, and the invention collects 1500 sets of data altogether, wherein 500 sets of normal working condition data, 500 sets of leakage working condition data and 500 sets of valve adjusting working condition data. 400 sets of data under each working condition are selected as training sets, and 100 sets of data are selected as test sets. The length of data acquisition is 1500, the sampling rate is 5KHz, and the fixed duration of each section of acoustic signal is 5s. The leakage range of the leakage points is 0.5-15.4%, the range from low leakage to high leakage is covered, and the effectiveness of the neural network model in each leakage interval can be effectively verified.
Before the acquired data is input into a convolutional network model, wavelet analysis denoising is needed, and before a network is constructed, preprocessing operation is needed for the input picture data set. Noise interference and small pressure fluctuations in the pipeline can produce a large amount of invalid characteristic information, which results in the inability of the convolutional neural network to converge. Wavelet denoising is a key element in signal recognition. And then loading the processed information into a convolutional network model for training, and classifying and identifying various working conditions.
Extracting characteristic parameters according to the data of each section, and inputting the characteristic parameters into a trained neural network; the extracted characteristic parameters comprise time domain characteristics: average, variance, energy, average amplitude, root mean square, square root amplitude, peak value coefficient, shape parameter, skewness parameter, pulse factor, valley degree factor, kurtosis factor.
The data after the wavelet analysis and noise reduction processing is input into an AlexNet convolutional neural network structure as shown in fig. 2. In order to solve the gradient dispersion problem generated by deeper models, the overfitting phenomenon is reduced, the generalization capability of the models is increased, and the AlexNet convolutional neural network model is improved as follows:
the convolution kernel adopts a one-dimensional convolution kernel, and the original convolution kernel 11×11 in the first convolution layer is adjusted to be 1×7. After the dimension and width of the convolution kernel are adjusted, finer features can be extracted through filtering of the small convolution kernel;
the LRN layer is deleted, and because the parameters of the LRN layer need to be cross-validated, the improvement of the model is also very limited in practice, and training time is increased;
the number of neurons is reduced on the full-connection layer according to the complexity of the data, the first full-connection layer is adjusted to 1000 from 4096 nerve units of the original structure, and a large number of redundant weights are reduced;
the weight parameter optimization is carried out by adopting Adam, the algorithm adopts the principle of self-adaption of learning rate, compared with a random gradient descent algorithm with fixed learning rate, the weight of the neural network can be iteratively updated based on training data, the convergence speed of the neural network is accelerated, and the updating of the super-parameters is not influenced by gradient telescopic transformation.
The modified alexent parameters are detailed in the table below.
After model improvement, the preprocessed data set needs to be input into the network for training. In order to observe the performance of the model, the change of the loss function and the accuracy with the number of iterations needs to be analyzed, see fig. 3 and fig. 4 for details. When the training times are 200, the accuracy of the training set and the testing set starts to be stable. The loss function of the model training set fluctuates less and eventually tends to be stable. The robustness of the model on the training set is good, but relatively large oscillation occurs to the loss function in the test set, which is related to the sample capacity of the test set, but the recognition accuracy of the model can still be maintained above 96%, and the model has strong diagnosis capability. The data processed by the normal working condition and the leakage working condition can be basically attached to the original data through training, and the processed invalid noise point can be well restrained and the main characteristics of the original signal can be reflected. The method aims at improving the structure of the AlexNet convolutional neural network heat supply pipeline leakage detection model aiming at one-dimensional data, establishes a leakage fault diagnosis and identification model, has an average identification rate of 98.39% for various working conditions, has a highest identification rate of 99.3%, and can effectively diagnose the leakage working conditions of the pipeline.
The above-described embodiments are intended to illustrate the present invention, not to limit it, and any modifications and variations made to the present invention within the object of the present invention and the scope of the claims fall within the scope of the present invention.
Claims (4)
1. A heating pipe leak detection method based on an improved AlexNet convolutional neural network, the method comprising the steps of:
1) Collecting normal working condition data, leakage working condition data and valve adjusting working condition data of a pipe network;
2) The acquired data is subjected to filtering processing through wavelet analysis, interference of pipeline noise is reduced, and various characteristic parameters which can be used for representing leakage signals are extracted according to each section of data;
3) Constructing a leakage signal AlexNet convolutional neural network identification model according to the extracted characteristic parameters;
4) Aiming at the gradient dispersion problem of the model, the one-dimensional leakage signal AlexNet convolutional neural network model is improved to relieve the overfitting phenomenon, and the generalization capability of the model is improved;
5) Training and parameter optimization are carried out on the improved model of the leakage signal AlexNet convolutional neural network.
2. The heating pipeline leakage detection method based on the improved AlexNet convolutional neural network according to claim 1, wherein the pipeline leakage data collected in the step 1 are sampled at a rate of 5KHz, and the fixed duration of each section of acoustic signal is 5s.
3. The heating pipe leakage detection method based on the improved AlexNet convolutional neural network according to claim 1, wherein the wavelet analysis is performed on the pressure and flow data in the step 2, and the characteristic parameters for the leakage signal characterization are extracted, and the method comprises the following steps:
step 2.1: the data denoising of wavelet analysis is an analysis method of time domain and frequency domain, the wavelet coefficient can be obtained by convolution operation of a wavelet basis function and an original signal, which is a function cluster generated by translating and stretching operation of a mother wavelet, and the mathematical expression is as follows:
wherein m is a scale factor, and the function of m is to stretch the mother wavelet function; n is a parameter of the translation factor reflecting translation; psi (x) is the mother wavelet;
step 2.2: the mathematical expression of discretizing the wavelet is as follows:
step 2.3: extracting time domain features of the leakage signal, including average value, variance, energy, average amplitude, root mean square, square root amplitude, effective value and effective value entropy;
step 2.4: extracting the shape characteristics of the leakage signal, including peak value coefficient, shape parameter, skewness parameter, pulse factor, valley factor, kurtosis and kurtosis factor;
step 2.5: extracting frequency domain characteristics of the leakage signal, including peak frequency, bandwidth and center frequency of the leakage signal.
4. The heating pipe leakage detection method based on the AlexNet convolutional neural network according to claim 1, wherein the step 4 comprises the steps of improving a heating pipe leakage detection model by using the AlexNet convolutional neural network, and the improvement of the heating pipe leakage detection model comprises the following steps:
step 4.1: the convolution kernel adopts a one-dimensional convolution kernel, and the original convolution kernel 11 multiplied by 11 in the first convolution layer is adjusted to be 1 multiplied by 7; not only is the dimension adjusted, but also the width of the convolution kernel is reduced, and finer features can be extracted through the filtering of the small convolution kernel;
step 4.2: the LRN layer is deleted, and because the parameters of the LRN layer need to be cross-validated, the improvement of the model is also very limited in practice, and training time is increased;
step 4.3: the number of neurons is reduced on the full-connection layer according to the complexity of the data, the first full-connection layer is adjusted to 1000 from 4096 nerve units of the original structure, and a large number of redundant weights are reduced;
step 4.4: the weight parameter optimization is carried out by adopting Adam, the algorithm adopts the principle of self-adaption of learning rate, compared with a random gradient descent algorithm with fixed learning rate, the weight of the neural network can be iteratively updated based on training data, the convergence speed of the neural network is accelerated, and the updating of the super-parameters is not influenced by gradient telescopic transformation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310222531.4A CN116428529A (en) | 2023-03-09 | 2023-03-09 | Heating pipeline leakage detection method based on AlexNet convolutional neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310222531.4A CN116428529A (en) | 2023-03-09 | 2023-03-09 | Heating pipeline leakage detection method based on AlexNet convolutional neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116428529A true CN116428529A (en) | 2023-07-14 |
Family
ID=87080500
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310222531.4A Pending CN116428529A (en) | 2023-03-09 | 2023-03-09 | Heating pipeline leakage detection method based on AlexNet convolutional neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116428529A (en) |
-
2023
- 2023-03-09 CN CN202310222531.4A patent/CN116428529A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111734961B (en) | Natural gas pipeline leakage detection method | |
Dong et al. | Denoising the optical fiber seismic data by using convolutional adversarial network based on loss balance | |
CN111238814B (en) | Rolling bearing fault diagnosis method based on short-time Hilbert transform | |
CN106500735B (en) | A kind of compressed sensing based FBG signal adaptive restorative procedure | |
CN110751073A (en) | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning | |
Wang et al. | Prediction model of natural gas pipeline crack evolution based on optimized DCNN-LSTM | |
Harmouche et al. | Long-term monitoring for leaks in water distribution networks using association rules mining | |
CN113076920B (en) | Intelligent fault diagnosis method based on asymmetric domain confrontation self-adaptive model | |
CN115824519B (en) | Comprehensive diagnosis method for valve leakage faults based on multi-sensor information fusion | |
CN114034481A (en) | Fault diagnosis system and method for rolling mill gearbox | |
CN116805061A (en) | Leakage event judging method based on optical fiber sensing | |
Zhang et al. | Complementary ensemble adaptive local iterative filtering and its application to rolling bearing fault diagnosis | |
Li et al. | Knowledge features enhanced intelligent fault detection with progressive adaptive sparse attention learning for high-power diesel engine | |
CN117633588A (en) | Pipeline leakage positioning method based on spectrum weighting and residual convolution neural network | |
CN116913316A (en) | Power transformer typical fault voiceprint diagnosis method based on Mosaic data enhancement | |
CN112183260A (en) | One-way valve fault diagnosis method based on total variation noise reduction and RQA | |
CN116428529A (en) | Heating pipeline leakage detection method based on AlexNet convolutional neural network | |
Li et al. | PSO-MCKD-MFFResnet based fault diagnosis algorithm for hydropower units | |
CN116753471A (en) | Water supply pipeline leakage multi-domain feature extraction and fusion identification method | |
CN113569989B (en) | TI-TSDCN model construction method for stage equipment fault diagnosis | |
Chen et al. | Pipeline leak AE signal denoising based on improved SSA-K-α index-VMD-MD | |
CN116541771A (en) | Unbalanced sample bearing fault diagnosis method based on multi-scale feature fusion | |
CN114528869A (en) | Fan blade damage identification method | |
CN112683533A (en) | Signal enhancement method and system for bearing fault diagnosis | |
Yang et al. | Improving long-term guided wave damage detection with measurement resampling |
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 |