CN116612388B - Blocking removing method and system for oil production well - Google Patents
Blocking removing method and system for oil production well Download PDFInfo
- Publication number
- CN116612388B CN116612388B CN202310869306.XA CN202310869306A CN116612388B CN 116612388 B CN116612388 B CN 116612388B CN 202310869306 A CN202310869306 A CN 202310869306A CN 116612388 B CN116612388 B CN 116612388B
- Authority
- CN
- China
- Prior art keywords
- well
- congestion
- feature map
- deep
- scale
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 230000000903 blocking effect Effects 0.000 title claims abstract description 62
- 238000004519 manufacturing process Methods 0.000 title claims description 37
- 239000011159 matrix material Substances 0.000 claims description 25
- 230000004927 fusion Effects 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 24
- 239000000463 material Substances 0.000 claims description 23
- 238000004458 analytical method Methods 0.000 claims description 17
- 238000009826 distribution Methods 0.000 claims description 15
- 230000002146 bilateral effect Effects 0.000 claims description 11
- 238000001914 filtration Methods 0.000 claims description 11
- 239000000203 mixture Substances 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 238000011084 recovery Methods 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 5
- 230000004931 aggregating effect Effects 0.000 claims description 4
- 238000011176 pooling Methods 0.000 claims description 4
- 230000001131 transforming effect Effects 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims 1
- 239000003129 oil well Substances 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 15
- 230000000694 effects Effects 0.000 description 10
- 239000000126 substance Substances 0.000 description 10
- 230000008569 process Effects 0.000 description 9
- 239000003795 chemical substances by application Substances 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 229910052500 inorganic mineral Inorganic materials 0.000 description 4
- 239000011707 mineral Substances 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 238000005406 washing Methods 0.000 description 4
- 239000001993 wax Substances 0.000 description 4
- 239000002253 acid Substances 0.000 description 3
- 239000003929 acidic solution Substances 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- VEXZGXHMUGYJMC-UHFFFAOYSA-N Hydrochloric acid Chemical compound Cl VEXZGXHMUGYJMC-UHFFFAOYSA-N 0.000 description 2
- QAOWNCQODCNURD-UHFFFAOYSA-N Sulfuric acid Chemical compound OS(O)(=O)=O QAOWNCQODCNURD-UHFFFAOYSA-N 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 239000012459 cleaning agent Substances 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 239000013078 crystal Substances 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 238000000197 pyrolysis Methods 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 239000002904 solvent Substances 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000002076 thermal analysis method Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000001712 DNA sequencing Methods 0.000 description 1
- 238000004566 IR spectroscopy Methods 0.000 description 1
- 238000005481 NMR spectroscopy Methods 0.000 description 1
- 238000002441 X-ray diffraction Methods 0.000 description 1
- 230000002378 acidificating effect Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 239000003637 basic solution Substances 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000009614 chemical analysis method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000004455 differential thermal analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000007789 gas Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000004949 mass spectrometry Methods 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 230000002906 microbiologic effect Effects 0.000 description 1
- 238000012543 microbiological analysis Methods 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003960 organic solvent Substances 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000007790 scraping Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000002411 thermogravimetry Methods 0.000 description 1
- 238000004506 ultrasonic cleaning Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
- E21B43/00—Methods or apparatus for obtaining oil, gas, water, soluble or meltable materials or a slurry of minerals from wells
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
- E21B47/00—Survey of boreholes or wells
- E21B47/002—Survey of boreholes or wells by visual inspection
-
- 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
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/52—Scale-space analysis, e.g. wavelet analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/806—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mining & Mineral Resources (AREA)
- Geology (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Environmental & Geological Engineering (AREA)
- Fluid Mechanics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geochemistry & Mineralogy (AREA)
- Geophysics (AREA)
- Image Analysis (AREA)
Abstract
A method and system for removing the blocking of oil well are disclosed. Firstly, acquiring an in-well blockage image acquired by a camera, then extracting an optimized multi-scale in-well blockage deep feature map from the in-well blockage image, and then determining a blockage removal scheme based on the optimized multi-scale in-well blockage deep feature map. In this way, the degree of well plugging can be intelligently assessed and an appropriate plugging removal protocol selected.
Description
Technical Field
The present disclosure relates to the field of oil well unblocking, and more particularly, to an oil well unblocking method and system thereof.
Background
Oil recovery wells tend to suffer from various levels of plugging due to long runs and deposit buildup. In order to ensure the normal productivity of the oil production well and prolong the service life of the equipment, the problem of blockage in the well needs to be solved in time.
However, different plugging levels often require different plugging removal schemes, and thus accurate assessment of the plugging level in the well and selection of the appropriate plugging removal scheme is a critical issue.
Traditional evaluation methods mainly rely on manual observation and empirical judgment, and have subjectivity and limitation. Thus, an optimized oil recovery well de-plugging scheme is desired.
Disclosure of Invention
In view of this, the present disclosure proposes a method and system for unblocking a production well, which can intelligently evaluate the degree of blockage in the well and select an appropriate unblocking scheme.
According to an aspect of the present disclosure, there is provided a method for unblocking an oil production well, including: acquiring an in-well blockage image acquired by a camera; extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and determining a blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map.
In a method for unblocking a production well, extracting an optimized multi-scale intra-well congestion depth feature map from the intra-well congestion image, comprising: performing image enhancement on the well blockage image to obtain an enhanced well blockage image; extracting a multi-scale intra-well congestion deep feature map from the enhanced intra-well congestion image based on a deep convolutional neural network model; and optimizing the characteristic distribution of the multi-scale intra-well congestion deep characteristic map to obtain the optimized multi-scale intra-well congestion deep characteristic map.
In a method for unblocking a production well, image enhancement of the well blockage image to obtain an enhanced well blockage image, comprising: and carrying out bilateral filtering on the well blockage image to obtain the enhanced well blockage image.
In the oil production well blocking removal method, based on a deep convolutional neural network model, extracting a multi-scale intra-well blocking deep feature map from the enhanced intra-well blocking image, comprising: the enhanced in-well blocking image passes through a feature extractor based on a pyramid network to obtain an in-well blocking shallow layer feature map, an in-well blocking middle layer feature map and an in-well blocking deep layer feature map; and merging the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map and the in-well congestion deep layer feature map to obtain the multi-scale in-well congestion deep layer feature map.
In the oil production well blocking removal method, the in-well blocking shallow layer feature map, the in-well blocking middle layer feature map and the in-well blocking deep layer feature map are fused to obtain the multi-scale in-well blocking deep layer feature map, which comprises the following steps: and utilizing a self-adaptive fusion module to fuse the in-well congestion shallow layer characteristic map, the in-well congestion middle layer characteristic map and the in-well congestion deep layer characteristic map so as to obtain the multi-scale in-well congestion deep layer characteristic map.
In the oil production well blocking removal method, the self-adaptive fusion module is utilized to fuse the in-well blocking shallow layer characteristic diagram, the in-well blocking middle layer characteristic diagram and the in-well blocking deep layer characteristic diagram to obtain the multi-scale in-well blocking deep layer characteristic diagram, and the method comprises the following steps: fusing the in-well congestion shallow layer feature map and the in-well congestion middle layer feature map into a first input tensor, and then passing through a first convolution layer to obtain a middle layer coding feature map; fusing the middle layer coding feature map and the in-well congestion deep layer feature map into a second input tensor, and then passing through a second convolution layer to obtain a deep layer coding feature map; aggregating the in-well congestion shallow feature map, the middle layer coding feature map and the deep coding feature map into a multi-scale collage feature map along a channel dimension; and performing point convolution processing on the multi-scale collage feature map to obtain the multi-scale intra-well congestion deep feature map.
In the oil production well blocking removal method, performing feature distribution optimization on the multi-scale intra-well blocking deep feature map to obtain the optimized multi-scale intra-well blocking deep feature map, including: calculating a weighted feature vector of the multi-scale intra-well congestion deep feature map; and weighting each feature matrix of the multi-scale intra-well congestion deep feature map along the channel dimension based on the weighting feature vector to obtain the optimized multi-scale intra-well congestion deep feature map.
In the oil production well blocking removal method, calculating a weighted feature vector of the multi-scale intra-well blocking deep feature map comprises the following steps: calculating the weighted feature vector of the multi-scale intra-well congestion depth feature map according to the following weighted formula; wherein, the weighting formula is:firstly, linearly transforming each feature matrix channel of the multi-scale intra-well congestion deep feature map into +.>Square matrix of>Is the number of channels of the multi-scale intra-well congestion depth profile, < >>Is the converted multi-scale intra-well congestion depth profile along the channel dimension +.>Characteristic matrix->Is the vector obtained by global pooling of each feature matrix of the converted multi-scale intra-well congestion deep feature map,/v- >Is the conversion ofThe +.sup.th of the multi-scale downhole congestion depth profile>First->Characteristic value of the location->Representing addition by position +.>Representing multiplication by location>Representing subtraction by position +.>Representing the weighted feature vector.
In the oil production well blocking removal method, determining a blocking removal scheme based on the optimized multi-scale intra-well blocking deep feature map comprises the following steps: the optimized multi-scale intra-well congestion deep feature map is subjected to a classifier to obtain a classification result, and the classification result is used for representing an intra-well congestion degree label; and determining a de-plugging scheme based on the composition analysis of the plugging material in the well and the classification result.
According to another aspect of the present disclosure, there is provided a system for unblocking an oil recovery well, comprising: the image acquisition module is used for acquiring an in-well blockage image acquired by the camera; the extraction optimization module is used for extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and the blocking removal scheme determining module is used for determining a blocking removal scheme based on the optimized multi-scale intra-well congestion deep characteristic diagram.
According to an embodiment of the disclosure, an in-well blockage image acquired by a camera is firstly acquired, then an optimized multi-scale in-well congestion depth feature map is extracted from the in-well blockage image, and then a blockage removal scheme is determined based on the optimized multi-scale in-well congestion depth feature map. In this way, the degree of well plugging can be intelligently assessed and an appropriate plugging removal protocol selected.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 shows a flow chart of a method of unblocking a production well according to an embodiment of the present disclosure.
Fig. 2 shows an architectural schematic diagram of a production well unblocking method according to an embodiment of the present disclosure.
Fig. 3 shows a flowchart of substep S120 of a production well unblocking method according to an embodiment of the present disclosure.
Fig. 4 shows a flowchart of substep S122 of the production well unblocking method according to an embodiment of the present disclosure.
Fig. 5 shows a flowchart of sub-step S1222 of a production well unblocking method according to an embodiment of the present disclosure.
Fig. 6 shows a flowchart of sub-step S123 of a production well unblocking method according to an embodiment of the present disclosure.
Fig. 7 shows a flowchart of sub-step S130 of a production well unblocking method according to an embodiment of the present disclosure.
Fig. 8 illustrates a block diagram of a production well unblocking system according to an embodiment of the present disclosure.
Fig. 9 illustrates an application scenario diagram of a production well unblocking method according to an embodiment of the present disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this disclosure and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following detailed description in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order not to obscure the present disclosure.
In view of the technical problems, the technical idea of the present disclosure is to recommend a deblocking scheme based on a blocking substance analysis technique and a blocking degree evaluation. In particular, the severity of the degree of blockage can affect the selection of the method of unblocking. For example, for a light blockage, simpler methods such as flushing or acid washing, etc. may be used; for severe plugging, more complex unplugging methods, such as mechanical unplugging or jet dredging, may be required. Furthermore, the selection of the deblocking method also requires consideration of the nature of the blocking material. Among these, common plugging materials may be deposits, oil stains, minerals, waxes, etc., and because of the different physical and chemical characteristics of the different materials, different plugging removal methods are typically required to interact with them.
It should be noted that, for the unblocking of the deposit, a physical method including washing, scraping or vibrating, etc. may be employed to remove or break up the deposit, or a chemical method may use a dissolving agent or a cleaning agent to dissolve or decompose the deposit. For the blocking removal of the oil dirt, a dissolving agent or a cleaning agent can be used for dissolving or decomposing the oil dirt, and in addition, physical methods such as high-pressure water jet or ultrasonic cleaning can be used for removing or crushing the oil dirt. For mineral deblocking, common methods include acid washing or solvent treatment, and acid washing may use an acidic solution such as hydrochloric acid or sulfuric acid to dissolve or decompose minerals; the dissolving agent treatment may use a suitable dissolving agent, such as an acidic or basic solution, to dissolve the minerals. For deblocking of wax, a pyrolysis process or a solvent treatment may be employed, and the pyrolysis process may use hot oil or steam to heat the wax and melt or evaporate it; the dissolving agent treatment may use a suitable dissolving agent, such as an organic solvent or an acidic solution, to dissolve the wax. It should be noted that in selecting the deblocking method, the severity of the degree of blocking and the nature of the blocking material should be considered in combination to determine the most appropriate deblocking scheme.
Fig. 1 shows a flow chart of a method of unblocking a production well according to an embodiment of the present disclosure. Fig. 2 shows an architectural schematic diagram of a production well unblocking method according to an embodiment of the present disclosure. As shown in fig. 1 and 2, a method for unblocking a production well according to an embodiment of the present disclosure includes the steps of: s110, acquiring an in-well blockage image acquired by a camera; s120, extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and S130, determining a blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map.
Specifically, in the technical scheme of the disclosure, an in-well blockage image acquired by a camera is firstly acquired, and bilateral filtering is performed on the in-well blockage image to obtain an enhanced in-well blockage image. Wherein the purpose of bilateral filtering of the well-occlusion image is to reduce noise interference in the image. It should be appreciated that for cameras used in the well unblocking method, a camera with high definition, good light adaptability, high temperature resistance, corrosion resistance, reliable transmission and storage capabilities, and capable of adapting to the well environment may be selected. The image quality of the camera should be high enough to clearly capture the blockage situation in the well so as to accurately analyze and judge; the condition that the light is insufficient or the light is changed strongly exists in the internal environment of the oil extraction well, and the camera needs to have good light adaptability and can work normally under various light conditions; the temperature inside the oil extraction well is high, and corrosive media may exist, so the camera needs to have high temperature resistance and corrosion resistance so as to ensure reliable operation under severe environments; the camera needs to be able to reliably transmit image data and have sufficient storage capacity for subsequent data processing and analysis; the camera is sized and shaped to accommodate the tubing and wellbore configuration of the production well for ease of installation and operation.
And then, the enhanced in-well blockage image passes through a feature extractor based on a pyramid network to obtain an in-well blockage shallow layer feature map, an in-well blockage middle layer feature map and an in-well blockage deep layer feature map. That is, a feature extractor is constructed using a pyramid network to capture multi-level feature information. And then, fusing the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map and the in-well congestion deep layer feature map to obtain a multi-scale in-well congestion deep layer feature map.
In one specific example of the present disclosure, the fusion of the in-well congestion shallow feature map, the in-well congestion middle layer feature map, and the in-well congestion deep feature map is performed using an adaptive fusion module. The coding process of the self-adaptive fusion module specifically comprises the following steps: firstly, fusing the in-well congestion shallow layer feature map and the in-well congestion middle layer feature map into a first input tensor, and then passing through a first convolution layer to obtain a middle layer coding feature map; then, fusing the middle layer coding feature map and the in-well congestion deep layer feature map into a second input tensor, and then passing through a second convolution layer to obtain a deep layer coding feature map; then, the in-well congestion shallow layer feature map, the middle layer coding feature map and the deep coding feature map are aggregated into a multi-scale collage feature map along a channel dimension; and then, carrying out point convolution processing on the multi-scale collage feature map to obtain the multi-scale intra-well congestion deep feature map. Thus, the self-adaptive fusion module can be utilized to retain multi-level information. Meanwhile, in order to not excessively increase the parameter number of the model and keep the channel number unchanged, multi-level characteristic fusion can be performed, and multi-level information is fully utilized. That is, the processing manner can fully utilize the shallow features in the in-well congestion shallow feature map, so that the shallow features are not depleted as the number of network layers increases.
Accordingly, in one possible implementation, as shown in fig. 3, extracting an optimized multi-scale intra-well congestion depth profile from the intra-well congestion image includes: s121, performing image enhancement on the well blockage image to obtain an enhanced well blockage image; s122, extracting a multi-scale intra-well congestion deep feature map from the enhanced intra-well congestion image based on a deep convolutional neural network model; and S123, performing feature distribution optimization on the multi-scale intra-well congestion deep feature map to obtain the optimized multi-scale intra-well congestion deep feature map.
More specifically, in step S121, image enhancement is performed on the in-well blockage image to obtain an enhanced in-well blockage image, including: and carrying out bilateral filtering on the well blockage image to obtain the enhanced well blockage image. It should be appreciated that Bilateral filtering (Bitemporal Filter) is an image filtering algorithm that is used to smooth images and reduce the effects of noise. Which performs smoothing processing on an image while maintaining edge information. Conventional linear filtering algorithms (e.g., mean or gaussian) can obscure the edge information of the image, resulting in loss of detail of the image. While bilateral filtering retains edge information by assigning weights of filters to neighboring pixels in consideration of similarity between pixels. The weights of the bilateral filter consist of two parts: spatial domain weights and gray domain weights. The spatial domain weight measures the distance between pixels, with closer pixels having higher weights; the gray domain weight measures the gray difference between pixels, with pixels of similar gray having higher weights. In this way, the bilateral filter is able to preserve edge detail while smoothing the image. In the technical scheme of the disclosure, the purpose of bilateral filtering of the well blockage image is to reduce noise interference in the image so as to obtain an enhanced well blockage image. By bilateral filtering, details of edges of the blocking object can be reserved while the image is smoothed, and accuracy of determination of a subsequent feature extraction and blocking removal scheme is improved.
More specifically, in step S122, as shown in fig. 4, based on the deep convolutional neural network model, a multi-scale intra-well congestion deep feature map is extracted from the enhanced intra-well congestion image, including: s1221, enabling the enhanced in-well blockage image to pass through a feature extractor based on a pyramid network to obtain an in-well blockage shallow layer feature map, an in-well blockage middle layer feature map and an in-well blockage deep layer feature map; and S1222, merging the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map and the in-well congestion deep layer feature map to obtain the multi-scale in-well congestion deep layer feature map. It should be appreciated that a pyramid network-based feature extractor is a neural network structure for extracting multi-scale features that is designed to process images with different scale information to capture different levels of features. Pyramid networks are typically composed of multiple parallel convolution layers, each having a different receptive field (field) size, which different convolution layers can capture features of different scales. The feature extractor of the pyramid network gradually extracts low-level features and high-level features of the image by carrying out rolling and pooling operations on different levels, the features of the lower level generally contain more detail information, the features of the higher level are more abstract and semantic, and a more comprehensive and richer feature representation can be obtained by fusing the features of the different levels. In the described enhanced in-well occlusion image, the pyramid network-based feature extractor may obtain in-well occlusion shallow feature maps, in-well occlusion middle layer feature maps, and in-well occlusion deep feature maps by extracting features at different levels. These feature maps can be used for subsequent image processing and analysis to enable evaluation of well plugging conditions and formulation of solutions.
Further, fusing the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map and the in-well congestion deep layer feature map to obtain the multi-scale in-well congestion deep layer feature map, including: and utilizing a self-adaptive fusion module to fuse the in-well congestion shallow layer characteristic map, the in-well congestion middle layer characteristic map and the in-well congestion deep layer characteristic map so as to obtain the multi-scale in-well congestion deep layer characteristic map.
More specifically, as shown in fig. 5, the in-well congestion shallow feature map, the in-well congestion middle layer feature map and the in-well congestion deep feature map are fused by using an adaptive fusion module to obtain the multi-scale in-well congestion deep feature map, including: s12221, fusing the in-well congestion shallow layer feature map and the in-well congestion middle layer feature map into a first input tensor, and then passing through a first convolution layer to obtain a middle layer coding feature map; s12222, fusing the middle layer coding feature map and the in-well congestion deep layer feature map into a second input tensor, and then passing through a second convolution layer to obtain a deep layer coding feature map; s12223, aggregating the in-well congestion shallow layer feature map, the middle layer coding feature map and the deep layer coding feature map into a multi-scale collage feature map along a channel dimension; and S12224, performing point convolution processing on the multi-scale collage feature map to obtain the multi-scale intra-well congestion deep feature map.
It should be understood that the adaptive fusion module plays a key role in the processing of the in-well congestion image, and by fusing the in-well congestion shallow layer feature map, the middle layer feature map and the deep layer feature map, the adaptive fusion module can generate a multi-scale in-well congestion deep layer feature map, which has the following specific roles: 1. the feature expression capability is improved, the in-well congestion shallow feature map, the middle layer feature map and the deep feature map capture information of different layers respectively, and the self-adaptive fusion module can improve the feature expression capability by fusing the feature maps, so that the generated multi-scale in-well congestion deep feature map is richer and more accurate. 2. The multi-scale information fusion module fuses the in-well congestion shallow layer feature map and the middle layer feature map into a first input tensor, obtains the middle layer coding feature map through a first convolution layer, fuses the middle layer coding feature map and the in-well congestion deep layer feature map into a second input tensor, and obtains the deep layer coding feature map through a second convolution layer, so that the multi-scale information fusion can comprehensively utilize feature information of different layers, and the effect of in-well congestion image processing is improved. 3. The self-adaptive fusion module is used for aggregating the in-well congestion shallow feature map, the middle-layer coding feature map and the deep coding feature map along the channel dimension to generate a multi-scale collage feature map, and the aggregation operation can keep the information of different scale features and plays a key role in point convolution processing. 4. The self-adaptive fusion module carries out point convolution processing on the multi-scale collage feature map to obtain a final multi-scale intra-well congestion deep feature map, and the point convolution processing can effectively extract the spatial relationship among the features and further optimize the processing result of the intra-well congestion image. The self-adaptive fusion module can improve the accuracy and effect of processing the congestion image in the well by fusing and processing the feature images of different layers, and provides more effective feature representation for subsequent analysis and decision.
It is worth mentioning that the point convolution process (Pointwise Convolution) is an operation in a convolutional neural network, also called point-by-point convolution or 1x1 convolution, which is a convolution operation with a convolution kernel size of 1x 1. The effect of the point convolution process is to perform an independent convolution operation on each pixel point of the input feature map, which generates an output feature map by linearly combining and non-linearly transforming features in the channel dimension of each pixel point. The convolution kernel size of the point convolution processing is 1x1, so that the space size of the feature map is not changed in the convolution operation, and only the channel dimension is operated. The point convolution process has a variety of applications in neural networks that can be used to adjust the number of channels of a feature map, by varying the number of channels to control the dimensionality and expressivity of the feature. In addition, the point convolution process can also be used for feature map fusion, and features from different sources are combined linearly to obtain a richer and diversified feature representation. In the process of generating the multi-scale intra-well congestion deep feature map, point convolution processing is used for processing the multi-scale collage feature map to obtain the multi-scale intra-well congestion deep feature map. Specifically, the point convolution operation generates a final multi-scale intra-well congestion deep feature map by performing independent convolution operation on each pixel point of the multi-scale collage feature map, so that the features can be further integrated and extracted while the space size is kept unchanged.
It is noted that in this disclosure, in addition to the fusion method described in S1222, the following method may be used to fuse the shallow, middle and deep congestion feature maps in the well to obtain a multi-scale deep congestion feature map in the well: 1. average fusion is carried out on the in-well congestion shallow feature map, the middle layer feature map and the deep feature map pixel by pixel, and an average fused multi-scale in-well congestion deep feature map is obtained; 2. weighting fusion, namely distributing a weight for each feature map, and then carrying out weighting fusion on the shallow feature map, the middle feature map and the deep feature map of the well congestion according to the weights to obtain a multi-scale well congestion deep feature map of the weighting fusion, wherein the weights can be set according to the importance of the feature maps, for example, the weights can be determined according to the contribution of the feature maps to the degree of the well congestion; the feature cascade connects the shallow layer feature map, the middle layer feature map and the deep layer feature map of the well congestion according to the channel dimension to form a deeper feature map, so that the information of each feature map can be reserved, and the feature maps are combined into a richer multi-scale well congestion deep layer feature map.
More specifically, in step S123, as shown in fig. 6, the optimizing the feature distribution of the multi-scale intra-well congestion depth feature map to obtain the optimized multi-scale intra-well congestion depth feature map includes: s1231, calculating a weighted feature vector of the multi-scale intra-well congestion deep feature map; and S1232, weighting each feature matrix of the multi-scale intra-well congestion deep feature map along the channel dimension based on the weighted feature vector to obtain the optimized multi-scale intra-well congestion deep feature map.
Further, the multi-scale intra-well congestion depth feature map is passed through a classifier to obtain a classification result, wherein the classification result is used for representing an intra-well congestion degree label. That is, the classifier is used for carrying out class boundary division and determination on the high-dimensional data manifold of the multi-scale intra-well congestion depth feature map so as to obtain a label for representing the degree of the intra-well congestion. In practical applications, a de-plugging scheme is determined based on the composition analysis of the plugging material in the well and the classification results. By combining the component analysis of the blocking substances, the blocking removal scheme can be more accurately determined, and the blocking removal efficiency and success rate are improved.
The classification and determination of the high-dimensional data manifold refers to analyzing and processing the high-dimensional data of the congestion deep feature map in the multi-scale well through a classifier to determine boundaries among different categories and divide the boundaries into different categories. In the in-well congestion image processing, these categories may represent different in-well congestion level labels. A high-dimensional dataform refers to a data distribution in high-dimensional space that describes the relationship and structure between data points, each pixel of which can be considered a data point in the high-dimensional space in multi-scale well congestion depth feature map in the well congestion image processing, and the relationship and structure between these data points can be analyzed by a classifier. The classifier is a machine learning model that can learn the mapping from input data (multi-scale in-well congestion depth feature map) to output labels (in-well congestion level labels), by training the classifier, it can learn the boundaries between different classes and can assign new data points to corresponding classes. Therefore, class boundaries are divided and determined through the high-dimensional data manifold, and the classifier can map the class boundaries to corresponding in-well congestion degree labels according to the characteristics of the multi-scale in-well congestion deep feature map. Therefore, in practical application, the degree of congestion in the well can be expressed according to the classification result, and the blocking removal scheme is determined by combining the component analysis of blocking substances, so that the blocking removal efficiency and the success rate are improved.
Among them, the composition analysis based on the plugging material in the well is obtained by a plugging material analysis technique, which is a technique for determining the composition and properties of the plugging material in a pipe, a line or a device, which can help engineers and operators to understand the source, composition and characteristics of the plugging material, thereby taking corresponding measures for cleaning or preventing. The following are some of the commonly used techniques for analysis of plugging materials: physical observation and inspection, by directly observing and inspecting the characteristics of appearance, shape, color, etc. of the clogging substance, the composition and source thereof can be preliminarily judged, for example, whether it is a solid, liquid or gas clogging substance can be determined by visual observation; chemical analysis, in which the chemical components and compositions of the plugging material can be determined by collecting a sample of the plugging material, and common chemical analysis methods include mass spectrometry, infrared spectrometry, nuclear magnetic resonance analysis, and the like; observing by an optical microscope, and observing the microstructure and morphology of the blocking material by using the optical microscope, so that information about the tissue structure, crystal morphology, particle size and the like of the blocking material can be obtained; x-ray diffraction analysis, wherein the crystal structure and the components of the blocking material can be determined by measuring the diffraction mode of the blocking material on X-rays; thermal analysis, using thermal analysis techniques such as differential thermal analysis, thermogravimetric analysis, etc., to determine thermal properties of the plugging material such as melting point, thermal stability, etc.; microbiological analysis, for the possible presence of clogging materials caused by microorganisms, can be performed, such as culture, colony counting, DNA sequencing, etc., to determine the microbiological species and quantity thereof. These plugging material analysis techniques may be used in conjunction with one another to obtain more comprehensive and accurate plugging material information.
Accordingly, in one possible implementation, as shown in fig. 7, determining the blocking removal scheme based on the optimized multi-scale intra-well congestion depth profile includes: s131, the optimized multi-scale intra-well congestion deep feature map is passed through a classifier to obtain a classification result, wherein the classification result is used for representing an intra-well congestion degree label; and S132, determining a blocking removal scheme based on the component analysis of the blocking substances in the well and the classification result. It should be appreciated that the classifier learns classification rules using a given class and known training data and then classifies (or predicts) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
In the technical solution of the present disclosure, the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map, and the in-well congestion deep layer feature map express shallow layer, middle layer, and deep layer image semantic features of the enhanced in-well congestion image, respectively, and since they are obtained from a feature extractor based on a pyramid network, the shallow layer, middle layer, and deep layer image semantic features also correspond to different feature extraction scales. Under the situation, when the intra-well congestion shallow feature map, the intra-well congestion middle layer feature map and the intra-well congestion deep feature map are fused through the self-adaptive module to obtain the multi-scale intra-well congestion deep feature map, the overall feature distribution association effect of the multi-scale intra-well congestion deep feature map is poor due to the difference between the expression scale and the expression depth among the feature matrices of the intra-well congestion shallow feature map, the intra-well congestion middle layer feature map and the intra-well congestion deep feature map, and the accuracy of the classification result obtained by the classifier of the multi-scale intra-well congestion deep feature map is affected due to the overall distribution difference of the feature matrices when the feature representation of the feature matrices is enriched through increasing the channel number.
Based on the above, the global feature distribution association effect of the multi-scale intra-well congestion deep feature map can be improved by weighting each feature matrix of the multi-scale intra-well congestion deep feature map along the channel dimension, and the weighted feature vector performs constrained directional bias through the static scene of each feature matrix so as to perform self-tuning structuring on the feature matrix for calculation.
Accordingly, in one possible implementation, calculating the weighted feature vector of the multi-scale intra-well congestion depth feature map includes: calculating the weighted feature vector of the multi-scale intra-well congestion depth feature map according to the following weighted formula; wherein, the weighting formula is:firstly, linearly transforming each feature matrix channel of the multi-scale intra-well congestion deep feature map into +.>Square matrix of>Is the number of channels of the multi-scale intra-well congestion depth profile, < >>Is the converted multi-scale intra-well congestion depth profile along the channel dimension +.>Characteristic matrix->Is the vector obtained by global pooling of each feature matrix of the converted multi-scale intra-well congestion deep feature map,/v- >Is the +.f. of the converted multi-scale intra-well congestion depth profile>First->Characteristic value of the location->Representing addition by position +.>Representing multiplication by location>Representing subtraction by position +.>Representing the weighted feature vector.
That is, for the multi-scale intra-well congestion depth profile with the weighted feature vectorCan pass through each static scene matrix along the channel dimension of the multi-scale intra-well congestion depth feature map when weighting the respective feature matrices of (a)Relative to channel control vector->The method comprises the steps of supporting self-tuning of static feature scenes by using directional bias vector quantities for expressing channel dimension association, so that structuring of the high-dimensional feature manifold is carried out based on a specific convex polyhedron family (convex polytopes family) corresponding to the feature scenes expressed by various feature matrixes and the high-dimensional feature manifold of the multi-scale well congestion deep feature map, explicit association between the image semantic expression of the scenerization of various feature matrixes and the model feature extraction expression of the channel dimension is improved, and therefore the global feature distribution time sequence association effect of the multi-scale well congestion deep feature map is improved.
It should be appreciated that the global feature distribution correlation effect of the feature map can be improved by weighting each feature matrix of the multi-scale intra-well congestion depth feature map along the channel dimension. The weighting mode can be adjusted according to the importance of the feature matrix, so that the more important feature matrix plays a larger role in calculation, and the expression capacity and the discrimination of the congestion deep feature map in the multi-scale well are improved. The weighted feature vector can carry out self-tuning structural calculation on the feature matrix through the constrained directional bias of the static scene of each feature matrix. The method means that the feature matrix can be better adapted to the characteristics and changes of the congestion scene in the well by restraining and adjusting the feature matrix, and the robustness and accuracy of the features are improved. Therefore, the global feature distribution association effect of the multi-scale intra-well congestion deep feature map can be improved, and the expression capacity and the distinguishing degree of the feature map are enhanced. And by directional deviation and self-tuning structural calculation, the feature matrix is better adapted to the characteristics and changes of the congestion scene in the well, and the robustness and accuracy of the features are improved. In other words, the method can improve the effect of processing the congestion image in the well and improve the accuracy and success rate of the blockage removal scheme.
In summary, the method for unblocking a production well according to the embodiments of the present disclosure may intelligently evaluate the degree of blockage in the well and select an appropriate unblocking scheme.
Fig. 8 illustrates a block diagram of a production well unblocking system 100 according to an embodiment of the present disclosure. As shown in fig. 8, a production well unblocking system 100 according to an embodiment of the present disclosure includes: an image acquisition module 110 for acquiring an in-well blockage image acquired by the camera; the extraction optimization module 120 is configured to extract an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and a blocking removal scheme determining module 130, configured to determine a blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described oil well unblocking system 100 have been described in detail in the above description of the oil well unblocking method with reference to fig. 1 to 7, and thus, repetitive descriptions thereof will be omitted.
As described above, the oil well unblocking system 100 according to the embodiment of the present disclosure may be implemented in various wireless terminals, such as a server or the like having an oil well unblocking algorithm. In one possible implementation, the oil recovery well unblocking system 100 according to embodiments of the present disclosure may be integrated into the wireless terminal as one software module and/or hardware module. For example, the production well unblocking system 100 may be a software module in the operating system of the wireless terminal or may be an application developed for the wireless terminal; of course, the production well unblocking system 100 can also be one of a number of hardware modules of the wireless terminal.
Alternatively, in another example, the production well deblock system 100 and the wireless terminal may be separate devices, and the production well deblock system 100 may be connected to the wireless terminal via a wired and/or wireless network and communicate interactive information in a agreed-upon data format.
Fig. 9 illustrates an application scenario diagram of a production well unblocking method according to an embodiment of the present disclosure. As shown in fig. 9, in this application scenario, first, an in-well blockage image (e.g., D shown in fig. 9) acquired by a camera (e.g., C shown in fig. 9) is acquired, and then the in-well blockage image is input to a server (e.g., S shown in fig. 9) where an oil recovery well blockage removal algorithm is deployed, wherein the server is capable of processing the in-well blockage image using the oil recovery well blockage removal algorithm to obtain a classification result for indicating a congestion degree label in a well.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (7)
1. The method for unblocking the oil extraction well is characterized by comprising the following steps of:
acquiring an in-well blockage image acquired by a camera;
extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and
determining a blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map;
extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image, wherein the method comprises the following steps of:
performing image enhancement on the well blockage image to obtain an enhanced well blockage image;
extracting a multi-scale intra-well congestion deep feature map from the enhanced intra-well congestion image based on a deep convolutional neural network model; and
Performing feature distribution optimization on the multi-scale intra-well congestion deep feature map to obtain an optimized multi-scale intra-well congestion deep feature map;
the method for optimizing the characteristic distribution of the multi-scale intra-well congestion deep characteristic map to obtain the optimized multi-scale intra-well congestion deep characteristic map comprises the following steps:
calculating a weighted feature vector of the multi-scale intra-well congestion deep feature map; and
weighting each feature matrix of the multi-scale intra-well congestion deep feature map along the channel dimension based on the weighting feature vector to obtain the optimized multi-scale intra-well congestion deep feature map;
the method for determining the blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map comprises the following steps:
the optimized multi-scale intra-well congestion deep feature map is subjected to a classifier to obtain a classification result, and the classification result is used for representing an intra-well congestion degree label; and
based on the composition analysis of the plugging material in the well and the classification results, a plugging removal protocol is determined.
2. The method of plugging removal for a production well of claim 1, wherein image enhancing the well plugging image to obtain an enhanced well plugging image comprises:
And carrying out bilateral filtering on the well blockage image to obtain the enhanced well blockage image.
3. The method of plugging removal for a production well of claim 2, wherein extracting a multi-scale intra-well congestion depth profile from the enhanced intra-well congestion image based on a depth convolutional neural network model comprises:
the enhanced in-well blocking image passes through a feature extractor based on a pyramid network to obtain an in-well blocking shallow layer feature map, an in-well blocking middle layer feature map and an in-well blocking deep layer feature map; and
and merging the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map and the in-well congestion deep layer feature map to obtain the multi-scale in-well congestion deep layer feature map.
4. The method of plugging removal for a production well of claim 3, wherein fusing the shallow in-well congestion profile, the middle in-well congestion profile, and the deep in-well congestion profile to obtain the multi-scale deep in-well congestion profile comprises:
and utilizing a self-adaptive fusion module to fuse the in-well congestion shallow layer characteristic map, the in-well congestion middle layer characteristic map and the in-well congestion deep layer characteristic map so as to obtain the multi-scale in-well congestion deep layer characteristic map.
5. The method of claim 4, wherein fusing the in-well congestion shallow layer feature map, the in-well congestion middle layer feature map, and the in-well congestion deep layer feature map to obtain the multi-scale in-well congestion deep layer feature map using an adaptive fusion module comprises:
fusing the in-well congestion shallow layer feature map and the in-well congestion middle layer feature map into a first input tensor, and then passing through a first convolution layer to obtain a middle layer coding feature map;
fusing the middle layer coding feature map and the in-well congestion deep layer feature map into a second input tensor, and then passing through a second convolution layer to obtain a deep layer coding feature map;
aggregating the in-well congestion shallow feature map, the middle layer coding feature map and the deep coding feature map into a multi-scale collage feature map along a channel dimension; and
and carrying out point convolution processing on the multi-scale collage feature map to obtain the multi-scale intra-well congestion deep feature map.
6. The method of claim 5, wherein calculating a weighted feature vector for the multi-scale intra-well congestion depth feature map comprises:
calculating the weighted feature vector of the multi-scale intra-well congestion depth feature map according to the following weighted formula;
Wherein, the weighting formula is:
，
firstly, linearly transforming each feature matrix channel of the multi-scale intra-well congestion deep feature map intoSquare matrix of>Is the number of channels of the multi-scale intra-well congestion depth profile, < >>Is the converted multi-scale intra-well congestion depth profile along the channel dimension +.>Characteristic matrix->Is the vector obtained by global pooling of each feature matrix of the converted multi-scale intra-well congestion deep feature map,/v->Is the +.f. of the converted multi-scale intra-well congestion depth profile>First->Characteristic value of the location->Representing addition by position +.>Representing multiplication by location>Representing subtraction by position +.>Representing the weighted feature vector.
7. An oil recovery well unblocking system, comprising:
the image acquisition module is used for acquiring an in-well blockage image acquired by the camera;
the extraction optimization module is used for extracting an optimized multi-scale intra-well congestion deep feature map from the intra-well congestion image; and
the blockage removal scheme determining module is used for determining a blockage removal scheme based on the optimized multi-scale intra-well congestion deep feature map;
Wherein, the deblocking scheme determination module is configured to:
performing image enhancement on the well blockage image to obtain an enhanced well blockage image;
extracting a multi-scale intra-well congestion deep feature map from the enhanced intra-well congestion image based on a deep convolutional neural network model; and
performing feature distribution optimization on the multi-scale intra-well congestion deep feature map to obtain an optimized multi-scale intra-well congestion deep feature map;
the method for optimizing the characteristic distribution of the multi-scale intra-well congestion deep characteristic map to obtain the optimized multi-scale intra-well congestion deep characteristic map comprises the following steps:
calculating a weighted feature vector of the multi-scale intra-well congestion deep feature map; and
weighting each feature matrix of the multi-scale intra-well congestion deep feature map along the channel dimension based on the weighting feature vector to obtain the optimized multi-scale intra-well congestion deep feature map;
the method for determining the blocking removal scheme based on the optimized multi-scale intra-well congestion deep feature map comprises the following steps:
the optimized multi-scale intra-well congestion deep feature map is subjected to a classifier to obtain a classification result, and the classification result is used for representing an intra-well congestion degree label; and
Based on the composition analysis of the plugging material in the well and the classification results, a plugging removal protocol is determined.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310869306.XA CN116612388B (en) | 2023-07-17 | 2023-07-17 | Blocking removing method and system for oil production well |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310869306.XA CN116612388B (en) | 2023-07-17 | 2023-07-17 | Blocking removing method and system for oil production well |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116612388A CN116612388A (en) | 2023-08-18 |
| CN116612388B true CN116612388B (en) | 2023-09-19 |
Family
ID=87683836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310869306.XA Active CN116612388B (en) | 2023-07-17 | 2023-07-17 | Blocking removing method and system for oil production well |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116612388B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116935180A (en) * | 2023-09-15 | 2023-10-24 | 东莞市将为防伪科技有限公司 | Information acquisition method and system for information of information code anti-counterfeiting label based on artificial intelligence |
| CN117173748B (en) * | 2023-11-03 | 2024-01-26 | 杭州登虹科技有限公司 | Video humanoid event extraction system based on humanoid recognition and humanoid detection |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101183795B1 (en) * | 2011-09-27 | 2012-09-17 | (주)포미트 | Apparatus and method for solving coal treatment system clogged up with coal |
| CN106958438A (en) * | 2017-02-20 | 2017-07-18 | 中国石油化工股份有限公司 | A kind of polymer flooding blocks the method for removing blockage of well |
| WO2018184397A1 (en) * | 2017-04-07 | 2018-10-11 | 中国石油大学(华东) | Integrated evaluation, testing and simulation apparatus for wellbore sand-control blocking and unblocking, and method |
| JP2019211814A (en) * | 2018-05-31 | 2019-12-12 | 株式会社東芝 | Congestion prediction device and congestion prediction method |
| CN112065341A (en) * | 2020-09-15 | 2020-12-11 | 北京尊宏科技发展有限公司 | Single-well physical blockage removal and intelligent remote control measurement and fine adjustment layered water injection process for water injection well |
| CN112101117A (en) * | 2020-08-18 | 2020-12-18 | 长安大学 | Expressway congestion identification model construction method and device and identification method |
| CN112127858A (en) * | 2020-11-30 | 2020-12-25 | 胜利油田东润机械工程有限责任公司 | Negative pressure flow-back blockage removing process for oil field |
| CN113887088A (en) * | 2021-11-08 | 2022-01-04 | 西南石油大学 | Method for determining dosage of blocking remover for removing reservoir barite pollution |
-
2023
- 2023-07-17 CN CN202310869306.XA patent/CN116612388B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101183795B1 (en) * | 2011-09-27 | 2012-09-17 | (주)포미트 | Apparatus and method for solving coal treatment system clogged up with coal |
| CN106958438A (en) * | 2017-02-20 | 2017-07-18 | 中国石油化工股份有限公司 | A kind of polymer flooding blocks the method for removing blockage of well |
| WO2018184397A1 (en) * | 2017-04-07 | 2018-10-11 | 中国石油大学(华东) | Integrated evaluation, testing and simulation apparatus for wellbore sand-control blocking and unblocking, and method |
| JP2019211814A (en) * | 2018-05-31 | 2019-12-12 | 株式会社東芝 | Congestion prediction device and congestion prediction method |
| CN112101117A (en) * | 2020-08-18 | 2020-12-18 | 长安大学 | Expressway congestion identification model construction method and device and identification method |
| CN112065341A (en) * | 2020-09-15 | 2020-12-11 | 北京尊宏科技发展有限公司 | Single-well physical blockage removal and intelligent remote control measurement and fine adjustment layered water injection process for water injection well |
| CN112127858A (en) * | 2020-11-30 | 2020-12-25 | 胜利油田东润机械工程有限责任公司 | Negative pressure flow-back blockage removing process for oil field |
| CN113887088A (en) * | 2021-11-08 | 2022-01-04 | 西南石油大学 | Method for determining dosage of blocking remover for removing reservoir barite pollution |
Non-Patent Citations (3)
| Title |
|---|
| 多脉冲高能气体压裂-热化学解堵综合增产技术;李伟翰;颜红侠;王世英;吕晓权;刘金义;;油田化学(第03期);全文 * |
| 注聚合物井堵塞机理分析及解堵剂研究;郑俊德, 张英志, 任华, 刘野;石油勘探与开发(第06期);全文 * |
| 连续油管冲砂解堵在水平井中的应用;俞宏伟;姜一超;;科学技术与工程(32);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116612388A (en) | 2023-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116612388B (en) | Blocking removing method and system for oil production well | |
| Tong et al. | Pavement defect detection with fully convolutional network and an uncertainty framework | |
| Tan et al. | Automatic detection of sewer defects based on improved you only look once algorithm | |
| Dong et al. | A deep-learning-based multiple defect detection method for tunnel lining damages | |
| Chatterjee et al. | Genetic algorithms for feature selection of image analysis-based quality monitoring model: An application to an iron mine | |
| Nayyeri et al. | Foreground–background separation technique for crack detection | |
| Mohammed Abdelkader et al. | Entropy-based automated method for detection and assessment of spalling severities in reinforced concrete bridges | |
| CN111369526B (en) | Multi-type old bridge crack identification method based on semi-supervised deep learning | |
| CN116682068B (en) | Oil well sand prevention operation construction monitoring method and system thereof | |
| CN110751644A (en) | Road surface crack detection method | |
| Zhao et al. | Recognition of flooding and sinking conditions in flotation process using soft measurement of froth surface level and QTA | |
| CN114399763B (en) | Single-sample and small-sample micro-body paleobiological fossil image identification method and system | |
| Johansen et al. | Towards detection and classification of microscopic foraminifera using transfer learning | |
| CN117586838A (en) | Cleaning agent for train brake valve and preparation method thereof | |
| Jiao et al. | A steerable pyramid autoencoder based framework for anomaly frame detection of water pipeline CCTV inspection | |
| Chand et al. | An intelligent system to identify coal maceral groups using markov-fuzzy clustering approach | |
| WO2020175666A1 (en) | Color filter inspection device, inspection device, color filter inspection method, and inspection method | |
| Singh et al. | Segmentation technique for the detection of Micro cracks in solar cell using support vector machine | |
| Myrans et al. | Combining classifiers to detect faults in wastewater networks | |
| Liu et al. | The Segmentation of Wear Particles Images Using J‐Segmentation Algorithm | |
| Matarneh et al. | Evaluation and optimisation of pre-trained CNN models for asphalt pavement crack detection and classification | |
| Gutiérrez-Mondragón et al. | Obstruction level detection of sewer videos using convolutional neural networks | |
| CN115374931A (en) | Deep neural network robustness enhancing method based on meta-countermeasure training | |
| Jang et al. | A defect inspection method for machine vision using defect probability image with deep convolutional neural network | |
| Tennakoon et al. | Visual Inspection of Storm-Water Pipe Systems using Deep Convolutional Neural Networks. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Yang Gang Inventor after: Zang Qiang Inventor after: Fang He Inventor after: Lu Xuehui Inventor after: Xie Shuanghui Inventor after: Zhang Xue Inventor before: Yang Gang Inventor before: Zang Qiang Inventor before: Fang He Inventor before: Lu Xuehui Inventor before: Xie Shuanghui Inventor before: Halheng tursong Inventor before: Shi Yongxiang Inventor before: Zhang Xue |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240110 Address after: Room 2, No. 26-8-234, Shengli Road, Karamay District, Karamay City, Xinjiang Uygur Autonomous Region 834000 Patentee after: Xinjiang Huayi Energy Development Co.,Ltd. Patentee after: Xinjiang Oilfield Heiyoushan Co.,Ltd. Address before: Room 2, No. 26-8-234, Shengli Road, Karamay District, Karamay City, Xinjiang Uygur Autonomous Region 834000 Patentee before: Xinjiang Huayi Energy Development Co.,Ltd. |