Old well measure layer intelligent recommendation method and device, equipment and medium

Abstract

Embodiments of the present application provide a method, device, equipment and medium for intelligent recommendation of a measure layer of an old well. The method comprises: obtaining interpretation result data of a plurality of potential layers to be processed and an engineering parameter file of a wellbore where each potential layer is located, the interpretation result data of each potential layer including depth information; cleaning and tokenizing the engineering parameter file of the wellbore where the potential layer is located to obtain processed text data; based on a preset knowledge entity extraction model, labeling and extracting objects, attributes and attribute values of the processed text data to obtain object information of the wellbore where the potential layer is located and attribute information and depth information of the wellbore where each potential layer is located; and according to the depth information of the plurality of potential layers, removing a potential layer having an intersection with the attribute information and depth information of the wellbore where any potential layer is located from the plurality of potential layers, and determining the remaining potential layers as measure layers. The method is used to improve the working efficiency of measure layer selection of an old well.

Description

Technical Field

[0001] This application relates to the field of well logging, and in particular to a method, apparatus, equipment and medium for intelligent recommendation of treatment layers in old wells. Background Technology

[0002] After years of exploitation, oilfields typically experience declining production and rising water cut. Coupled with reduced drilling investment, there is an urgent need to stabilize production through methods that tap potential and increase efficiency. Re-examining old wells, as a low-investment, quick-return, and highly efficient method for increasing reserves and production, is increasingly valued by oilfield exploration and development departments and has become an effective way for various oilfields to increase reserves and production.

[0003] Conventional re-evaluation techniques for old wells in oil and gas reservoirs typically involve two processes: re-evaluating each section of the old well to select potential layers; and comparing engineering parameters such as perforation sections and casing damage locations of the old well to recommend intervention layers from the potential layers.

[0004] In existing technologies, when recommending potential layers, researchers often review engineering parameter documents for various old wells and compare these parameters with the depth of the potential layers to make recommendations. However, this manual review and well-by-well comparison method is extremely inefficient when reviewing multiple wells. Summary of the Invention

[0005] This application provides a method, apparatus, equipment, and medium for intelligent recommendation of old well treatment layers, which aims to improve the efficiency of old well treatment layer selection.

[0006] In a first aspect, embodiments of this application provide an intelligent recommendation method for old well management measures, including:

[0007] Acquire the interpretation results data of multiple potential layers to be processed and the engineering parameter files of the wellbore where each potential layer is located. The interpretation results data of each potential layer include depth information and interpretation layer number.

[0008] The engineering parameter file of the wellbore containing the potential layer is cleaned and segmented to obtain processed text data;

[0009] Based on a preset knowledge entity extraction model, the processed text data is labeled and extracted for objects, attributes and attribute values ​​to obtain the objects in the wellbore where the potential layer is located, as well as the attribute information and depth information of each wellbore where the potential layer is located.

[0010] Based on the depth information and interpretation layer number of the multiple potential layers, potential layers that have intersections with the attribute information and depth information of any wellbore containing any potential layer are removed, and the remaining potential layers are determined as the measure layers.

[0011] In one possible implementation, the interpretation results data for each potential layer also include: values ​​of multiple evaluation indicators for the potential layer;

[0012] The method further includes:

[0013] Based on the values ​​of multiple evaluation indicators for each potential layer, the TOPSIS (Topology-Topology-Sorting-Ideal-Point) method is used to sort the multiple potential layers and determine the recommendation priority of each potential layer.

[0014] In one possible implementation, acquiring the interpretation results data of the multiple potential layers to be processed and the engineering parameter files of the wellbore containing each potential layer includes:

[0015] The terminal device receives a request for recommendation of old well measures layers, which includes interpretation results data of the multiple potential layers and engineering parameter files of the wellbore where each potential layer is located.

[0016] Accordingly, the method further includes:

[0017] The old well measure layer recommendation results are returned to the terminal device. The old well measure layer recommendation results include: measure layer and recommendation priority of each measure layer.

[0018] In one possible implementation, the process of cleaning and segmenting the engineering parameter file of the wellbore containing the potential layer to obtain processed text data includes:

[0019] Data cleaning is performed on the engineering parameter files of the multiple potential layers to remove non-standard characters from the text, resulting in cleaned text data.

[0020] Based on a pre-defined main dictionary in the petroleum field, the cleaned text data is segmented into words to obtain segmented text data.

[0021] Based on a preset stop word library, characters without actual semantic meaning are removed from the segmented text data to obtain the processed text data; wherein, the stop word library includes multiple preset characters without actual meaning.

[0022] In one possible implementation, the pre-defined knowledge entity extraction model is used to annotate and extract objects, attributes, and attribute values ​​from the processed text data to obtain the objects in the wellbore containing the potential layer, the attribute information of each wellbore containing the potential layer, and the depth information, including:

[0023] Based on the pre-trained Bi-LSTM-CRF model, the processed text data is labeled with objects, attributes, and attribute values ​​to obtain labeled text data.

[0024] Extract the objects of the wellbore containing the potential layer and the attribute and depth information of each wellbore containing the potential layer from the labeled text data;

[0025] The object of the wellbore where the potential layer is located is the well number;

[0026] The property information for each potential layer in the wellbore includes at least one of perforation, casing damage, plunger, and cementing quality.

[0027] In one possible implementation, the step of ranking the multiple potential layers using the TOPSIS (Topology-Order-Sorting-Ideal-Point) method based on the values ​​of multiple evaluation indicators for each potential layer, and determining the recommendation priority of each potential layer, includes:

[0028] A decision matrix is ​​constructed based on the values ​​of multiple evaluation indicators for each potential layer, and the decision matrix includes the value of each evaluation indicator for each potential layer.

[0029] The decision matrix is ​​then subjected to forward transformation and normalization to obtain a standardized matrix;

[0030] The maximum value of each evaluation index in the standardized matrix is ​​determined as the positive ideal point, and the minimum value of each evaluation index is determined as the negative ideal point.

[0031] For each potential layer, calculate the first distance between the value of each evaluation index of the potential layer and the corresponding positive ideal point, and the second distance between the value of each index and the corresponding negative ideal point;

[0032] Based on the first and second distances corresponding to the values ​​of each evaluation index in each potential layer, the multiple potential layers are sorted to determine the recommendation priority of each potential layer.

[0033] In one possible implementation, the decision matrix is ​​subjected to forwarding and normalization processes to obtain a standardized matrix, including:

[0034] The decision matrix is ​​then forward-oriented to obtain the forward-oriented matrix.

[0035] Based on the preset weights of each evaluation index, the matrix after positive transformation is normalized to obtain a standardized matrix.

[0036] Secondly, embodiments of this application provide an intelligent recommendation device for old well treatment layers, comprising:

[0037] The data input module is used to acquire the interpretation results data of multiple potential layers to be processed and the engineering parameter file of the wellbore where each potential layer is located. The interpretation results data of each potential layer includes depth information and interpretation layer number.

[0038] The model processing module is used to clean and segment the engineering parameter file of the wellbore where the potential layer is located to obtain processed text data;

[0039] The text annotation and extraction module is used to annotate and extract objects, attributes and attribute values ​​from the processed text data based on a preset knowledge entity extraction model, so as to obtain the objects of the wellbore where the potential layer is located and the attribute information and depth information of each wellbore where the potential layer is located.

[0040] The measure layer determination module is used to eliminate potential layers that have intersections with the attribute information and depth information of any wellbore containing any potential layer, based on the depth information and interpretation layer number of the multiple potential layers, and determine the remaining potential layers as measure layers.

[0041] In one possible implementation, this application provides an intelligent recommendation device for old well management layers, which further includes:

[0042] The sorting module is used to sort the multiple potential layers according to the values ​​of multiple evaluation indicators for each potential layer, using the TOPSIS (Topology-Topology-Sorting-Ideal-Point) method, and to determine the recommendation priority of each potential layer.

[0043] In one possible implementation, the data input module is specifically used for:

[0044] The terminal device receives a request for recommendation of old well measures layers, which includes interpretation results data of the multiple potential layers and engineering parameter files of the wellbore where each potential layer is located.

[0045] Accordingly, this application provides an intelligent recommendation device for old well treatment layers, which further includes:

[0046] Result return module: used to return the old well measure layer recommendation results to the terminal device. The old well measure layer recommendation results include: measure layer and recommendation priority of each measure layer.

[0047] In one possible implementation, the model processing module is specifically used for:

[0048] The engineering parameter file of the wellbore containing the potential layer is cleaned to remove non-standard characters from the text, resulting in cleaned text data.

[0049] Based on a pre-defined main dictionary in the petroleum field, the cleaned text data is segmented into words to obtain segmented text data.

[0050] Based on a preset stop word library, characters without actual semantic meaning are removed from the segmented text data to obtain the processed text data; wherein, the stop word library includes multiple preset characters without actual meaning.

[0051] In one possible implementation, the text annotation and extraction module is specifically used for:

[0052] Based on the pre-trained Bi-LSTM-CRF model, the processed text data is labeled with objects, attributes, and attribute values ​​to obtain labeled text data.

[0053] Extract the objects of the wellbore containing the potential layer and the attribute and depth information of each wellbore containing the potential layer from the labeled text data;

[0054] The object of the wellbore where the potential layer is located is the well number;

[0055] The property information for each potential layer in the wellbore includes at least one of perforation, casing damage, plunger, and cementing quality.

[0056] In one possible implementation, the sorting module is specifically used for:

[0057] A decision matrix is ​​constructed based on the values ​​of multiple evaluation indicators for each potential layer, and the decision matrix includes the value of each evaluation indicator for each potential layer.

[0058] The decision matrix is ​​then subjected to forward transformation and normalization to obtain a standardized matrix;

[0059] The maximum value of each evaluation index in the standardized matrix is ​​determined as the positive ideal point, and the minimum value of each evaluation index is determined as the negative ideal point.

[0060] For each potential layer, calculate the first distance between the value of each evaluation index of the potential layer and the corresponding positive ideal point, and the second distance between the value of each index and the corresponding negative ideal point;

[0061] Based on the first and second distances corresponding to the values ​​of each evaluation index in each potential layer, the multiple potential layers are sorted to determine the recommendation priority of each potential layer.

[0062] In one possible implementation, the sorting module is specifically used for:

[0063] The decision matrix is ​​then forward-oriented to obtain the forward-oriented matrix.

[0064] Based on the preset weights of each evaluation index, the matrix after positive transformation is normalized to obtain a standardized matrix.

[0065] Thirdly, embodiments of this application provide an intelligent recommendation device for old well management measures, comprising: a memory and a processor;

[0066] The memory stores computer-executed instructions;

[0067] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0068] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0069] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0070] The intelligent recommendation method, apparatus, equipment, and medium for old well measure layers provided in this application first acquire the depth information and interpretation layer position from the interpretation results data of multiple potential layers to be processed, as well as the engineering parameter file of the wellbore where each potential layer is located. Then, based on knowledge entity extraction technology, the engineering parameter file of the wellbore where each potential layer is located is cleaned, segmented, annotated, and parameters are extracted to directly obtain the objects, attribute information, and depth information of the wellbore where each potential layer is located. Finally, the depth information of the potential layer is compared with the depth information corresponding to the attribute information of the wellbore where the potential layer is located obtained from the parameter extraction. If the two depth information overlaps in any potential layer, that potential layer is removed, and the remaining potential layers are determined as measure layers. This application achieves the technical effect of improving the efficiency of old well measure layer selection by automatically extracting key parameters from the engineering parameter file using a cloud server and comparing the depth information in the obtained key parameters with the depth information of the potential layers. Attached Figure Description

[0071] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0072] Figure 1 A flowchart illustrating an embodiment of the intelligent recommendation method for old well management layers provided in this application;

[0073] Figure 2 A flowchart illustrating Embodiment 2 of the intelligent recommendation method for old well management layers provided in this application;

[0074] Figure 3 A schematic diagram illustrating the application scenario of the intelligent recommendation method for old well measures layer provided in this application;

[0075] Figure 4 A flowchart illustrating Embodiment 3 of the intelligent recommendation method for old well management layers provided in this application;

[0076] Figure 5 This is a schematic diagram of the structure of an embodiment of the intelligent recommendation device for old well treatment layers provided in this application;

[0077] Figure 6 This is a schematic diagram of the structure of Embodiment 2 of the intelligent recommendation device for old well treatment layers provided in this application;

[0078] Figure 7 A schematic diagram of the structure of the intelligent recommendation device for the old well treatment layer provided in this application.

[0079] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0080] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0081] First, let me explain the terms used in this application:

[0082] Potential reservoirs refer to previously underdeveloped or unidentified oil and gas reservoirs that have been identified as having further exploration or production value after reassessment, analysis, and the application of technical means in oil and gas wells that have already undergone exploration and / or production operations (i.e., old wells).

[0083] Measures layer: refers to oil and gas reservoirs where actual extraction measures can be taken.

[0084] To better understand the technical solution provided in this application, a detailed introduction to the background technology will be provided below.

[0085] After years of exploitation, oilfields typically experience declining production and rising water cut. Coupled with reduced drilling investment, there is an urgent need to tap potential and increase efficiency to stabilize production. Re-evaluation of old wells, as a low-investment, quick-return, and highly efficient method for increasing reserves and production, is increasingly valued by oilfield exploration and development departments and has become an effective approach for increasing reserves and production in various oilfields. In recent years, oilfield units have developed a series of logging-based methods for re-evaluating old wells through scientific research. By conducting multifaceted technical research on reservoir lithology, physical properties, and hydrocarbon-bearing characteristics, they have developed re-evaluation technologies for re-evaluating oil and gas reservoirs.

[0086] Existing methods for re-evaluating old wells typically involve two stages. The first stage, based on fully utilizing existing data from the old well and combining it with modern logging interpretation and evaluation techniques, re-evaluates reservoirs that were misjudged due to limitations in the technology available at the time. This identifies key factors for determining oil and gas content and allows for rapid logging interpretation and evaluation of wells within the block, selecting layers with significant potential. The second stage, building upon the comprehensive evaluation of the old well reservoirs and the selection of potential layers, requires a comprehensive analysis of the wellbore conditions. For the selected potential layers, further examination of wellbore engineering parameters is necessary to determine whether the potential layer has already been addressed or cannot be addressed further. Layers that have been removed from the selected potential layers are then eliminated. Finally, the remaining potential layers are recommended as the layers for remediation of the old well, providing a basis for subsequent re-perforation and other measures.

[0087] In existing technologies, the selection of intervention layers for old wells requires researchers to collect and review relevant wellbore engineering documents based on a comprehensive evaluation of the reservoir. This involves obtaining engineering parameters such as the perforation section and casing damage location, comparing the depth information obtained from these parameters with the depth of potential layers, and then recommending intervention layers. This existing method of reviewing old wells, which relies on researchers collecting various wellbore engineering documents and reviewing them well-by-well to obtain parameters, is extremely inefficient when reviewing a large number of wells and dealing with a heavy workload.

[0088] To address the technical problems mentioned above, this application provides an intelligent recommendation method for old well intervention layers. Based on knowledge entity extraction technology, it cleans, segments, annotates, and extracts parameters from the engineering parameter files of multiple potential layers within the wellbore. This directly obtains the objects within each potential layer's wellbore, along with the attribute and depth information of each wellbore. Finally, it directly compares the depth information of the potential layers with the depth information corresponding to the attribute information of the wellbore within the extracted potential layers, eliminating potential layers with overlapping depths and identifying the remaining potential layers as intervention layers. This application utilizes a cloud server to automatically extract key parameters from the engineering parameter files and compares the depth information in the obtained key parameters with the depth information of the potential layers, thereby improving the efficiency of old well review and effectively solving the problem of low recommendation efficiency for intervention layers when manually reviewing multiple old wells.

[0089] The intelligent recommendation method for old well measures provided in this application can be applied to cloud servers on cloud platforms. This application does not limit the application scenarios of this method.

[0090] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0091] Figure 1 The flowchart of the intelligent recommendation method for old well measures layer provided in this application is shown in the following figure. Figure 1 As shown, the method includes:

[0092] S101. Obtain the interpretation results data of multiple potential layers to be processed and the engineering parameter files of the wellbore where the multiple potential layers are located. The interpretation results data of each potential layer include depth information and interpretation layer position.

[0093] In this solution, the multiple potential layers to be processed are layers with exploitable potential that have been preliminarily identified by business personnel based on relevant historical data and new evaluation technologies. Furthermore, after preliminarily identifying multiple potential layers to be processed, business personnel can upload the relevant information of these potential layers to the terminal device and hand it over to the cloud server for processing. The cloud server performs text processing, parameter extraction, and depth comparison on the acquired information and finally recommends the appropriate action layer.

[0094] In this step, operational personnel can upload the interpretation results data of multiple potential layers to be processed, as well as the engineering parameter files of the wellbore containing each potential layer, to the terminal device. The cloud server can then obtain the basic data for recommending the measure layer, namely the interpretation results data of multiple potential layers and the engineering parameter files of the wellbore containing each potential layer. These multiple potential layers can originate from a single wellbore or multiple wellbores; this application does not impose any restrictions on this.

[0095] It should be understood that the interpretation data of potential layers includes the depth information and interpretation horizon of the potential layer. Therefore, the depth of the potential layer segment can be determined based on the interpretation data for subsequent depth comparison. Engineering parameter files refer to documents used to record and store parameters related to wellbore engineering during previous well construction and maintenance, such as engineering logging evaluation reports, well completion summary reports, cementing quality evaluation reports, and well testing results reports. Key information in engineering parameter files, such as perforation depth, casing damage location, and plunger location, is used to indicate which segments in older wells have been treated or cannot be treated.

[0096] In the specific implementation of this solution, for a single wellbore, business personnel can simultaneously upload multiple engineering parameter files for that wellbore, and the cloud server can correspondingly obtain these multiple engineering parameter files.

[0097] S102. Clean and segment the engineering parameter file of the wellbore where the potential layer is located to obtain the processed text data;

[0098] In this step, the obtained engineering parameter file needs to be preprocessed to remove unconventional characters and characters without practical meaning. Unconventional characters can include special symbols, control characters, and non-standard characters that do not conform to the regular text format; characters without practical meaning can include conjunctions, prepositions, pronouns, auxiliary words, quantifiers, interjections, and other words that do not have actual meaning.

[0099] In one specific implementation, the engineering parameter file of the wellbore containing the potential layer can be cleaned and segmented as follows to obtain processed text data:

[0100] The engineering parameter file of the wellbore containing the potential layer is cleaned to remove non-standard characters, resulting in cleaned text data. Then, based on a pre-defined main dictionary in the petroleum field, the cleaned text data is segmented into words, resulting in segmented text data. Based on a pre-defined stop word library, characters without actual semantic meaning in the segmented text data are removed, resulting in the processed text data. The stop word library includes multiple pre-defined characters without actual meaning.

[0101] In this embodiment, since the engineering parameter file of the wellbore involves many petroleum industry technical terms, the text data cannot be accurately segmented using a conventional dictionary. Therefore, it is necessary to pre-configure a petroleum industry-specific dictionary for segmentation by adding petroleum industry technical terms to improve the accuracy of segmentation and, consequently, the accuracy of subsequent extraction of knowledge entities from the text data. Furthermore, the pre-set stop word library can be a conventional one, such as a general Chinese stop word list or a petroleum industry stop word list.

[0102] S103. Based on the preset knowledge entity extraction model, the processed text data is labeled and extracted to obtain the objects in the wellbore where the potential layer is located, the attribute information and depth information of each wellbore where the potential layer is located.

[0103] In this step, based on a model capable of extracting knowledge entities, the processed text data is annotated, namely, object annotation, attribute annotation, and attribute value annotation, and the object, attribute information, and depth information of the wellbore containing the potential layer are extracted from the annotated text data.

[0104] Among them, object annotation refers to annotating the identification information (i.e., object) of the well shaft, which is used to distinguish the information extracted from the text corresponding to the well shaft from the information extracted from the text corresponding to other well shafts; for example, the object of the well shaft can be the well number; attribute annotation refers to annotating the attribute information in the text data, which is used to indicate the historical construction or maintenance type of the well shaft; attribute value annotation refers to annotating the depth information corresponding to the attribute information, which is used to indicate the depth of the historical construction or maintenance of the well shaft.

[0105] The knowledge entity extraction model used in this application can be a general text annotation and parameter extraction model, such as Hidden Markov Model, Conditional Random Field, Deep Learning Model (such as Convolutional Neural Network Model, Recurrent Neural Network, etc.) and Transfer Learning Model, etc., which can perform text annotation and parameter extraction. This application does not limit this.

[0106] In one specific implementation, a pre-trained Bidirectional Long Short-Term Memory-Conditional Random Field (Bi-LSTM-CRF) model can be used to label and extract objects, attributes, and attribute values ​​from the processed text data, thereby obtaining the objects in the wellbores containing potential layers, the attribute information of each wellbore containing potential layers, and the depth information. The detailed steps are as follows:

[0107] Step 1: Based on the pre-trained Bi-LSTM-CRF model, label the processed text data with objects, attributes, and attribute values ​​to obtain labeled text data;

[0108] In this scheme, the Bi-LSTM-CRF model is obtained by pre-training a pre-constructed Bi-LSTM-CRF model. The model structure consists of five layers: an input layer, a word embedding layer, a Bidirectional Long Short-Term Memory (Bi-LSTM) encoding layer, a Conditional Random Field (CRF) layer, and an output layer.

[0109] In the specific implementation of this scheme, during the model training phase, three metrics can be used to evaluate the model: precision, recall, and F1 score.

[0110] Step 2: Extract the objects of the wellbores containing potential layers and the attribute information and depth information of each wellbore containing a potential layer from the labeled text data; wherein, the objects of the wellbores containing potential layers are well numbers; the attribute information of each wellbore containing a potential layer includes at least one of perforation, casing damage, plunger and cementing quality.

[0111] In this step, based on a pre-trained Bi-LSTM-CRF model, the attribute information of the wellbore containing each potential layer extracted from the text data may include at least one of perforation, casing damage, plunger, and cementing quality. Depth information refers to the depth information corresponding to the attribute information, such as perforation depth, casing damage depth, plunger depth, and cementing quality depth.

[0112] It should be understood that in the specific implementation of this scheme, the perforation depth information can be extracted from the oil testing results report in the engineering parameter file of the wellbore where the potential layer is located; the depth information with the attribute of casing damage can be extracted from the engineering logging evaluation report in the engineering parameter file; the plunger depth information with the attribute of plunger can be extracted from the well completion summary report in the engineering parameter file; and the depth information of cement cementing quality can be extracted from the cement cementing quality evaluation report in the engineering parameter file.

[0113] S104. Based on the depth information and interpretation layer number of multiple potential layers, remove potential layers that have intersection with the attribute information and depth information of any potential layer in the wellbore, and determine the remaining potential layers as the measure layers.

[0114] In this step, for any potential layer, the depth information obtained from the interpretation results data is compared with the depth information corresponding to the wellbore attribute information of the potential layer obtained through parameter extraction. If the two depth information have an intersection, the potential layer is removed from the multiple potential layers, and the remaining potential layers that are not removed are identified as the measure layers.

[0115] For example, if the depth information of a potential layer is 1130-1135 meters, the attribute information and depth information extracted from the engineering parameter file of the wellbore containing this potential layer are: perforation depth 1204-1206 meters; casing damage depth 1134 meters; plunger depth 950 meters. Since the depth information of the potential layer overlaps with the depth of the casing damage attribute, this potential layer is removed from the list of potential layers.

[0116] The intelligent recommendation method for old well remediation layers provided in Embodiment 1 of this application first obtains the depth information of the interpretation results data of multiple potential layers to be processed, as well as the engineering parameter files of the wellbore where each potential layer is located. Based on knowledge entity extraction technology, the method cleans, segments, annotates, and extracts parameters from the engineering parameter files of the wellbore where each potential layer is located, directly obtaining the objects, attribute information, and depth information of the wellbore where each potential layer is located. Finally, the method compares the depth information of the potential layer with the depth information corresponding to the attribute information of the wellbore where the potential layer is located obtained by parameter extraction. If the two depth information have an intersection in any potential layer, the potential layer is removed, and the remaining potential layers are determined as remediation layer measures. The method automatically extracts the key parameters of the engineering parameter files and compares the depth information in the obtained key parameters with the depth information of the potential layers, thereby achieving the technical effect of improving the efficiency of old well review.

[0117] This application also provides a second embodiment of an intelligent recommendation method for old well measures layers. Based on the above embodiments, since the interpretation results data of each potential layer also include the values ​​of multiple evaluation indicators of the potential layer, the method provided in this application further includes:

[0118] Based on the values ​​of multiple evaluation indicators for each potential layer, the TOPSIS (Topology-Topology-Sorting-Ideal-Point) method is used to sort the multiple potential layers and determine the recommendation priority of each potential layer.

[0119] In this solution, the method provided in this application, in addition to recommending the measure layers as described in Example 1, can further rank multiple potential layers using the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) based on the values ​​of multiple evaluation indicators in the interpretation results data of each potential layer. This determines the recommendation priority of each potential layer, and the final recommended measure layers will inherit this recommendation priority. Business personnel can determine the mining order of multiple measure layers based on the recommendation priority of each measure layer.

[0120] It should be understood that ranking multiple potential layers based on the values ​​of multiple evaluation indicators in the interpreted data of each potential layer is a comprehensive consideration of the exploitation potential of these layers. Therefore, implementing exploitation and other measures on layers with higher recommended priority will yield greater beneficial effects.

[0121] In one specific implementation, the multiple potential layers are ranked using the Top-Approximation-to-Ideal-Point Ranking (TOPSIS) method based on the values ​​of multiple evaluation indicators for each potential layer, thereby determining the recommendation priority of each potential layer. Figure 2 The flowchart of Embodiment 2 of the intelligent recommendation method for old well measures layer provided in this application is as follows: Figure 2 As shown, the method specifically includes:

[0122] S201. Construct a decision matrix based on the values ​​of multiple evaluation indicators for each potential layer. The decision matrix includes the value of each evaluation indicator for each potential layer.

[0123] In this step, the evaluation indicators are pre-determined potential layer evaluation indicators based on theoretical analysis and expert experience; the values ​​of multiple evaluation indicators for each potential layer can be obtained from the interpreted data of each potential layer. The constructed decision matrix can be represented by the following formula:

[0124]

[0125] In the above formula, A represents the decision matrix; m represents the number of evaluation indicators; n represents the number of potential layers; f ij Let represent the value of the j-th evaluation index of the i-th potential layer, where i = 1, 2, ..., n, j = 1, 2, ..., m.

[0126] The value of any evaluation index for a given segment is obtained by averaging all the values ​​of that evaluation index within that segment.

[0127] For example, when recommending potential layers for low-saturation oil and gas reservoirs, the evaluation indicators for potential layers include at least three of the following: water cut, production, porosity, permeability, effective thickness, oil saturation, and formation pressure.

[0128] S202. Perform forward transformation and normalization on the decision matrix to obtain a standardized matrix;

[0129] In this step, when the evaluation indicators include both high-quality and low-quality indicators, it is necessary to perform trend-following transformation on the evaluation indicators, converting high-quality indicators into low-quality indicators or vice versa. Therefore, the decision matrix needs to be forward-oriented. Similarly, to facilitate subsequent calculations, the decision matrix also needs to be normalized to obtain a standardized matrix.

[0130] In one specific implementation, the decision matrix can be positiveized using the following method:

[0131] For each value f in decision matrix A ij For low-quality indicators whose values ​​correspond to relative numbers, their values ​​are processed using the reciprocal method, as shown in the following formula:

[0132]

[0133] This represents the value of the j-th evaluation index of the i-th potential layer after positive transformation.

[0134] For each value f in decision matrix A ij For low-quality indicators whose values ​​are relative, the values ​​are processed using the difference method, as shown in the following formula:

[0135]

[0136] For each value f in decision matrix A ij For high-quality indicators, the positive transformation of their values ​​is as follows:

[0137]

[0138] The above calculation method is used to calculate the value of each forward-processed element in decision matrix A. Then, the decision matrix after positive transformation is obtained.

[0139] In one specific implementation, the forward-processed decision matrix can be processed using the following method. Standardization process:

[0140] For the decision matrix after positive transformation The normalization of each value in the equation is as follows:

[0141]

[0142] In the above formula, This represents the value of the j-th evaluation index of the i-th potential layer after normalization.

[0143] Optionally, in one specific implementation, the matrix after positive transformation can be normalized based on the preset weight of each evaluation index.

[0144] Correspondingly, for the decision matrix after positive transformation The normalization of each value in the equation is as follows:

[0145]

[0146] In the above formula, w j This represents the weight of the j-th evaluation indicator.

[0147] Specifically, the weights for each evaluation indicator are pre-determined using the following method:

[0148]

[0149] Among them, a sj This indicates that the importance of the s-th evaluation indicator obtained in advance is relative to the j-th evaluation indicator, which can be obtained in advance by experts using the Santy scale method. This represents the weight vector of the j-th evaluation index.

[0150] In this implementation, by considering the importance of each evaluation indicator to the potential layer recommendation process, the rationality of the subsequently determined recommendation priorities is improved.

[0151] The decision matrix after forward processing is calculated using the above calculation method. Each normalized value Then, the standardized matrix is ​​obtained.

[0152] S203. The maximum value of each evaluation indicator in the standardization matrix is ​​determined as the positive ideal point, and the minimum value of each evaluation indicator is determined as the negative ideal point.

[0153] In one specific implementation, the positive and negative ideal points of each evaluation index are determined using the following formula:

[0154]

[0155] In the above formula, The set of positive ideal points representing the evaluation index; The set of negative ideal points representing the evaluation index; This represents the positive ideal point of the j-th evaluation index; This represents the negative ideal point of the j-th evaluation index.

[0156] S204. For each potential layer, calculate the first distance between the value of each evaluation index of the potential layer and the corresponding positive ideal point, and the second distance between the value of the corresponding negative ideal point;

[0157] In one specific implementation, the first distance and the second distance are calculated using the following formula:

[0158]

[0159]

[0160] In the above formula, The value of the j-th evaluation index of the i-th potential layer is the first distance from the corresponding positive ideal point. The value of the j-th evaluation index of the i-th potential layer is represented by the second distance between it and the corresponding negative ideal point.

[0161] S205. Based on the first distance and second distance corresponding to the value of each evaluation index of each potential layer, sort the multiple potential layers and determine the recommendation priority of each potential layer.

[0162] In one specific implementation, the distance from each potential layer to the positive ideal point and the distance from each potential layer to the negative ideal point are determined based on the first distance and the second distance corresponding to the value of each evaluation index of each potential layer. This can be calculated using the following formula:

[0163]

[0164] In the above formula, S i + S represents the distance from the i-th potential layer to the positive ideal point; i - This represents the distance from the i-th potential layer to the negative ideal point.

[0165] Furthermore, based on the distance from each potential layer to the positive ideal point and the distance from each potential layer to the negative ideal point, the recommendation coefficient (relative closeness to the positive ideal point) of each potential layer is determined, which can be calculated using the following formula:

[0166]

[0167] In the above formula, This represents the recommendation coefficient for the i-th potential layer.

[0168] Furthermore, based on the recommendation coefficient of each potential layer, multiple potential layers are ranked to determine the recommendation priority of each potential layer. Specifically, potential layers with a high relative closeness to the positive ideal point have a higher recommendation priority.

[0169] The intelligent recommendation method for old well intervention layers provided in Embodiment 2 of this application, based on Embodiment 1, further ranks the multiple potential layers using the Topology-Topology-Sorting (TOPSIS) method based on the values ​​of multiple evaluation indicators for each potential layer, thereby determining the recommendation priority of each potential layer. Based on a comprehensive consideration of multiple evaluation indicators, the exploitation potential of multiple potential layers is ranked, providing guidance for actual exploitation operations. Operational personnel can further select higher-priority intervention layers for exploitation and other measures based on their recommendation priorities, and can also determine the order in which to take exploitation and other measures for multiple intervention layers based on their recommendation priorities.

[0170] Embodiment 3 of this application provides an intelligent recommendation method for old well measure layers. Based on the above embodiments, the application scenarios of the intelligent recommendation method for old well measure layers provided in this application are as follows: Figure 3 As shown, in this application scenario, business personnel control terminal device 301 to send a recommendation request for old well measure layers to cloud server 302. Cloud server 302 processes the interpretation results data of multiple potential layers carried in the request and the engineering parameter files of the wellbore where each potential layer is located to obtain the old well measure layer recommendation results, including the recommended measure layers and the recommendation priority of each measure layer. Finally, the old well measure layer recommendation results are returned to terminal device 301 for business personnel to view and provide them with guidance.

[0171] Figure 4 This is a flowchart illustrating an intelligent recommendation method for old well management measures provided in Embodiment 3 of this application, as shown below. Figure 4 As shown in Embodiment 3 of this application, a method for intelligent recommendation of old well management measures also includes:

[0172] S401. Receive the old well measure layer recommendation request sent by the terminal equipment. The old well measure layer recommendation request includes the interpretation results data of multiple potential layers and the engineering parameter file of the wellbore where each potential layer is located.

[0173] In this step, the business personnel first upload the interpretation results data of multiple potential layers and the engineering parameter files of the wellbore where each potential layer is located through the terminal device. The terminal device then sends the corresponding old well measure layer recommendation request to the cloud server. This request includes the interpretation results data of multiple potential layers uploaded by the business personnel and the engineering parameter files of the wellbore where each potential layer is located.

[0174] S402. Based on the values ​​of multiple evaluation indicators for each potential layer, the TOPSIS method is used to rank the multiple potential layers and determine the recommendation priority of each potential layer.

[0175] S403. Clean and segment the engineering parameter file of the wellbore where the potential layer is located to obtain the processed text data.

[0176] S404. Based on the preset knowledge entity extraction model, the processed text data is labeled and extracted to obtain the objects in the wellbore where the potential layer is located, as well as the attribute information and depth information of the wellbore where each potential layer is located.

[0177] S405. Compare the depth information of multiple potential layers with the attribute information and depth information of any wellbore containing a potential layer.

[0178] S406. If the depth information of a potential layer intersects with the attribute information and depth information of any wellbore containing a potential layer, then that potential layer shall be removed.

[0179] S407. If the depth information of a potential layer does not intersect with the attribute information and depth information of any wellbore containing a potential layer, the potential layer shall be recommended as a measure layer.

[0180] The execution process of steps S402-S407 is similar to that of Embodiment 1 and Embodiment 2, and the effect is similar to that of the above embodiments. Therefore, this embodiment will not be described again.

[0181] S408. Return the recommended results of the old well measure layer to the terminal device. The recommended results of the old well measure layer include: the measure layer and the recommendation priority of each measure layer.

[0182] In this step, the cloud server returns the recommended results of the old well measure layer to the terminal device for business personnel to view and provide guidance. The recommended results of the old well measure layer include the measure layer and the recommendation priority of each measure layer.

[0183] It should be understood that in the specific implementation of this scheme, the recommended priority of each measure layer is an update and inheritance of the recommended priority of its corresponding potential layer. For example, step S303 determines the recommended priorities of segments A and B of wellbore 1 and segment C of wellbore 2 in the potential layers as level 1, level 2, and level 3, respectively. After processing in steps S304 to S305, segments A and C from multiple potential layers are recommended as measure layers, segment B is removed, and the priorities of segments A and C are updated to obtain the measure layer recommendation results and their corresponding priorities. Therefore, in the returned old well measure layer recommendation results, the recommended measure layers are: segment A of wellbore 1 and segment C of wellbore 2, with priorities of level 1 and level 2, respectively.

[0184] It should be understood that in the specific implementation of this solution, the recommended results for old well treatment layers may also include the depth of each treatment layer. By returning the depth of each treatment layer as a result to the terminal for business personnel to view, business personnel no longer need to search through the interpretation results data based on the wellbore number and layer number to find the depth information of the recommended treatment layers, thus improving the work efficiency of business personnel.

[0185] Embodiment 3 of this application provides an intelligent recommendation method for old well measure layers. The cloud server first receives the interpretation results data of multiple potential layers and the engineering parameter files of the wellbore where each potential layer is located, carried in the old well measure layer recommendation request sent by the terminal device. Based on the interpretation results data of multiple potential layers, the TOPSIS method is used to sort the multiple potential layers and determine the recommendation priority of each potential layer. Then, the engineering parameter files of the wellbore where each potential layer is located are cleaned and preprocessed by word segmentation. The preprocessed text data is then annotated and extracted to obtain the attribute information and depth information of the wellbore where each potential layer is located. Finally, based on the depth information of each potential layer obtained from the interpretation results data of multiple potential layers and the attribute information and depth information of the wellbore where each potential layer is located, a depth judgment is performed. Potential layers that have intersections with the attribute information and depth information of any potential layer in the wellbore are removed. The remaining potential layers are determined as measure layers. The recommended measure layers and their priorities are returned to the terminal as the old well measure layer recommendation results for business personnel to view, which improves the efficiency of measure layer selection.

[0186] The following is a specific embodiment five of the intelligent recommendation method for old well measure layers provided in this application.

[0187] In this embodiment, the evaluation indicators (potential layer evaluation indicators) include: water cut, production rate, sublayer permeability, sublayer porosity, sublayer oil saturation, sublayer effective thickness, sublayer formation pressure, and sublayer prediction. Among these, sublayer permeability, sublayer porosity, sublayer effective thickness, sublayer oil saturation, and sublayer formation pressure are all high-optimal indicators; while water cut is a low-optimal indicator. The potential layers pre-determined by the operations personnel are: the section with layer number G16+7 in the wellbore of Well-5; the section with layer number S351 in the wellbore of Well-7; the section with layer number G16+5 in the wellbore of Well-4; and the section with layer number S2101 in the wellbore of Well-2.

[0188] The specific execution steps are as follows:

[0189] Step 1: Receive an old well intervention layer recommendation request sent by the terminal device. The old well intervention layer recommendation request includes interpretation results data of the multiple potential layers and engineering parameter files of the wellbore where each potential layer is located. The interpretation results data includes values ​​of multiple evaluation indicators; the engineering parameter files of each wellbore where the potential layer is located include engineering logging evaluation reports, well completion summary reports, and well testing results reports for each wellbore.

[0190] Specifically, the potential layer depth and values ​​of multiple evaluation indicators are obtained from the interpretation results data of multiple potential layers, as shown in Table 1.

[0191] Table 1

[0192]

[0193] Step 3: Using the TOPSIS method to approximate the ideal point ranking, calculate the layer recommendation coefficient for each potential layer, and derive the recommendation priority for each potential layer based on the ranking results, as shown in Table 2.

[0194] Table 2

[0195] hashtag Floor number Recommendation coefficient for potential layer Priority Well-5 G16+7 0.5699 1 Well-2 S2101 0.4992 2 Well-7 S351 0.4952 3 Well-4 G16+5 0.3827 4

[0196] Step 2: Clean and segment the engineering parameter files of the wellbore containing the potential layer to obtain processed text data. Based on a preset knowledge entity extraction model, label and extract objects, attributes, and attribute values ​​from the processed text data to obtain the objects of the wellbore containing the potential layer, as well as the attribute information and depth information of each wellbore containing the potential layer. The extracted wellbore objects, attribute information, and depth information are shown in Table 3.

[0197] Table 3

[0198] hashtag Perforation depth depth of loss plunger depth Well-5 1100-1105 967 1050 Well-2 1235-1240 / 1148 Well-7 1204-1206 1134 950 Well-4 1209-1218 / 1350

[0199] Step 3: Based on the depth information of multiple potential layers, remove potential layers that have intersections with the attribute information and depth information of any wellbore containing any potential layer, and determine the remaining potential layers as the measure layers.

[0200] After comparative calculations, it was determined that the depth of segment S2101 in well-2 overlaps with the perforation depth of well-2 within the identified potential layers; the depth of segment S351 in well-7 overlaps with the casing damage depth of well-7; the depths of the remaining segments do not overlap with the perforation depth, casing damage depth, or plunger depth in their respective wells. Therefore, segments S2101 in well-2 and S351 in well-7 are removed from the potential layers. Segments G16+7 in well-5 and G16+5 in well-4 are recommended as treatment layers.

[0201] Step 4: Return the recommended results of the old well measure layer to the terminal device. The recommended results of the old well measure layer include: the measure layer and the recommendation priority of each measure layer.

[0202] Specifically, the recommended results for the old well measures layer include:

[0203] The recommended treatment layers are: the G16+7 section of Well-5, with an operating priority of 1; and the G16+5 section of Well-4, with an operating priority of 2.

[0204] This application provides a specific embodiment five of an intelligent recommendation method for old well intervention layers. Based on the TOPSIS method, it processes the interpretation results data of multiple potential layers to obtain the recommendation priority of the potential layers. Furthermore, based on a knowledge entity extraction model, it extracts parameters from the engineering parameter files of the wellbore containing each potential layer, obtaining the depth information of each potential layer, as well as the attribute and depth information of the wellbore containing each potential layer. Finally, after depth comparison, intervention layers are recommended. This method improves the efficiency of selecting intervention layers for old wells.

[0205] Figure 5 This is a schematic diagram of the structure of an embodiment of the intelligent recommendation device for old well management layers provided in this application, as shown below. Figure 5 As shown, the intelligent recommendation device 50 for old well treatment layers provided in this embodiment includes:

[0206] The data input module 501 is used to acquire the interpretation results data of multiple potential layers to be processed and the engineering parameter file of the wellbore where each potential layer is located. The interpretation results data of each potential layer includes depth information.

[0207] The model processing module 502 is used to clean and segment the engineering parameter file of the wellbore where the potential layer is located to obtain processed text data;

[0208] The text annotation and extraction module 503 is used to annotate and extract objects, attributes and attribute values ​​from the processed text data based on a preset knowledge entity extraction model, so as to obtain the objects of the wellbore where the potential layer is located and the attribute information and depth information of each wellbore where the potential layer is located.

[0209] The measure layer determination module 504 is used to eliminate potential layers that have intersections with the attribute information and depth information of any wellbore containing any potential layer, based on the depth information of the multiple potential layers, and determine the remaining potential layers as measure layers.

[0210] In one possible implementation, the data input module 501 is specifically used for:

[0211] The terminal device receives a request for recommendation of old well measures layers, which includes interpretation results data of the multiple potential layers and engineering parameter files of the wellbore where each potential layer is located.

[0212] In one possible implementation, the model processing module 502 is specifically used for:

[0213] The engineering parameter file of the wellbore containing the potential layer is cleaned to remove non-standard characters from the text, resulting in cleaned text data.

[0214] Based on a pre-defined main dictionary in the petroleum field, the cleaned text data is segmented into words to obtain segmented text data.

[0215] Based on a preset stop word library, characters without actual semantic meaning are removed from the segmented text data to obtain the processed text data; wherein, the stop word library includes multiple preset characters without actual meaning.

[0216] In one possible implementation, the text annotation and extraction module 503 is specifically used for:

[0217] Based on the pre-trained Bi-LSTM-CRF model, the processed text data is labeled with objects, attributes, and attribute values ​​to obtain labeled text data.

[0218] Extract the objects of the wellbore containing the potential layer and the attribute and depth information of each wellbore containing the potential layer from the labeled text data;

[0219] The object of the wellbore where the potential layer is located is the well number;

[0220] The property information for each potential layer in the wellbore includes at least one of perforation, casing damage, plunger, and cementing quality.

[0221] The intelligent recommendation device for old well measures layer provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0222] Figure 6 This is a schematic diagram of the structure of Embodiment 2 of the intelligent recommendation device for old well measures layer provided in this application, as shown below. Figure 6 As shown, the intelligent recommendation device 50 for old well treatment layers provided in this embodiment includes:

[0223] The sorting module 505 is used to sort the multiple potential layers according to the depth information of each potential layer and the values ​​of multiple evaluation indicators, using the TOPSIS method of approximating the ideal point, and to determine the recommendation priority of each potential layer.

[0224] Accordingly, this application embodiment provides an intelligent recommendation device 50 for old well management layers, which further includes:

[0225] Result return module 506: used to return the old well measure layer recommendation result to the terminal device, the old well measure layer recommendation result includes: measure layer and recommendation priority of each measure layer.

[0226] In one possible implementation, the sorting module 505 is specifically used for:

[0227] A decision matrix is ​​constructed based on the values ​​of multiple evaluation indicators for each potential layer, and the decision matrix includes the value of each evaluation indicator for each potential layer.

[0228] The decision matrix is ​​then subjected to forward transformation and normalization to obtain a standardized matrix;

[0229] The maximum value of each evaluation index in the standardized matrix is ​​determined as the positive ideal point, and the minimum value of each evaluation index is determined as the negative ideal point.

[0230] For each potential layer, calculate the first distance between the value of each evaluation index of the potential layer and the corresponding positive ideal point, and the second distance between the value of each index and the corresponding negative ideal point;

[0231] Based on the first and second distances corresponding to the values ​​of each evaluation index in each potential layer, the multiple potential layers are sorted to determine the recommendation priority of each potential layer.

[0232] In one possible implementation, the sorting module 505 is specifically used for:

[0233] The decision matrix is ​​then forward-oriented to obtain the forward-oriented matrix.

[0234] Based on the preset weights of each evaluation index, the matrix after positive transformation is normalized to obtain a standardized matrix.

[0235] The intelligent recommendation device for old well measures layer provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0236] Figure 7 A schematic diagram of the intelligent recommendation device for the old well treatment layer provided in this application. (See attached diagram.) Figure 7 As shown, the electronic device 60 provided in this embodiment includes at least one processor 601 and a memory 602. Optionally, the device 60 further includes a communication component 603. The processor 601, memory 602, and communication component 603 are connected via a bus 604.

[0237] In a specific implementation, at least one processor 601 executes computer execution instructions stored in memory 602, causing at least one processor 601 to perform the above-described method.

[0238] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0239] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0240] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0241] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0242] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0243] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0244] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0245] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0246] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0247] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0248] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0249] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0250] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0251] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for intelligently recommending treatment layers in old wells, characterized in that, include: Acquire the interpretation results data of multiple potential layers to be processed and the engineering parameter files of the wellbore where each potential layer is located. The interpretation results data of each potential layer includes depth information. The engineering parameter file of the wellbore containing the potential layer is cleaned and segmented to obtain processed text data; Based on a preset knowledge entity extraction model, the processed text data is labeled and extracted for objects, attributes and attribute values ​​to obtain the objects in the wellbore where the potential layer is located, as well as the attribute information and depth information of each wellbore where the potential layer is located. Based on the depth information of the multiple potential layers, potential layers that intersect with the attribute information and depth information of any wellbore containing any potential layer are removed, and the remaining potential layers are determined as the measure layers.

2. The method according to claim 1, characterized in that, The interpretation results data for each potential layer also include: the values ​​of multiple evaluation indicators for that potential layer; The method further includes: Based on the values ​​of multiple evaluation indicators for each potential layer, the TOPSIS (Topology-Topology-Sorting-Ideal-Point) method is used to sort the multiple potential layers and determine the recommendation priority of each potential layer.

3. The method according to claim 2, characterized in that, The process of acquiring interpretation results data for multiple potential layers to be processed and engineering parameter files for the wellbore containing each potential layer includes: The terminal device receives a request for recommendation of old well measures layers, which includes interpretation results data of the multiple potential layers and engineering parameter files of the wellbore where each potential layer is located. Accordingly, the method further includes: The old well measure layer recommendation results are returned to the terminal device. The old well measure layer recommendation results include: measure layer and recommendation priority of each measure layer.

4. The method according to any one of claims 1 to 3, characterized in that, The process of cleaning and segmenting the engineering parameter file of the wellbore containing the potential layer yields processed text data, including: The engineering parameter file of the wellbore containing the potential layer is cleaned to remove non-standard characters from the text, resulting in cleaned text data. Based on a pre-defined main dictionary in the petroleum field, the cleaned text data is segmented into words to obtain segmented text data. Based on a preset stop word library, characters without actual semantic meaning are removed from the segmented text data to obtain the processed text data; wherein, the stop word library includes multiple preset characters without actual meaning.

5. The method according to any one of claims 1 to 3, characterized in that, The pre-defined knowledge entity extraction model annotates and extracts objects, attributes, and attribute values ​​from the processed text data to obtain the objects in the wellbore containing the potential layer, the attribute information of each wellbore containing the potential layer, and the depth information, including: Based on the pre-trained Bi-LSTM-CRF model, the processed text data is labeled with objects, attributes, and attribute values ​​to obtain labeled text data. Extract the objects of the wellbore containing the potential layer and the attribute and depth information of each wellbore containing the potential layer from the labeled text data; The object of the wellbore where the potential layer is located is the well number; The property information for each potential layer in the wellbore includes at least one of perforation, casing damage, plunger, and cementing quality.

6. The method according to claim 2, characterized in that, The step involves ranking the multiple potential layers using the TOPSIS (Topology-Topology-Sorting-Ideal-Point) method based on the values ​​of multiple evaluation indicators for each potential layer, and determining the recommendation priority for each potential layer, including: A decision matrix is ​​constructed based on the values ​​of multiple evaluation indicators for each potential layer, and the decision matrix includes the value of each evaluation indicator for each potential layer. The decision matrix is ​​then subjected to forward transformation and normalization to obtain a standardized matrix; The maximum value of each evaluation index in the standardized matrix is ​​determined as the positive ideal point, and the minimum value of each evaluation index is determined as the negative ideal point. For each potential layer, calculate the first distance between the value of each evaluation index of the potential layer and the corresponding positive ideal point, and the second distance between the value of each index and the corresponding negative ideal point; Based on the first and second distances corresponding to the values ​​of each evaluation index in each potential layer, the multiple potential layers are sorted to determine the recommendation priority of each potential layer.

7. The method according to claim 6, characterized in that, The decision matrix is ​​subjected to forward transformation and normalization to obtain a standardized matrix, including: The decision matrix is ​​then forward-oriented to obtain the forward-oriented matrix. Based on the preset weights of each evaluation index, the matrix after positive transformation is normalized to obtain a standardized matrix.

8. A smart recommendation device for old well treatment layers, characterized in that, include: The data input module is used to acquire the interpretation results data of multiple potential layers to be processed and the engineering parameter files of the wellbore where each potential layer is located. The interpretation results data of each potential layer includes depth information. The model processing module is used to clean and segment the engineering parameter file of the wellbore where the potential layer is located to obtain processed text data; The text annotation and extraction module is used to annotate and extract objects, attributes and attribute values ​​from the processed text data based on a preset knowledge entity extraction model, so as to obtain the objects of the wellbore where the potential layer is located and the attribute information and depth information of each wellbore where the potential layer is located. The measure layer determination module is used to eliminate potential layers that have intersections with the attribute information and depth information of any wellbore containing any potential layer, based on the depth information of the multiple potential layers, and determine the remaining potential layers as measure layers.

9. A smart recommendation device for old well treatment layers, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images