CN117189082A - Intelligent identification and prediction method for sand production information of gas well - Google Patents
Intelligent identification and prediction method for sand production information of gas well Download PDFInfo
- Publication number
- CN117189082A CN117189082A CN202311153876.5A CN202311153876A CN117189082A CN 117189082 A CN117189082 A CN 117189082A CN 202311153876 A CN202311153876 A CN 202311153876A CN 117189082 A CN117189082 A CN 117189082A
- Authority
- CN
- China
- Prior art keywords
- sand
- noise reduction
- vibration
- signal
- prediction
- 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
- 239000004576 sand Substances 0.000 title claims abstract description 120
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 36
- 230000009467 reduction Effects 0.000 claims abstract description 31
- 238000012544 monitoring process Methods 0.000 claims abstract description 20
- 238000004364 calculation method Methods 0.000 claims abstract description 16
- 230000004044 response Effects 0.000 claims abstract description 15
- 230000003044 adaptive effect Effects 0.000 claims abstract description 4
- 230000005514 two-phase flow Effects 0.000 claims abstract 5
- 230000005284 excitation Effects 0.000 claims abstract 2
- 238000001228 spectrum Methods 0.000 claims description 19
- 238000013527 convolutional neural network Methods 0.000 claims description 15
- 238000000354 decomposition reaction Methods 0.000 claims description 13
- 238000005457 optimization Methods 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 9
- 230000004913 activation Effects 0.000 claims description 8
- 230000001133 acceleration Effects 0.000 claims description 4
- 238000011176 pooling Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000001427 coherent effect Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 230000007774 longterm Effects 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 claims 1
- 238000011161 development Methods 0.000 abstract description 4
- 238000007781 pre-processing Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 6
- 230000005012 migration Effects 0.000 description 6
- 238000013508 migration Methods 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000008034 disappearance Effects 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000013145 classification model Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000016273 neuron death Effects 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 101000709114 Homo sapiens SAFB-like transcription modulator Proteins 0.000 description 1
- 102100032664 SAFB-like transcription modulator Human genes 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to the field of gas well development engineering, in particular to an intelligent identification and prediction method for sand production information of a gas well. The system mainly comprises a sand-out signal preprocessing module and a sand-out information intelligent recognition and prediction module. The sand-out signal preprocessing module is an EMD adaptive noise reduction algorithm optimized based on an improved wavelet packet threshold noise reduction algorithm; the sand output information intelligent recognition and prediction module is a CNN-LSTM model driven by a multi-scale triaxial vibration response sequence, integrates spatial dimension characteristics and time sequence characteristics of multi-scale flow information contained in a gas-sand two-phase flow excitation triaxial vibration signal, evaluates and integrates influence weights of vibration in different monitoring directions on sand output amount through an entropy weight method, and accordingly achieves accurate calculation and prediction of the sand output amount under gas turbulence disturbance. The intelligent identification and prediction method for the sand production information of the gas well has higher precision for the real-time prediction of the sand production amount of the gas well.
Description
Technical Field
The invention relates to the field of gas well development engineering, in particular to an intelligent identification and prediction method for sand production information of a gas well.
Background
Sand production monitoring is one of key steps for realizing sand production management in a whole life production cycle, and accurate identification and prediction of sand grain information in a complex multiphase flow system are problems to be solved in sand production monitoring. Accurate monitoring of sand particles under gas turbulence is one of the major challenges faced in gas well development. Because the problems of damage, blockage, sand burying and the like of production equipment caused by sand production greatly limit the safe and efficient development of oil gas, the method for realizing the real-time early warning and monitoring of the sand production information of the shaft has important guiding significance for timely adjusting production system, reducing cost and enhancing efficiency.
CN110344816a discloses a sand-out monitoring method for oil and gas wells based on distributed optical fiber sound monitoring, which monitors sand-out conditions of oil and gas wells by installing a distributed optical fiber sound monitoring device in the oil and gas well to be monitored. The method can monitor the sand discharge conditions of all production intervals in real time, but has the defect that the method can only qualitatively judge the sand discharge degree of each interval and cannot quantitatively predict the sand discharge amount.
CN111305814a discloses a method for monitoring underwater sand production of deep water oil and gas well, which is characterized in that the method can simply and rapidly distinguish noise signals from sand production signals by comparing collected signals with preset fluid patterns, and has the advantages that the filtering of the noise signals is limited by the precision of a preset pattern, and meanwhile, the extraction of the sand production signal characteristics is lack of self-adaptability. In addition, the existing oil-gas well sand production prediction method based on the acoustic and vibration monitoring method generally uses a single-axis sensor, so that the problem of information loss exists when the fine sand signal is monitored under the interference of strong fluid noise.
The invention aims to construct an intelligent identification and prediction method for sand production information of a gas well. By fusing a series of multifrequency multi-scale triaxial vibration signal analysis methods and CNN-LSTM, the real-time accurate prediction of the sand yield under the turbulence of gas is realized, and technical support is provided for the safe and efficient production of a sand-producing gas well. At present, no relevant report is found on the method for intelligently identifying and predicting the sand production information of the gas well based on the combination of multi-frequency multi-scale triaxial vibration signal analysis and CNN-LSTM.
Disclosure of Invention
The invention aims to solve the technical problem of overcoming the defects of the prior art and providing an intelligent recognition and prediction method suitable for sand signal characteristic analysis, extraction and sand output information under the interference of gas turbulence noise.
In order to achieve the above purpose, the invention adopts the following technical scheme:
an intelligent identification and prediction method for sand production information of a gas well comprises the following steps:
and acquiring a sand-out signal by using a high-frequency triaxial acceleration vibration sensor.
As further optimization of the invention, the wavelet packet threshold value noise reduction based on multifrequency coherent analysis and statistical feature fusion driving optimization is carried out on the sand-out vibration signal, and the wavelet packet noise reduction threshold value is manually set according to the frequency domain response range of the sand-out signal, so that fluid noise signals which are aliased with the sand-out signal are accurately filtered.
Further, EMD self-adaptive decomposition is carried out on the sand-out signal after wavelet packet threshold noise reduction.
As a further optimization of the invention, the two-dimensional HHT time spectrum of each IMF component is calculated and drawn, and different sand migration information and noise response information characterized by different IMF components are qualitatively determined.
As further optimization of the invention, the energy duty ratio, impact strength, information complexity and sequence stability index of each IMF component are calculated, and multi-scale reconstruction is carried out on the IMF components with similar fractal characteristics and response characteristics.
Specifically, IMF components characterizing the gas directly carrying sand migration information are reconstructed into a sequence of microscale.
Specifically, IMF components characterizing the remaining sand migration behavior information are reconstructed into a mesoscale sequence.
Specifically, IMF components characterized as having dynamic stable noise response characteristics are reconstructed as a macro-scale sequence.
And integrating the power spectrum of the microscale sequence and the mesoscale sequence along a time axis. If the vibration energy value of the microscale sequence is positively correlated with the gas speed, the integrated energy value of the mesoscale time sequence is almost unchanged or dynamically stable with the gas speed, then the accurate division of different scale sequences is indicated; otherwise, the multi-scale sequence is divided and reconstructed again.
And converting the microscale sequence into a two-dimensional gray image, and inputting the two-dimensional gray image into a CNN branch of the CNN-LSTM sand-out information identification and prediction module constructed by the invention.
As a further optimization of the present invention, the CNN branch is a deep pure convolutional neural network, which has advantages over existing mature CNN models in that it includes:
1) The 'leakage ReLU' type activation function is introduced to replace the common ReLU and ELU type activation functions, so that the problems of neuronal death and gradient disappearance caused by using the ReLU and other activation functions are effectively relieved.
2) The full-connection layer necessary for the traditional CNN classification model is abandoned, global average pooling operation is adopted to replace the full-connection layer, and model overfitting risks caused by high parameter quantity of the full-connection layer in the traditional CNN classification model are reduced.
Further, extracting the characteristics of the input microscale sequence through the CNN branch, identifying the grain size of sand corresponding to the microscale sequence, and finally obtaining a sand yield correction coefficient A corresponding to different grain sizes through automatic searching and matching.
As a further optimization of the present invention, the LSTM branches are a shallow SLTM network, which is advantageous in that it includes:
1) The Dropout regularization technology is introduced to optimize the LSTM, so that the strong dependency relationship among neurons is effectively reduced, and the problem that long-distance time step information is difficult to transfer is solved.
2) The shallow LSTM network structure greatly improves the operation efficiency of the model.
Further, inputting the mesoscale sequence into the LSTM branch, predicting to obtain a power spectrum of the sand-out signal, and integrating the predicted power spectrum along a time axis to obtain a sand-out vibration energy level Q i 。
As a further optimization of the invention, the influence weights S of different monitoring direction vibration energy on the sand yield are determined based on an entropy weight method i 。
As a further optimization of the invention, the invention provides a gas well sand production amount calculation model, which comprises the following steps:
wherein C is sand G is the sand yield.
The C is subjected to sand And comparing the value with a preset sand yield threshold value, and if the value exceeds the preset threshold value, giving an alarm.
Compared with the prior art, the invention has the advantages and positive effects that the invention comprises:
1) The equipment for detecting the sand signal is a non-implanted high-frequency triaxial acceleration vibration sensor, and has the advantages that (1) the non-implanted installation method avoids the contact between the sensor and the migration sand in the shaft, and prolongs the service life of the sensor; (2) The high-frequency triaxial acceleration sensor effectively solves the problem of sand information loss commonly existing in the process of using a uniaxial acoustic sensor and a vibration sensor.
2) According to the wavelet packet noise reduction method based on multi-frequency statistical characteristics and coherence analysis optimization, which is provided by the invention, the noise response range can be finely determined through analysis of the energy distribution, impact strength, signal to noise ratio and frequency coherence of each sub-band, so that the manual accurate setting of the noise reduction threshold is realized, and the problems of excessive noise reduction or insufficient noise reduction existing when the traditional wavelet packet threshold noise reduction method processes non-steady sand-out signals are effectively solved.
3) The EMD adaptive noise reduction algorithm based on optimized wavelet packet preprocessing provided by the invention realizes the fusion of the wavelet packet high-frequency fine decomposition characteristic and the EMD adaptation, effectively solves the problem of excessive aliasing between a high frequency band and broadband background noise when the EMD processes a non-stationary sand-out signal, and improves the accuracy of extracting the sand-out signal characteristics.
4) According to the wellbore sand output quantity calculation model corrected based on the entropy weight method, the influence proportion of vibration energy in different monitoring directions on sand output quantity calculation is estimated by the entropy weight method, the problems of difference and contribution degree loss of vibration energy in different monitoring directions caused by using a mean value method and the like are effectively solved, and the sand output quantity calculation accuracy is improved.
5) The CNN module of the intelligent identification and prediction model for the sand production information of the CNN-LSTM constructed by the invention uses a 'leakage ReLU' type activation function, and the activation function effectively relieves the problems of neuronal death and gradient disappearance caused by using the activation functions such as ReLU and the like. Meanwhile, the necessary full-connection layer of the traditional CNN model is abandoned, global average pooling operation is adopted to replace the full-connection layer, the overfitting risk of the model due to the high parameter number of the full-connection layer is reduced, and the learning capacity of the model for large-scale learning tasks is improved;
6) The LSTM module of the CNN-LSTM sand-out information intelligent identification and prediction model constructed by the invention is a shallow LSTM prediction model optimized based on a Dropout regularization technology, and the model reduces strong dependency among neurons by randomly discarding the neurons in the training process, thereby reducing overfitting on a training set. Meanwhile, the problem that information of a long-distance time step is difficult to transfer is effectively solved, the problem of gradient disappearance is solved to a certain extent, and the long-term dependence of the sand signal power spectrum sequence can be better captured.
Drawings
FIG. 1 is a flow chart of an intelligent identification and prediction method for sand production information of a gas well;
FIG. 2 is a structural block diagram of an intelligent identification and prediction method for sand production information of a gas well, which is provided by the invention;
Detailed Description
The present invention will be specifically described below by way of exemplary embodiments. It is to be understood, however, that elements, structures, and features of one embodiment may be beneficially incorporated in other embodiments without further recitation;
in the description of the present invention, it should be noted that the positional or positional relationship indicated by the terms such as "inner", "outer", "upper", "lower", "front", "rear", etc. are based on the positional relationship shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or element in question must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Referring to fig. 1, the invention provides an intelligent identification and prediction method for sand production information of a gas well, which is integrated in modules 1-9 shown in fig. 2 and comprises the following steps:
s1: initializing and self-checking each module of the triaxial vibration response driven gas well sand production information identification and prediction method;
s2: setting acquisition parameters such as a signal sampling rate and the like through an acquisition parameter setting module 2 shown in fig. 2;
s3: the sand-out vibration signal is collected and stored through the vibration signal collection module 3 shown in fig. 2;
s4: the wavelet packet decomposition is carried out on the sand-out vibration signal through the wavelet packet noise reduction module 4 shown in fig. 2;
further, SNR, energy, kurtosis and coherence coefficient of each subband obtained by wavelet packet decomposition are calculated, and the SNR calculation formula is as follows:
wherein E is A For the energy value of the sand-out signal E B Is the energy value of the gas turbulence noise signal.
And calculating the coherence coefficient, the energy and the kurtosis of each sub-band obtained by decomposing the wavelet packet, and determining the sub-band representing the noise information by combining the response range of the characteristic frequency domain of the sand signal.
And manually setting the wavelet packet coefficient of the sub-frequency band representing the noise response information to 0 to realize the self-defined noise reduction threshold setting of the wavelet packet, and then reconstructing all the sub-frequency bands to obtain the wavelet packet noise reduction signal.
Calculating the energy ratio, the waveform similarity coefficient and the root mean square error of the original triaxial sand-out signal and the wavelet packet noise reduction signal, and comprehensively evaluating the noise reduction effect; if the energy ratio is greater than 0.85, the waveform similarity coefficient is greater than 0.9 and the root mean square error is less than 0.05, the noise reduction effect is good; otherwise, the number of decomposition layers, wavelet packet base, etc. are reset and wavelet packet decomposition is performed.
S5: the EMD multi-scale decomposition and reconstruction module 5 shown in fig. 2 performs EMD adaptive decomposition on the wavelet packet noise reduction signal, and calculates and draws the HHT time spectrum of each IMF component.
The Hurst of each IMF component is calculated as follows:
wherein f (t) is the Kth subsequence of length N, S k Is the standard deviation of the subsequence.
Information entropy, peak-to-peak value, and energy duty ratio of each IMF component are calculated.
The IMF component of the sand signal and the noise signal is characterized based on the peak-to-peak and energy duty cycle discrimination.
Based on information entropy (characterizing complexity of each IMF sequence) and Hurst characteristics (characterizing fractal characteristics of each IMF sequence), IMF components having the same characteristics are reconstructed. Specifically, reconstructing IMF1 components characterizing gas-carried sand migration directly into a microscale sequence; reconstructing IMF2-IMF3 components characterizing other sand migration forms into a mesoscale sequence; reconstructing IMF4-IMF10 components representing noise response information into a macro-scale sequence;
s6: converting the micro-scale vibration sequence into a two-dimensional gray level image by the intelligent identification module 6 of the grain size of the CNN sand grains shown in the figure 2, inputting the two-dimensional gray level image into the CNN branches, identifying the grain size of the sand grains represented by the scale sequence signal, and then automatically searching to obtain a sand yield correction coefficient A corresponding to the sand grains with the grain size;
s7: and converting the mesoscale vibration response sequences of the X, Y, Z monitoring directions into corresponding power spectrum sequences, and obtaining a predicted power spectrum of the sand production signal through the LSTM power spectrum sequence prediction module 7 shown in figure 2. Integrating the predicted power spectrum along a time axis to obtain sand-out vibration energy levels Q corresponding to three monitoring directions i 。
S8: the influence weight S of the vibration energy of the X, Y, Z monitoring directions on the sand yield calculation is calculated through the entropy weight sand yield calculation module 8 shown in fig. 2 i 。
S9: the sand output correction coefficient A and the vibration energy levels Q in three monitoring directions are corrected i Weights S of three monitoring directions i The sand output prediction and alarm module 9 shown in fig. 2 is input to predict and obtain the sand output C at the moment sand . The predicted sand value C sand Automatically comparing with a preset sand-out threshold value, if C sand If the sand quantity is larger than the preset threshold value, the system gives out excessive sand discharge early warning and simultaneously outputs a predicted sand quantity value C sand 。
Claims (6)
1. The intelligent identification and prediction method for the sand production information of the gas well is characterized by comprising the following steps of:
sand-out signal acquisition unit: the system comprises an acquisition parameter setting module and a vibration signal acquisition module, wherein the acquisition parameter setting module is used for acquiring a sand-out vibration signal excited by a gas-sand two-phase flow and comprises a high-frequency triaxial acceleration vibration sensor;
sand-out signal noise reduction unit: the system comprises a wavelet packet noise reduction module and an EMD multi-scale decomposition and reconstruction module, wherein the wavelet packet noise reduction module and the EMD multi-scale decomposition and reconstruction module are used for filtering noise signals contained in sand signals, and the system comprises a self-defined wavelet packet threshold noise reduction method based on the fusion optimization of statistical characteristics and multi-frequency coherent analysis and an EMD self-adaptive noise reduction algorithm driven and optimized by the self-defined wavelet packet threshold noise reduction method;
CNN-based sand grain size intelligent recognition unit: the deep convolutional neural network integrating global average pooling operation and a 'leakage ReLU' type activation function is used for intelligently identifying the grain size of sand grains in gas-sand two-phase flow, and automatically searching and determining a sand yield correction coefficient A corresponding to sand grains with different grain sizes according to the identified grain size of the sand grains;
sand-out signal power spectrum prediction unit based on LSTM: the method comprises the following steps of predicting the power spectrum of a real-time sand production signal under the drive of a mesoscale vibration response sequence excited by a gas-sand two-phase flow, and integrating the predicted power spectrum to obtain vibration energy levels in different monitoring directions, wherein the shallow LSTM neural network is optimized based on a Dropout regularization technology;
the sand output calculation and prediction unit of the gas well: the sand output calculation model based on the entropy weight method correction is constructed by inputting the sand output correction coefficient obtained by the sand grain size intelligent identification module and the vibration energy levels in different monitoring directions obtained by the sand output signal power spectrum prediction module into the constructed sand output calculation model, so that the accurate calculation and prediction of the sand output of the gas well can be realized.
2. The sand out signal noise reduction unit of claim 1, comprising a custom wavelet packet threshold noise reduction method that drives optimization based on statistical features and multi-frequency coherence analysis:
s1: determining a frequency domain response range of a gas-sand two-phase flow excitation sand vibration signal based on FFT;
s2: carrying out multi-layer wavelet packet decomposition on the collected sand-out vibration signals, and calculating the energy duty ratio, the signal-to-noise ratio, the kurtosis and the multi-frequency coherence coefficient of each sub-band;
s3: determining sub-bands representing noise signals through the S1 and the S2, manually setting wavelet packet coefficients of the sub-bands to 0, and then reconstructing each sub-band;
s4: and calculating the energy ratio, the waveform similarity coefficient and the root mean square error of the reconstructed signal and the original signal. If the energy ratio is greater than 0.85, the waveform similarity coefficient is greater than 0.9 and the root mean square error is less than 0.05, the noise reduction effect is good; otherwise, the number of decomposition layers, wavelet packet basis, etc. are reset and S2-S4 are performed.
3. The sand out signal noise reduction unit of claim 1, comprising an EMD adaptive noise reduction method optimized based on a custom wavelet packet threshold noise reduction method, comprising the steps of:
s1: inputting the sand-out vibration signal after noise reduction of the wavelet packet into an EMD for self-adaptive decomposition;
s2: calculating a two-dimensional HHT time spectrum of each IMF component, and qualitatively determining vibration information sources mainly characterized by different IMF components;
s3: calculating energy, hurst, information entropy and the like of each IMF component, and quantitatively evaluating vibration information sources represented by different IMF components;
s4, performing S4; and based on the S1, S2 and S3 analysis, performing multi-scale sequence reconstruction on IMF components representing the same vibration information source, and removing the scale sequence representing noise information.
4. The intelligent recognition unit for the sand grain size based on the CNN is characterized by comprising a deep pure convolutional neural network, wherein a global average pooling operation is adopted to replace a full-connection layer while a 'leak ReLU' type activation function is used, so that the calculation efficiency and generalization of a model are effectively improved, the sand grain size is recognized under the driving of a microscale sand-out response sequence, and then the sand-out correction coefficient A corresponding to sand grains with different grain sizes is automatically searched.
5. The LSTM based sand out signal power spectrum prediction unit of claim 1 comprising a Dropout regularization optimized shallow LSTM prediction model effective to mitigate over-fitting and long distance timesThe problem that the information of the interval is difficult to transfer is conducive to better capturing the long-term time dependence of the sand signal power spectrum sequence, the sand signal power spectrum is predicted under the drive of the mesoscale sand production response sequence, and the power spectrum is integrated along the time axis to obtain the corresponding sand production vibration energy Q i 。
6. The sand production calculation and prediction unit for a gas well according to claim 1, wherein the influence weight S of the vibration energy detected in different monitoring directions on the sand production calculation is calculated by adopting an entropy weight method i And the sand output correction coefficient A and the vibration energy Q are combined with the sand output correction coefficient i Inputting the constructed sand output quantity calculation model to obtain a sand output quantity predicted value C sand And C is combined with sand Automatically comparing with a preset sand yield threshold value, if C sand If the preset sand output threshold value is exceeded, the system alarms and outputs a predicted sand output value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311153876.5A CN117189082A (en) | 2023-09-08 | 2023-09-08 | Intelligent identification and prediction method for sand production information of gas well |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311153876.5A CN117189082A (en) | 2023-09-08 | 2023-09-08 | Intelligent identification and prediction method for sand production information of gas well |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117189082A true CN117189082A (en) | 2023-12-08 |
Family
ID=88982955
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311153876.5A Pending CN117189082A (en) | 2023-09-08 | 2023-09-08 | Intelligent identification and prediction method for sand production information of gas well |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117189082A (en) |
-
2023
- 2023-09-08 CN CN202311153876.5A patent/CN117189082A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110751073A (en) | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning | |
| CN108805059B (en) | Sparse regularization filtering and self-adaptive sparse decomposition gearbox fault diagnosis method | |
| CN110940539A (en) | Machine equipment fault diagnosis method based on artificial experience and voice recognition | |
| CN112036042B (en) | Power equipment abnormality detection method and system based on variational modal decomposition | |
| CN105834835A (en) | Method for monitoring tool wear on line based on multiscale principal component analysis | |
| CN111170103B (en) | Equipment fault identification method | |
| CN105626502B (en) | Plunger pump health evaluating method based on wavelet packet and laplacian eigenmaps | |
| EP2370837A1 (en) | Method and system for real time production management and reservoir characterization | |
| CN110375983B (en) | Valve fault real-time diagnosis system and method based on time series analysis | |
| CN103982378A (en) | Method for diagnosing surface icing faults of wind power generator blade of power system based on machine visual images | |
| CN109443766A (en) | A kind of heavy-duty vehicle gearbox gear Safety Analysis Method | |
| CN108757426B (en) | Fault diagnosis method for water injection plunger pump in petroleum mine field | |
| CN115014617B (en) | Cable-stayed bridge cable force synchronous monitoring method based on ground radar | |
| CN114707555A (en) | Bank collapse precursor infrasonic wave signal noise reduction processing method, characteristic identification method and device | |
| CN111204842A (en) | Method, device and system for realizing ultrafiltration membrane pollution evaluation through neural network | |
| CN117370879A (en) | Real-time online fault diagnosis method and system for wind turbine gearbox | |
| CN117782198B (en) | Highway electromechanical equipment operation monitoring method and system based on cloud edge architecture | |
| CN114861349A (en) | Rolling bearing RUL prediction method based on model migration and wiener process | |
| CN117189082A (en) | Intelligent identification and prediction method for sand production information of gas well | |
| CN117233547A (en) | Fusion recognition method and system for oil paper insulation partial discharge | |
| CN113686577B (en) | Bearing fault diagnosis method based on rapid nonlinear sparse spectrum | |
| CN118312836B (en) | Method and system for detecting bolt connection state between rigid cage guides | |
| CN112834194B (en) | Fault intelligent detection method based on soft target measurement under fault-free sample | |
| CN117367570B (en) | Intelligent fault diagnosis method for air compressor based on single-point acoustic signals | |
| CN116664573B (en) | Downhole drill rod number statistics method based on improved YOLOX |
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 |