Crack piece extraction method, electronic equipment, storage medium and device

Abstract

The invention discloses a crack piece extraction method, electronic equipment, a storage medium and a device. The method comprises the steps of performing conjoint analysis based on far sound wave data, logging data and three-dimensional seismic data, and determining a well seismic position; performing series attribute extraction based on the far sound wave data to obtain texture attributes and gradient structure tensor attributes; carrying out series attribute extraction based on the three-dimensional seismic data to obtain an ant body; extracting ant body crack pieces based on ant bodies; performing three-dimensional comprehensive analysis on the ant body based on texture attributes and gradient structure tensor attributes, and determining crack development positions; and constructing a three-dimensional geologic model, and extracting crack slices based on the three-dimensional geologic model. According to the method, collective characterization at the fractures and the vugles is achieved, collective characterization results at the fractures and the vugles can be improved, anomalies are more focused to a certain extent, the theory is rigorous and reliable, the operation process is simple and practical, the method is a requirement of complex reservoir exploration situation and a key step of well position arrangement, and the method has important significance on prediction and subsequent processing of fracture and vugles in complex work areas.

Description

Technical Field

[0001] This invention belongs to the field of geophysical exploration technology, and more specifically, relates to a method for extracting fracture sections, an electronic device, a storage medium, and an apparatus. Background Technology

[0002] With the increasing demand for oil, natural gas, and mineral exploration and development, and the advancement of logging technology, there is a need to detect geological structures and bodies (such as stratigraphic interfaces, fractures, faults, caverns, and ore bodies) within a range of several meters to tens of meters around the well. The technique for measuring these geological structures within the well is called remote detection technology. It increases the measurement range of conventional logging technology from about one meter around the well to tens of meters, with enormous application potential. This technology can be used to display geological interfaces intersecting with the well; it can be used to detect inclined stratigraphic interfaces, fractures, faults, pinch-outs, and internal structures of salt domes near the well; and it can be used to trace oil reservoir boundaries in horizontal wells. Fractured reservoirs are an important type of oil and gas reservoir in my country's oil and gas basins. These reservoirs generally have complex fracture systems, which not only affect the accumulation and migration of oil and gas but also have a significant impact on production and development at different stages. Characterizing fractured-cavitary reservoirs through remote acoustic data-based fracture-cavitary characterization technology is a key step in describing the needs of complex reservoir exploration and well location planning, and it is of great significance for predicting and subsequently processing fractured-cavitary bodies in complex work areas.

[0003] The information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to propose a method, electronic device, storage medium, and apparatus for extracting fracture fragments, thereby improving the characterization effect of fracture-cavity reservoirs based on far-sound data.

[0005] To achieve the above objectives, the present invention provides a method for extracting cracked sections, an electronic device, a storage medium, and an apparatus.

[0006] According to a first aspect of the present invention, a method for extracting cracked sections is provided, comprising:

[0007] The location of the well-seismic event was determined by joint analysis of far-sonic data, well logging data, and 3D seismic data.

[0008] Based on the far-sound data, concatenated attribute extraction is performed to obtain texture attributes and gradient structure tensor attributes.

[0009] Based on the aforementioned 3D seismic data, concatenated attribute extraction is performed to obtain the ant body;

[0010] Based on the described ant body, extract ant body fissures;

[0011] Based on the texture attributes and gradient structure tensor attributes, a three-dimensional comprehensive analysis of the ant body was performed to clarify the location of crack development.

[0012] A three-dimensional geological model was constructed, and fracture sections were extracted based on the three-dimensional geological model.

[0013] Optionally, the series attribute extraction based on the far-sound data includes:

[0014] The far-sound data is preprocessed based on two-dimensional empirical mode decomposition.

[0015] The texture attributes are extracted based on the preprocessed far-sound data and gray-level co-occurrence matrix;

[0016] The gradient structure tensor properties are extracted based on the preprocessed far-sound data.

[0017] Optionally, the concatenation attribute extraction based on the seismic data includes:

[0018] Chaotic attribute volumes are extracted based on the earthquake data;

[0019] Extract the variance attribute volume based on the chaotic attribute volume;

[0020] Ant bodies are extracted based on the variance attribute data.

[0021] Optionally, the step of constructing a three-dimensional geological model and extracting fracture sections based on the three-dimensional geological model includes:

[0022] A large-scale discrete distribution model of cracks is constructed based on crack slices in ants, and then a small-scale discrete distribution model of cracks is established.

[0023] The tracking parameters of the large-scale crack discrete distribution model and the small-scale crack discrete distribution model are set based on the texture properties and gradient structure tensor properties.

[0024] Based on geological knowledge and various attribute slices, the fault slices of the large-scale and small-scale fracture discrete distribution models are screened and edited to complete the extraction of the three-dimensional fracture model.

[0025] Optionally, the expression for the gray-level co-occurrence matrix is:

[0026] p(i,j)=∑{g(x,y)=ig(x+Δx,y+Δy)=j};

[0027] Where x and y are the horizontal and vertical coordinates of the seismic profile data, respectively, x = 0, 1, ..., L-1, where L is the number of gray levels in the seismic profile, and g(x,y) is the gray value at (x,y). Δx represents the increment of seismic data in the x-direction, and Δy represents the increment of seismic data in the y-direction.

[0028] Optionally, the extraction of gradient structure tensor properties based on the preprocessed far-sound data includes:

[0029] Based on specific filter operators, the preprocessed far-sound data is processed to obtain far-sound seismic data volume;

[0030] Calculate the gradient structure tensor based on the aforementioned far-sonic seismic data volume;

[0031] The gradient structure tensor is reconstructed to obtain the reconstructed gradient structure tensor;

[0032] By using different Gaussian window parameters, a smoothing test is performed on each component of the reconstructed gradient structure tensor to obtain the optimal reconstructed gradient structure tensor.

[0033] The eigenvalues ​​and eigenvectors of the gradient structure tensor are calculated based on the optimal reconstructed gradient structure tensor.

[0034] Optionally, the extraction of ant bodies based on the variance attribute body includes:

[0035] Ant bodies are extracted from the variance attribute body using at least one passive extraction method combined with at least one active extraction method.

[0036] According to a second aspect of the present invention, a slit extraction apparatus is provided, comprising:

[0037] The joint analysis module is used to perform joint analysis based on far-sonic data, well logging data, and 3D seismic data to determine the location of well-seismic events.

[0038] The first attribute extraction module is used to perform concatenated attribute extraction based on the far-sound data to obtain texture attributes and gradient structure tensor attributes.

[0039] The second attribute extraction module is used to extract serial attributes based on the three-dimensional seismic data to obtain the ant body;

[0040] The first extraction module is used to extract ant body fissures based on the ant body;

[0041] The three-dimensional comprehensive analysis module performs a three-dimensional comprehensive analysis of the ant body based on the texture attributes and gradient structure tensor attributes to clarify the location of crack development.

[0042] Modules for building three-dimensional geological models;

[0043] The second extraction module is used to extract fracture sections based on the three-dimensional geological model.

[0044] According to a third aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0045] At least one processor; and,

[0046] A memory communicatively connected to the at least one processor; wherein,

[0047] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the crack extraction method according to any of the first aspects.

[0048] According to a fourth aspect of the invention, a non-transitory computer-readable storage medium is provided, which stores computer instructions for causing a computer to perform the slit extraction method described in any of the first aspects.

[0049] The beneficial effects of this invention are as follows: This invention extracts texture attributes and gradient structure tensor attributes from far-sound data, and combines them with ant-like structures extracted from seismic data to achieve collective characterization of fractures and cavities. This can improve the collective characterization results of fractures and cavities, and to a certain extent make anomalies more focused. The theory is rigorous and reliable, and the operation process is simple and practical. It is a key step in describing the needs of complex reservoir exploration and well location layout, and it is of great significance for the prediction and subsequent processing of fractures and cavities in complex work areas.

[0050] The system of the present invention has other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description

[0051] The above and other objects, features and advantages of the present invention will become more apparent from the accompanying drawings, in which like reference numerals generally denote like parts.

[0052] Figure 1 A flowchart illustrating the steps of the crack fragment extraction method according to the present invention is shown.

[0053] Figure 2 a, Figure 2 b、 Figure 2 c and Figure 2Figure d shows a schematic diagram of far-sound detection data according to Embodiment 2 of the present invention and the reflection residual intensity attribute, amplitude envelope attribute and relative impedance attribute extracted from the anomaly point of the far-sound detection data.

[0054] Figure 3 a, Figure 3 b、 Figure 3 c and Figure 3 Figure d shows a schematic diagram of the extraction results of frequency and geometric attributes based on far-sound wave detection data according to Embodiment 2 of the present invention.

[0055] Figure 4 a and Figure 4 b shows a schematic diagram of the contrast before and after texture attribute extraction based on far-sound detection data according to Embodiment 2 of the present invention.

[0056] Figure 5 a and Figure 5 b shows a schematic diagram before and after gradient structure tensor attribute extraction based on far-sound wave detection data according to Embodiment 2 of the present invention.

[0057] Figure 6 A schematic diagram is shown of the inversion results of three-dimensional wave impedance inversion identification of ground seismic fractures and cavities based on the fracture sheet extraction method of this embodiment according to Embodiment 2 of the present invention. Detailed Implementation

[0058] The invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0059] like Figure 1 As shown, a method for extracting cracked sections according to the present invention includes:

[0060] The location of the well-seismic event was determined by joint analysis of far-sonic data, well logging data, and 3D seismic data.

[0061] Based on far-sound data, concatenated attribute extraction is performed to obtain texture attributes and gradient structure tensor attributes.

[0062] Ant bodies were obtained by extracting concatenated attributes from 3D seismic data.

[0063] Extracting ant body slits based on ant body structure;

[0064] Three-dimensional comprehensive analysis of ant bodies was performed based on texture attributes and gradient structure tensor attributes to clarify the location of crack development.

[0065] A three-dimensional geological model was constructed, and fracture sections were extracted based on the three-dimensional geological model.

[0066] Specifically, this invention uses joint analysis of acoustic wave data, well logging data, and 3D seismic data to determine well-seismic locations. It extracts cascaded attributes from acoustic wave data to obtain texture and gradient structure tensor attributes; it extracts cascaded attributes from 3D seismic data to obtain ant-like structures; it extracts fracture fragments from these ant-like structures; it performs 3D comprehensive analysis of the ant-like structures based on texture and gradient structure tensor attributes to clarify fracture development locations; it constructs a 3D geological model and extracts fracture fragments based on this model. This enables collective characterization of fracture-cavity areas, improving the results and making anomalies more focused. The theory is rigorous and reliable, and the operation is simple and practical. It is a key step in addressing the needs of complex reservoir exploration and well location planning, and is of great significance for predicting and processing fracture-cavity areas in complex work zones.

[0067] In one example, concatenated attribute extraction based on farsound data includes:

[0068] Preprocessing of far-sound data based on two-dimensional empirical mode decomposition;

[0069] Texture attributes are extracted based on preprocessed far-sound data and gray-level co-occurrence matrix;

[0070] Gradient structure tensor properties are extracted from preprocessed farsound data.

[0071] Specifically, the BEMD (Two-Dimensional Empirical Mode Decomposition) feature enhancement algorithm was used to preprocess the far-sound data, suppressing strong reflections from seismic profiles. Texture and structural tensor attribute extraction, as image processing techniques, more intuitively reflects the overall characteristics of the data. Extending this approach to the field of geophysics can serve as an important basis for seismic data interpretation; Gao et al. [12-14] The computation of texture properties has been applied in tomographic identification, as demonstrated by Knutsson et al.

[15] The concept of structural tensor was proposed and applied to attribute extraction of different geological bodies, by Bakker, Randen, and others. [16-18] Humans have extended the gradient structure tensor to anisotropic parameters and chaotic properties, as exemplified by Zhang Junhua et al. [19-20] Texture attributes and structural tensor attributes were applied in the corresponding work area. By calculating texture attributes and structural tensor attributes, the application of seismic attribute texture analysis in far-field acoustic detection was explored. Texture attribute extraction, Haralick et al.

[21] This paper proposes a method for calculating texture attributes using gray-level co-occurrence matrix (GLCM). By converting the image to grayscale, the similarity of pixels at a certain distance and direction is statistically analyzed to reflect the image's variation characteristics in different directions and intervals. Using the seismic response feature recognition results from BEMD signal decomposition as input, the image texture feature calculation method is extended to seismic texture attributes. Specifically, the gray-level correlation between pixels is used to realize the amplitude of the seismic data volume. Texture attributes are calculated using GLCM from different angles, gray levels, directions, and windows. By converting the image to grayscale and statistically analyzing the similarity of pixels at a certain distance and direction, the variation characteristics of the image in different directions and intervals are reflected. Applying this to seismic data processing, for a seismic profile described by grayscale, the probability of any two data points with a distance δ along a certain direction satisfying certain conditions is calculated as follows: p(i,j)=∑{g(x,y)=ig(x+Δx,y+Δy)=j}, x=0,1…L-1, where L is the grayscale level of the seismic profile; x and y are the horizontal and vertical coordinates of the seismic profile data, respectively; and g(x,y) is the grayscale value at (x,y) of the seismic profile. Δx represents the increment of seismic data in the x-direction, and Δy represents the increment of seismic data in the y-direction. The gray-level co-occurrence matrix (GLCM) statistically summarizes profile information. To intuitively obtain the changes in profile gray levels, i.e., the data variations, texture attribute values ​​derived from the GLCM are calculated: a. Energy: Energy is a measure of the uniformity of texture gray-level changes, reflecting the uniformity of gray-level distribution and texture coarseness. b. Entropy: Entropy measures the randomness of a texture. It reaches its maximum value when all values ​​in the co-occurrence matrix are equal; conversely, it is smaller if the values ​​in the co-occurrence matrix are highly uneven. c. Contrast ratio: Contrast ratio is the moment of inertia near the main diagonal of the gray-level co-occurrence matrix. It measures the distribution and local variations of matrix values ​​and reflects sharpness and the depth of texture grooves.

[0072] In one example, concatenated attribute extraction based on seismic data includes:

[0073] Extracting chaotic attribute volumes from seismic data;

[0074] Extracting variance attribute volume based on chaotic attribute volume;

[0075] Extracting ant bodies based on variance attribute volume.

[0076] In one example, the process of constructing a three-dimensional geological model and extracting fracture sections based on that model includes:

[0077] A large-scale discrete distribution model of cracks is constructed based on crack slices in ants, and then a small-scale discrete distribution model of cracks is established.

[0078] The tracking parameters for the large-scale crack discrete distribution model and the small-scale crack discrete distribution model are set based on texture attributes and gradient structure tensor attributes.

[0079] Based on geological knowledge and various attribute slices, fault sections of large-scale and small-scale fracture discrete distribution models are screened and edited to complete the extraction of three-dimensional fracture models.

[0080] Specifically, effective crack identification is a prerequisite for establishing an accurate three-dimensional geological model of cracks. Crack slices are extracted from attribute volumes to establish a large-scale discrete crack distribution model, and then a small-scale discrete crack distribution model is established. Tracking parameters are set based on the results of farsonic data attributes. Based on geological understanding and various attribute volume slices, fault slices are screened and edited to achieve the extraction of a three-dimensional fracture model.

[0081] In one example, the expression for the gray-level co-occurrence matrix is:

[0082] p(i,j)=∑{g(x,y)=ig(x+Δx,y+Δy)=j};

[0083] Where x and y are the horizontal and vertical coordinates of the seismic profile data, respectively, x = 0, 1...L-1, L is the number of gray levels in the seismic profile, and g(x,y) is the gray value at (x,y). Δx represents the increment of seismic data in the x-direction, and Δy represents the increment of seismic data in the y-direction.

[0084] In one example, the gradient structure tensor properties extracted from preprocessed farsound data include:

[0085] Based on specific filter operators, signal processing is performed on the preprocessed far-sound data to obtain far-sound seismic data volume;

[0086] Calculation of gradient structure tensor based on farsound seismic data volume;

[0087] The gradient structure tensor is reconstructed to obtain the reconstructed gradient structure tensor;

[0088] By using different Gaussian window parameters, a smoothing test is performed on each component of the reconstructed gradient structure tensor to obtain the optimal reconstructed gradient structure tensor.

[0089] The eigenvalues ​​and eigenvectors of the gradient structure tensor are calculated based on the optimal reconstructed gradient structure tensor.

[0090] Specifically, gradient structure tensor attribute extraction and analysis is a novel attribute analysis method introduced from image processing into seismic interpretation in recent years. It identifies different structural features or texture units (such as layered textures, chaotic textures, etc.) in seismic images. Texture variations in images actually represent anomalies in geological targets such as faults and fissures, and the degree of variation can be expressed by feature values ​​in the XYZ directions. The Gst (gradient structure tensor) is mainly used to calculate seismic tectonic attributes and predict fracture development zones. Based on existing algorithms, the extraction results of Gst attributes have been developed and optimized, specifically through the following steps: b. Signal processing of the profile using specific filter operators to enhance the boundary of far-sonic detection data; b. Calculation of the gradient structure tensor: Calculating the directional derivative of the gradient vector at each point in the 3D seismic data volume:

[0091]

[0092]

[0093] in, For the gradient operator, σ g Here, μ(x) is the scale parameter, x is the x-coordinate of the seismic data, y is the y-coordinate of the seismic data, and z is the z-coordinate of the seismic data.

[0094] Reconstructing the gradient structure tensor T

[22] (Bakker, 2002):

[0095]

[0096] d. Here, different Gaussian window parameters are used to flatten each component of the gradient structure tensor.

[0097] Slip test, select the optimal result:

[0098]

[0099] in, For the average gradient structure tensor result, T ij For gradient structure tensors, G(x,σ) is the convolution operator. T ) is the Gaussian kernel function. denoted as the average directional derivative in different directions.

[0100] e. Calculate the eigenvalues ​​and eigenvectors of the gradient structure tensor:

[0101] |Tυ-λυ|=0;

[0102] Where υ is the eigenvector and λ is the eigenvalue.

[0103] Gradient structure tensor attribute extraction and analysis were performed on the enhanced far-sound data. Based on the geological and geophysical characteristics of the far-sound detection data, the attributes extracted by Gst have a certain reflection of the underground geological bodies.

[0104] In one example, extracting ant bodies based on variance attribute bodies includes:

[0105] Ants are extracted from variance attribute data using at least one passive extraction method combined with at least one active extraction method.

[0106] Specifically, chaotic attribute volumes are extracted from the seismic data volume of the target area, and variance attribute volumes are extracted from these chaotic attribute volumes. Ant bodies are then extracted from the variance attribute volumes using at least one passive extraction method combined with at least one active extraction method. Based on these ant bodies, ant body fracture slices are extracted. Ant bodies are the data volumes extracted using the ant algorithm. Passive and active extraction methods represent the passive and active algorithms within the ant algorithm, respectively. The passive algorithm is a conservative method for estimating fractures, only estimating along the direction of stronger signal. The active algorithm allows ants to detect fractures in a more flexible way, and is effective for detecting both large and small fractures. In the process of fracture slice extraction, a simple single extraction of ant bodies can reflect the three-dimensional distribution characteristics of fractures to some extent, but the signal is poor and the continuity is poor, failing to accurately reflect the actual distribution of fracture slices.

[0107] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the invention. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present invention can be combined with each other.

[0108] Example 1

[0109] This embodiment provides a method for extracting fractured sections, including:

[0110] The location of the well-seismic event was determined by joint analysis of far-sonic data, well logging data, and 3D seismic data.

[0111] Based on far-sound data, concatenated attribute extraction is performed to obtain texture attributes and gradient structure tensor attributes.

[0112] Ant bodies were obtained by extracting concatenated attributes from 3D seismic data.

[0113] Extracting ant body slits based on ant body structure;

[0114] Three-dimensional comprehensive analysis of ant bodies was performed based on texture attributes and gradient structure tensor attributes to clarify the location of crack development.

[0115] A three-dimensional geological model was constructed, and fracture sections were extracted based on the three-dimensional geological model.

[0116] Serial attribute extraction based on far-sound data includes:

[0117] Preprocessing of far-sound data based on two-dimensional empirical mode decomposition;

[0118] Texture attributes are extracted based on preprocessed far-sound data and gray-level co-occurrence matrix;

[0119] Gradient structure tensor properties are extracted from preprocessed farsound data.

[0120] Seismic data-based concatenation attribute extraction includes:

[0121] Extracting chaotic attribute volumes from seismic data;

[0122] Extracting variance attribute volume based on chaotic attribute volume;

[0123] Extracting ant bodies based on variance attribute volume.

[0124] Based on the construction of a three-dimensional geological model, and the extraction of fracture sections based on the three-dimensional geological model, the following are included:

[0125] A large-scale discrete distribution model of cracks is constructed based on crack slices in ants, and then a small-scale discrete distribution model of cracks is established.

[0126] The tracking parameters for the large-scale crack discrete distribution model and the small-scale crack discrete distribution model are set based on texture attributes and gradient structure tensor attributes.

[0127] Based on geological knowledge and various attribute slices, fault sections of large-scale and small-scale fracture discrete distribution models are screened and edited to complete the extraction of three-dimensional fracture models.

[0128] The expression for the gray-level co-occurrence matrix is:

[0129] p(i,j)=∑{g(x,y)=ig(x+Δx,y+Δy)=j};

[0130] Where x and y are the horizontal and vertical coordinates of the seismic profile data, respectively, x = 0, 1, ..., L-1, where L is the number of gray levels in the seismic profile, and g(x,y) is the gray value at (x,y). Δx represents the increment of seismic data in the x-direction, and Δy represents the increment of seismic data in the y-direction.

[0131] The gradient structure tensor properties extracted from the preprocessed far-sound data include:

[0132] Based on specific filter operators, signal processing is performed on the preprocessed far-sound data to obtain far-sound seismic data volume;

[0133] Calculation of gradient structure tensor based on farsound seismic data volume;

[0134] The gradient structure tensor is reconstructed to obtain the reconstructed gradient structure tensor;

[0135] By using different Gaussian window parameters, a smoothing test is performed on each component of the reconstructed gradient structure tensor to obtain the optimal reconstructed gradient structure tensor.

[0136] The eigenvalues ​​and eigenvectors of the gradient structure tensor are calculated based on the optimal reconstructed gradient structure tensor.

[0137] Extracting ant bodies based on variance attribute volume includes:

[0138] Ants are extracted from variance attribute data using at least one passive extraction method combined with at least one active extraction method.

[0139] Example 2

[0140] This embodiment provides a method for extracting fractured sections, including:

[0141] The first step is to jointly analyze the far-sonic data with well logging data and seismic data to determine the location of the well-seismic event.

[0142] The second step is to extract the serial attributes based on far-sound data;

[0143] Attribute extraction technology has been widely applied in various fields such as seismic tectonic interpretation, stratigraphic analysis, reservoir characteristic description, and reservoir dynamic monitoring. Its role in oil and gas exploration and development is increasingly significant. According to the definition, attributes can be divided into four main categories: geometric attributes, kinematic attributes, dynamic attributes, and statistical attributes. This embodiment starts with amplitude, frequency, and geometric attributes, analyzing the quality of the identification effect of different attribute analysis results on the characterization of fractures and cavities using acoustic waves, and then performs attribute optimization based on this analysis. Geological anomalies in the acoustic detection profile (…) Figure 2 (a blue box) is the target for multi-attribute research, extracting different attributes for comparison, such as Figure 2-Figure 2 As shown in d. Based on the enhanced reflection intensity, amplitude envelope, relative impedance, and other amplitude-related attributes, there is a further improvement in the identification of geological structures.

[0144] Frequency and geometric attributes were extracted from the far-field acoustic detection data, and the results are as follows: Figure 3 a- Figure 3 As shown in d, the recognition effect of frequency-based and geometric attributes is poor. Frequency-based attributes are limited by factors such as spectrum, detection depth and oil and gas content, while geometric attributes generally require three-dimensional data, which is not met by far-sound detection data.

[0145] Research and development of amplitude-based technologies applicable to far-sonic sounding data were conducted. Based on the data characteristics, image-based texture feature research was carried out to enhance the feasibility of identifying geological anomalies such as cracks through texture features. The BEMD feature enhancement algorithm was used for preprocessing of far-sonic data, suppressing strong reflections from seismic profiles. Effectively highlighting far-sonic waves to enhance seismic information is key to solving this problem. Texture attribute and structural tensor attribute extraction, as image processing techniques, more intuitively reflect the overall characteristics of the data. Extending these techniques to the geophysical field can serve as important evidence for seismic data interpretation. Gao et al. [12-14] The computation of texture properties has been applied in tomographic identification, as demonstrated by Knutsson et al.

[15] The concept of structural tensor was proposed and applied to attribute extraction of different geological bodies, by Bakker, Randen, and others. [16-18] Humans have extended the gradient structure tensor to anisotropic parameters and chaotic properties, as exemplified by Zhang Junhua et al. [19-20] Texture attributes and structural tensor attributes were applied in the corresponding work areas. By calculating texture attributes and structural tensor attributes, the application of seismic attribute texture analysis in far-field acoustic detection was explored.

[0146] Texture attribute extraction: Haralick et al.

[21] This paper proposes a method for calculating texture attributes using gray-level co-occurrence matrix (GLCM). By converting the image to grayscale, the similarity of pixels at a certain distance and direction is statistically analyzed to reflect the image's variation characteristics in different directions and intervals. Using the seismic response feature recognition results from BEMD signal decomposition as input, the image texture feature calculation method is extended to seismic texture attributes. Specifically, the gray-level correlation between pixels is used to realize the amplitude of the seismic data volume. Texture attributes are calculated using GLCM from different angles, gray levels, directions, and windows. By converting the image to grayscale and statistically analyzing the similarity of pixels at a certain distance and direction, the variation characteristics of the image in different directions and intervals are reflected. Applying this to seismic data processing, for a seismic profile described by grayscale, the probability of any two data points with a distance δ along a certain direction satisfying certain conditions is calculated as follows: p(i,j)=∑{g(x,y)=ig(x+Δx,y+Δy)=j}, x=0,1…L-1, where L is the grayscale level of the seismic profile; x and y are the horizontal and vertical coordinates of the seismic profile data, respectively; and g(x,y) is the grayscale value at (x,y) of the seismic profile. The gray-level co-occurrence matrix (GLCM) statistically analyzes profile information. To intuitively obtain the changes in gray levels, i.e., data, within the profile, texture attribute values ​​derived from the GLCM are calculated.

[0147] a. Energy: Energy is a measure of the uniformity of texture grayscale changes, reflecting the degree of uniformity of grayscale distribution and the coarseness of the texture.

[0148] b. Entropy: Entropy measures the randomness of a texture. It reaches its maximum value when all values ​​in the co-occurrence matrix are equal; conversely, it is smaller if the values ​​in the co-occurrence matrix are highly uneven.

[0149] c. Contrast ratio: Contrast ratio is the moment of inertia near the main diagonal of the gray-level co-occurrence matrix. It measures the distribution and local variations of matrix values ​​and reflects sharpness and the depth of texture grooves.

[0150] Gradient structure tensor attribute extraction and analysis is a novel attribute analysis method introduced from image processing into seismic interpretation in recent years. It identifies different structural features or texture units (such as layered textures, chaotic textures, etc.) in seismic images. Texture variations in images actually represent anomalies in geological targets such as faults and fissures, and the degree of variation can be expressed by feature values ​​in the XYZ directions. The Gst (gradient structure tensor) is mainly used to calculate seismic tectonic attributes and predict fracture development zones. Based on existing algorithms, the extraction results of Gst attributes have been developed and optimized, specifically through the following steps: b. Signal processing of the profile using specific filter operators to enhance the boundary of far-sonic detection data; b. Calculation of the gradient structure tensor: Calculating the directional derivative of the gradient vector at each point in the 3D seismic data volume:

[0151]

[0152] in, For the gradient operator, σ g Here, μ(x) is the scale parameter, x is the x-coordinate of the seismic data, y is the y-coordinate of the seismic data, and z is the z-coordinate of the seismic data.

[0153] Texture attribute extraction results are as follows Figure 4 As shown in b.

[0154] Reconstructing the gradient structure tensor T

[22] (Bakker, 2002):

[0155]

[0156] d. Here, different Gaussian window parameters are used to flatten each component of the gradient structure tensor.

[0157] Slip test, select the optimal result:

[0158]

[0159] in, For the average gradient structure tensor result, T ij For gradient structure tensors, G(x,σ) is the convolution operator. T ) is the Gaussian kernel function. denoted as the average directional derivative in different directions.

[0160] e. Calculate the eigenvalues ​​and eigenvectors of the gradient structure tensor:

[0161] |Tυ-λυ|=0;

[0162] Where υ is the eigenvector and λ is the eigenvalue.

[0163] Gradient structure tensor attribute extraction and analysis were performed on enhanced far-sound data. Based on the geological and geophysical characteristics of the far-sound data, the attributes extracted through Gaussian kernel (Gst) reflect the subsurface geological bodies to a certain extent. Gradient structure tensor attribute extraction was performed on the model under different Gaussian kernel parameters. Through comparison, we believe that 0.5 and 2 yielded more ideal results. Figure 5 a and Figure 5 As shown in b.

[0164] Based on seismic data attribute concatenation extraction, a chaotic attribute volume is extracted from the seismic data volume of the target area, and a variance attribute volume is extracted from the chaotic attribute volume. Ant bodies are then extracted from the variance attribute volume using at least one passive extraction method combined with at least one active extraction method. Based on these ant bodies, ant body fracture slices are extracted. The ant body refers to the data volume extracted using the ant algorithm. The passive and active extraction methods represent the passive and active algorithms within the ant algorithm, respectively. The passive algorithm is a conservative method for estimating fractures, only estimating along the direction of stronger signal. The active algorithm allows ants to detect fractures in a more flexible way, and is effective for detecting both large and small fractures. In the process of fracture slice extraction, a simple single extraction of ant bodies can reflect the three-dimensional distribution characteristics of fractures to some extent, but the signal is poor and the continuity is poor, failing to accurately reflect the actual distribution of fracture slices.

[0165] The third step is to constrain the crack properties based on farsonic data and clarify the three-dimensional development location of the cracks.

[0166] Based on the attribute extraction results from farsonic data, a three-dimensional comprehensive analysis of the crack attributes extracted from seismic data is performed to clarify the location of crack development, realize the comprehensive zonal characterization of cracks and cavities by combining multiple information, and determine whether cracks have developed by combining the crack and cavity extraction results, delete unclear cracks, and clarify the crack development results.

[0167] Step 4: Extraction of the fractured sections.

[0168] Effective fracture identification is a prerequisite for establishing an accurate three-dimensional geological model of fractures. Fracture slices are extracted from attribute volumes to establish a large-scale discrete fracture distribution model, followed by a small-scale discrete fracture distribution model for reservoir description and comprehensive characterization. Tracking parameters are set based on the results of farsonic data attributes. Based on geological understanding and various attribute volume slices, fault slices are screened and edited to achieve the extraction of a three-dimensional fracture model.

[0169] The crack extraction method of this embodiment was used to conduct three-dimensional wave impedance inversion identification of ground seismic fractures and cavities in a certain work area. The inversion results confirmed the existence of fractures and cavities on the south side of the well. Ground seismic crack identification was also conducted, and the identification results confirmed the existence of cracks on the north side of the well. Figure 6 As shown.

[0170] Example 3

[0171] This embodiment provides a fractured section extraction device, including:

[0172] The joint analysis module is used to perform joint analysis based on far-sonic data, well logging data, and 3D seismic data to determine the location of well-seismic events.

[0173] The first attribute extraction module is used to extract concatenated attributes based on far-sound data to obtain texture attributes and gradient structure tensor attributes.

[0174] The second attribute extraction module is used to extract concatenated attributes based on three-dimensional seismic data to obtain ant bodies;

[0175] The first extraction module is used to extract ant body cracks based on ant bodies;

[0176] The three-dimensional comprehensive analysis module performs a three-dimensional comprehensive analysis of the ant body based on texture attributes and gradient structure tensor attributes to clarify the location of crack development.

[0177] Modules for building three-dimensional geological models;

[0178] The second extraction module is used to extract fracture sections based on a three-dimensional geological model.

[0179] Example 4

[0180] This disclosure also provides an electronic device, which includes:

[0181] At least one processor; and,

[0182] A memory communicatively connected to the at least one processor; wherein,

[0183] The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform the cracked section extraction method in Embodiment 1.

[0184] An electronic device according to embodiments of the present disclosure includes a memory and a processor. The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0185] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of this disclosure, the processor is used to execute computer-readable instructions stored in the memory.

[0186] Those skilled in the art will understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this disclosure.

[0187] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0188] Example 5

[0189] This disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the cracked section extraction method of Embodiment 1.

[0190] A computer-readable storage medium according to embodiments of the present disclosure stores non-transitory computer-readable instructions. When these non-transitory computer-readable instructions are executed by a processor, all or part of the steps of the methods described in the foregoing embodiments of the present disclosure are performed.

[0191] The aforementioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or portable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).

[0192] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A method of extracting a crack sheet, characterized by, The method comprises the following steps: carrying out joint analysis based on far sound wave data, logging data and three-dimensional seismic data to determine well-seismic position; extracting texture attribute and gradient structure tensor attribute based on the far sound wave data; extracting ant body based on the three-dimensional seismic data; extracting ant body crack piece based on the ant body; carrying out three-dimensional comprehensive analysis on the ant body based on the texture attribute and gradient structure tensor attribute to determine crack development position; constructing three-dimensional geological model and extracting crack piece based on the three-dimensional geological model.

2. The crack sheet extraction method of claim 1, wherein, The step of extracting attribute based on the far sound wave data comprises the following steps: preprocessing the far sound wave data based on two-dimensional empirical mode decomposition; extracting the texture attribute based on the preprocessed far sound wave data and gray level co-occurrence matrix; extracting the gradient structure tensor attribute based on the preprocessed far sound wave data.

3. The crack sheet extraction method of claim 1, wherein, The step of extracting attribute based on the seismic data comprises the following steps: extracting chaotic attribute body based on the seismic data; extracting variance attribute body based on the chaotic attribute body; extracting ant body based on the variance attribute body.

4. The crack sheet extraction method of claim 1, wherein, The step of constructing three-dimensional geological model and extracting crack piece based on the three-dimensional geological model comprises the following steps: constructing large-scale crack discrete distribution model based on ant body crack piece and then constructing small-scale crack discrete distribution model; setting tracking parameters of the large-scale crack discrete distribution model and small-scale crack discrete distribution model based on the texture attribute and gradient structure tensor attribute; filtering and editing fault piece of the large-scale crack discrete distribution model and small-scale crack discrete distribution model based on geological cognition and various attribute body slices to complete extraction of three-dimensional fracture model.

5. The crack sheet extraction method of claim 2, wherein, The expression of the gray level co-occurrence matrix is: p(i,j) = Σ{g(x,y) = i.g(x+Δx,y+Δy) = j}; Wherein, x and y are the horizontal and vertical coordinates of the profile data respectively, x = 0, 1...L-1, L is the gray level of the seismic profile, g(x, y) is the gray value at (x, y) Δx is the seismic data x direction increment, and Δy is the seismic data y direction increment.

6. The crack sheet extraction method of claim 2, wherein, The step of extracting gradient structure tensor attribute based on the preprocessed far sound wave data comprises the following steps: carrying out signal processing on the preprocessed far sound wave data based on specific filter operator to obtain far sound wave seismic data body; calculating gradient structure tensor based on the far sound wave seismic data body; reconstructing the gradient structure tensor to obtain reconstructed gradient structure tensor; carrying out smoothing test on each component of the reconstructed gradient structure tensor by using different Gaussian window parameters to obtain optimal reconstructed gradient structure tensor; calculating eigenvalue and eigenvector of the gradient structure tensor based on the optimal reconstructed gradient structure tensor.

7. The crack sheet extraction method of claim 3, wherein, The step of extracting ant body based on the variance attribute body comprises the following steps: extracting ant body from the variance attribute body by combining at least one passive extraction mode with at least one active extraction mode.

8. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the crack piece extraction method of any one of claims 1-7.

9. A non-transitory computer-readable storage medium, comprising: The non-transitory computer readable storage medium stores computer instructions for causing a computer to execute the fracture patch extraction method of any one of claims 1-7.

10. A fracture sheet extraction device characterized by, Comprise: Joint analysis module, for joint analysis based on far acoustic wave data, logging data and three-dimensional seismic data, determine well seismic position; First attribute extraction module, for attribute extraction in series based on the far acoustic wave data, obtain texture attribute and gradient structure tensor attribute; Second attribute extraction module, for attribute extraction in series based on the three-dimensional seismic data, obtain ant body; First extraction module, for extracting ant body fracture patch based on the ant body; Three-dimensional comprehensive analysis module, based on the texture attribute and gradient structure tensor attribute, three-dimensional comprehensive analysis is carried out on the ant body, and the fracture development position is clear; Construction module, for constructing three-dimensional geological model; Second extraction module, for extracting fracture patch based on the three-dimensional geological model.

Images