Deep strike-slip fracture fracture zone identification method

Abstract

The invention discloses a deep strike-slip fracture fracture zone identification method, and the method comprises the steps: converting original seismic data into a frequency-wavenumber domain through Fourier transform, carrying out the multi-scale decomposition and downsampling through a radial filter cascade architecture, converting the data into a time-space domain through a directional basis filter, and carrying out the filtering. And acquiring an optimal direction by adopting an optimization strategy, and constructing characterization attributes based on neighborhood attribute differences to extract multi-scale fracture attributes. And introducing a plurality of level set functions to construct an energy functional model, and solving through a dynamic weight adjustment algorithm to realize internal and external band identification. And constructing a model by means of a topological relation based on distance-angle constraint to analyze the three-dimensional space topology. The problems that traditional data processing is not fine and model analysis is insufficient are solved, accurate recognition of the deep strike-slip fracture fracture zone is achieved, and the method has important application value in the fields of oil and gas exploration and development and the like.

Description

Technical Field

[0001] The present invention relates to the field of deep-ultra-deep oil and gas exploration and development, and in particular to a method for identifying deep strike-slip fault zones. Background Art

[0002] The identification of deep strike-slip fault zones is of great significance to oil and gas exploration and development. With growing energy demand and the expansion of infrastructure construction, accurate identification of deep strike-slip fault zones provides a key basis for identifying fractured reservoirs in oil and gas resources, mitigating potential risks. However, current identification technologies struggle to meet the high-precision requirements under complex geological conditions, necessitating urgent technological innovation.

[0003] Existing methods for identifying deep strike-slip fault zones suffer from significant deficiencies in data processing. Traditional methods often rely on a single, fixed filtering strategy to process seismic data, failing to perform a multi-scale, refined decomposition based on the data's frequency characteristics. When faced with weak signals generated by deep geological structures, these methods struggle to effectively separate the different frequency components, resulting in useful information being drowned out by noise. This prevents the full capture of the subtle characteristics of the fault zone, compromising the accuracy of subsequent analysis.

[0004] Existing technologies have significant limitations in model construction and analysis. Most identification models rely solely on a single or limited set of attributes, failing to fully consider the interrelationships between multiple attributes. This makes it difficult to accurately depict the complex morphology of fracture zones. Furthermore, the lack of effective analysis of the three-dimensional spatial topological relationships within fracture zones makes it impossible to establish spatial connections between fracture zones, making it difficult to accurately depict the spatial distribution of fracture zones in practical exploration and development. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides a method for identifying a deep strike-slip fault zone.

[0006] The technical solution adopted by the present invention is a method for identifying a deep strike-slip fault zone, comprising the following steps: Step S1: The original seismic data is mapped to the frequency-wavenumber domain using the Fourier transform mechanism. A cascaded architecture consisting of multiple radial filters with frequency attenuation characteristics is used to perform multi-scale decomposition processing on the seismic data in the frequency-wavenumber domain. In the process of increasing the scale step by step, the data volume is reduced using the alternate row and column downsampling technique to obtain multiple sets of scale data covering different frequency characteristics. Step S2: Convert the data of each scale from the frequency-wavenumber domain to the time-space domain through a directional basis filter. Based on the theoretical system of controllable directional filters, the arbitrary directional filter is characterized by a combination of basis filters and weighting functions, and multi-directional filtering operations are performed on the data of each scale. Step S3: For the data set after direction filtering, use the optimization solution strategy to perform a calculation process on each data point to obtain the direction that can make the filtered amplitude reach the peak as the optimal direction, and simultaneously record the weighting function parameters corresponding to the optimal direction; Step S4: Based on the sum of the absolute values of the amplitudes of the directional filtered data in the directions parallel to and perpendicular to the seismic event, potential correlations between the sum and geological features are explored. A weighted coefficient calculation mechanism based on neighborhood attribute differences is introduced to construct an attribute index system for characterizing strike-slip fault zones, and multi-scale fault attribute extraction is performed. Step S5: Introduce multiple level set function elements, extend the signed pressure function to multi-classification problem scenarios, consider the importance of different attributes in the recognition process, construct an attribute weight matrix, and build an energy functional model suitable for multi-phase level set based on the above elements; Step S6: using an iterative optimization algorithm with a dynamic weight adjustment function to solve the energy functional model, and performing contour segmentation operations on the multi-scale fault attributes based on the solution results to achieve the identification and determination of the outer and inner zones of the strike-slip fault fracture zone; Step S7: For the identified strike-slip fault zones, a model is constructed using a topological relationship based on the distance-angle dual constraint condition, and the topological relationship between the fault zones is analyzed in the three-dimensional space dimension to construct a three-dimensional topological structure. Step S8: Based on the constructed three-dimensional spatial topological relationship, using oil and gas accumulation analysis, the favorable area of oil and gas reservoirs in the strike-slip fault fracture zone is characterized to complete the prediction analysis of the distribution trend of the strike-slip fault fracture zone.

[0007] Furthermore, in step S1, among the multiple radial filters of the cascade architecture, the first The output data of the radial filter is taken as the The input data of the radial filter, the cutoff frequency of the two adjacent radial filters and satisfy relationship, in which is the frequency attenuation coefficient, and its value range is ; During the downsampling process, The sampling interval of scale data and scale number follow The law, among which, is the initial sampling interval, The sampling interval growth coefficient is greater than 1, and the downsampling operation adopts a sampling strategy with variable row and column intervals, and the row and column sampling interval ratio satisfies ,in, For the The row and column sampling ratio coefficient of each scale.

[0008] Furthermore, in step S2, based on the angle-weighted directional basis filter combination model, scale data in any direction Filtered data , its calculation expression is ,in, is the total number of basis filters; For the direction Next The weighting function of the basis filter satisfies ; For the The scale corresponding to Basis filter, weighting function According to direction With the base filter direction Angle The specific formula is , and introduce the angle correction factor , adjust the weighting function, the adjusted weighting function ,in, For the Angle correction coefficient of each basis filter.

[0009] Furthermore, in step S3, the optimization algorithm adopts a gradient search model based on adaptive step size and direction dynamic correction. In the search process, the step size According to the gradient modulus of the current point Perform dynamic adjustment, the adjustment formula is ,in, is the step size adjustment coefficient, In order to prevent the denominator from being a very small positive number of zero, a direction correction factor is introduced , when the search direction changes twice in a row with an angle greater than the set threshold When the search direction is corrected, the corrected search direction for ,in, To correct the search direction, is the estimated value of the optimal direction of the target; and in the search process, according to the rate of change of the gradient direction Step size adjustment coefficient Perform secondary adjustment, the adjustment formula is ,in, is the step size coefficient adjustment factor.

[0010] Furthermore, in step S4, the weighted coefficient based on the neighborhood difference To calculate, first determine scale data points Neighborhood range , by calculating the attribute variance of the data points in the neighborhood , combined with the attribute mean of the neighborhood data points , the weighted coefficient calculation formula is ,in, is the number of scales considered, the characterization properties of the strike-slip fault zone Expressed as ,in, Respectively The sum of the absolute values of the amplitudes of the filtering results of the scales in the directions parallel and perpendicular to the seismic phase axis, and the neighborhood gradient weight is introduced at the same time Correct the representation attributes, and the corrected representation attributes ,in, For the scale data points The attribute gradient of .

[0011] Furthermore, in step S5, the energy functional based on the multiphase level set Expressed as ,in, is the total number of level set functions; is the total number of attributes; For the level set function; for point The seismic attribute value at ; For the level set region The mean of the attributes, the attribute weight matrix Elements According to the attributes In the level set Coefficient of variation within a region Calculation, the calculation formula is ,in, For the level set region The standard deviation of the attributes, the attribute association weight matrix is introduced , the energy functional is corrected, and the corrected energy functional ,in, For attributes With attributes The association weight coefficient.

[0012] Furthermore, in step S6, the iterative optimization algorithm based on dynamic weight adjustment is used in the iterative process. The energy change at each iteration Make adjustments, and the adjustment formula is ,in, is the number of iterations, is the weight adjustment coefficient, and the energy change threshold is introduced ,when When , the iteration process is terminated; and during the iteration process, the amplitude is adjusted according to the weight Weight adjustment coefficient Feedback adjustment is performed, and the adjustment formula is ,in, is the feedback regulation factor.

[0013] Furthermore, in step S7, the topological relationship model based on the distance-angle constraint is constructed to determine the relationship between any two fractured and broken zones. and , its topological relationship coefficient Expressed as ,in, is the minimum distance between two regions, obtained by calculating the Euclidean distance between the boundary points of the two regions; is the maximum distance threshold set; is the average angle difference between the two regions, which is obtained by calculating the angle between the main directions of the two regions and taking the average; is the distance and angle weight coefficient, and the area weight is introduced The topological relationship coefficient is corrected, and the corrected topological relationship coefficient ,in, Respectively for regions and The area, is the regional area weight coefficient.

[0014] Beneficial effects: The present invention proposes a method for identifying deep strike-slip fault zones. In terms of data processing, existing methods are unable to finely decompose seismic data. The present invention uses Fourier transform to map data to the frequency-wavenumber domain, and performs multi-scale decomposition through multiple radial filter cascade architectures. At the same time, it adopts alternate row and column downsampling technology, which can not only effectively separate characteristic signals of different frequencies, but also reduce the amount of data and avoid useful information from being interfered with by noise. Compared with the traditional single filtering strategy, it greatly improves the precision and effectiveness of data processing. At the model construction and analysis level, traditional technologies have not fully considered the fusion of multiple attributes and the spatial topological relationship of fault zones. The present invention introduces multiple level set functions and constructs an attribute weight matrix, fuses multiple attributes to construct an energy functional model, and accurately depicts the complex morphology of fault zones; by constructing a model based on the topological relationship of distance-angle constraints, the three-dimensional spatial topological correlation between fault zones is analyzed. These technical means comprehensively make up for the shortcomings of traditional methods. In the fields of oil and gas resource exploration, fault zones can be identified more accurately, providing a reliable basis for the identification of sweet spots in oil and gas exploration and drilling engineering decisions, and significantly improving the accuracy and practicality of identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the method steps of the present invention; Figure 2 Schematic diagram of the original seismic profile and small, medium and large scale fault attributes of the present invention; Figure 3 A schematic diagram of a multiphase level set according to the present invention; Figure 4 This is a schematic diagram of the inner and outer zones and the inner zone of the strike-slip fault zone of the present invention. DETAILED DESCRIPTION

[0016] It should be noted that, unless there is a conflict, the embodiments in this application and the features described in the embodiments can be combined with each other. The application is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] like Figure 1 As shown, a method for identifying a deep strike-slip fault zone includes the following steps: Step S1: The original seismic data is mapped to the frequency-wavenumber domain using the Fourier transform mechanism. A cascaded architecture consisting of multiple radial filters with frequency attenuation characteristics is used to perform multi-scale decomposition processing on the seismic data in the frequency-wavenumber domain. In the process of increasing the scale step by step, the data volume is reduced using the alternate row and column downsampling technique to obtain multiple sets of scale data covering different frequency characteristics. Specifically, step S1 is the data-based processing stage of the entire identification method, which aims to convert the original seismic data into a form more suitable for subsequent analysis. The original seismic data contains rich but complex information, and it is difficult to obtain effective features through direct analysis. Through the Fourier transform mechanism, the original seismic data is converted from the time-space domain to the frequency-wavenumber domain, so that the frequency components in the data can be clearly displayed. On this basis, a cascade architecture composed of multiple radial filters with frequency attenuation characteristics is used to perform multi-scale decomposition of the data. There is a correlation between the cutoff frequencies of two adjacent radial filters, and the cutoff frequency of the latter filter is the cutoff frequency of the previous one. times ,This design can gradually decompose the data from high frequency to low frequency, ,so as to obtain multiple sets of scale data with different frequency ,features.

[0018] In the multi-scale decomposition process, as the scale increases, the amount of data will continue to increase. In order to balance the amount of data processing and the degree of information retention, the alternate row and column downsampling technology is used to reduce the amount of data. The sampling interval of scale data and scale number Related, follow The law ( Greater than 1), and adopts a sampling strategy with variable row and column intervals, and the row and column sampling interval ratio is determined by This downsampling method can not only effectively reduce the amount of data and lower the computational burden of subsequent processing, but also retain the key information in the data to the greatest extent, providing high-quality data for the subsequent accurate identification of deep strike-slip fault zones.

[0019] Step S2: Convert the data of each scale from the frequency-wavenumber domain to the time-space domain through a directional basis filter. Based on the theoretical system of controllable directional filters, the arbitrary directional filter is characterized by a combination of basis filters and weighting functions, and multi-directional filtering operations are performed on the data of each scale. Specifically, step S2 receives the data processed by S1. The core task is to convert the data of each scale from the frequency-wavenumber domain to the time-space domain and perform multi-directional filtering operations. The data domain conversion is achieved through the directional basis filter. This process is based on the controllable directional filter theory system and uses the combination of basis filter and weighting function to represent any directional filter. scale data in any direction The filter data calculation of is based on the angle-weighted directional basis filter combination model. The total number of basis filters is , in the direction Next Weighting function of a basis filter , satisfying the condition that the sum of all weights is 1, For the The scale corresponding to Basis filter. Weighting function According to the direction With the base filter direction Calculate the angle and introduce the angle correction factor Make adjustments.

[0020] This multi-directional filtering method can extract geological features in different directions, fully tapping into the data's potential information in all directions. Because deep strike-slip fault zones can exist in different directions, this filtering operation can enhance the characteristic signals associated with these zones in the data and suppress irrelevant noise. This makes the filtered data more representative of the actual conditions within these zones, providing more accurate data support for subsequent determination of optimal orientations and construction of characterization attributes.

[0021] Step S3: For the data set after direction filtering, use the optimization solution strategy to perform a calculation process on each data point to obtain the direction that can make the filtered amplitude reach the peak as the optimal direction, and simultaneously record the weighting function parameters corresponding to the optimal direction; Specifically, step S3 is based on the data after direction filtering in step S2, and by applying the optimization solution strategy, determines the direction that can make the filtered amplitude reach the peak value for each data point, that is, the optimal direction, and records the corresponding weighting function parameters. A gradient search model based on adaptive step size and dynamic correction of direction is used to calculate the optimal direction. During the search process, the step size will be dynamically adjusted according to the gradient modulus of the current point. When the gradient modulus is large, the step size becomes smaller for a more precise search; when the gradient modulus is small, the step size becomes larger to speed up the search. At the same time, a direction correction factor is introduced. When the angle of change of the search direction is greater than the set threshold for two consecutive times, the search direction is corrected. In addition, during the search process, the step size adjustment coefficient will be adjusted twice according to the rate of change of the gradient direction.

[0022] This complex and sophisticated search mechanism avoids falling into local optima, quickly and accurately finding the optimal orientation for each data point and ensuring that the optimal orientation truly reflects the directional characteristics of the fault zone. Recording the weighting function parameters corresponding to the optimal orientation is crucial; these parameters can help accurately utilize the data information in that orientation during subsequent attribute construction and analysis, further improving the accuracy and reliability of identifying the directional characteristics of deep strike-slip fault zones.

[0023] Step S4: Based on the sum of the absolute values of the amplitudes of the directional filtered data in the directions parallel to and perpendicular to the seismic event, potential correlations between the sum and geological features are explored. A weighted coefficient calculation mechanism based on neighborhood attribute differences is introduced to construct an attribute index system for characterizing strike-slip fault zones, and multi-scale fault attribute extraction is performed. Specifically, step S4 deeply explores the potential correlation between the directional filtering data and the geological characteristics based on the directional information determined in S3, constructs an attribute index system for characterizing the strike-slip fault zone, and then realizes the extraction of multi-scale fault attributes. Based on the sum of the absolute values of the amplitude of the directional filtering data in the direction parallel to the seismic phase axis and perpendicular to the seismic phase axis, a weighting coefficient calculation mechanism based on neighborhood differences is introduced. When calculating the weighting coefficient, first determine the neighborhood range of the data point, and obtain the weighting coefficient by calculating the attribute variance and mean of the data points in the neighborhood. The characterization attributes of the strike-slip fault zone are obtained by a preset calculation method of the weighting coefficient and the absolute values of the amplitude in the parallel and vertical directions, and the neighborhood gradient weight is introduced to correct the characterization attributes, and finally obtain the multi-scale fault attributes, such as Figure 2 As shown, (a) is the original seismic profile, (b), (c), and (d) are the small-scale, medium-scale, and large-scale fault attributes, respectively.

[0024] This construction method fully considers the local characteristics and differences of the data. Because the geological characteristics of the fracture zone vary at different locations, neighborhood-based calculations and corrections can more accurately characterize the properties of the fracture zone. Applying this operation to multi-scale data can comprehensively reflect the characteristics of the fracture zone at different scales. Small-scale data can capture the subtle structure of the fracture zone, while large-scale data can grasp the overall shape. This provides rich and comprehensive information for subsequent identification and analysis, helping to more accurately define the scope of the fracture zone.

[0025] Step S5: Introduce multiple level set function elements, extend the signed pressure function to multi-classification problem scenarios, consider the importance of different attributes in the recognition process, construct an attribute weight matrix, and build an energy functional model suitable for multi-phase level set based on the above elements; Specifically, step S5 introduces multiple level set function elements and applies the symbol pressure function to multi-classification problem scenarios, such as Figure 3 As shown in the figure, the level set method regards the closed curve of the two-dimensional curve in the image as the zero contour line of the two-dimensional curve in the three-dimensional space, and the function describing the two-dimensional curve is called the level set function. represents the interior of the curve, the level set is outside the curve, In particular, the level set function is introduced ,make For region 1, For region 2, For region 3, three classifications can be achieved. Considering the varying importance of different attributes in the identification process, an attribute weight matrix is constructed. Based on these elements, an energy functional model suitable for multiphase level sets is constructed. The energy functional based on multiphase level sets involves multiple parameters, including the number of level set functions, the number of attributes, the level set function, the seismic attribute values, and the mean of the attributes within each level set region. The elements of the attribute weight matrix are calculated based on the coefficient of variation of the attributes within the level set region. An attribute association weight matrix is also introduced to modify the energy functional.

[0026] Because the characteristics of deep strike-slip fault zones are complex and difficult to accurately describe using a single attribute, this energy functional model fuses multi-attribute information, assigns corresponding weights to different attributes, and considers the correlation between attributes to more accurately describe the characteristics of fault zones. This transforms the problem of identifying fault zones into the problem of minimizing the energy functional. By continuously adjusting the model parameters to minimize the energy functional, the accurate modeling of the fault zone is achieved. This provides a theoretical basis and computational model for subsequent delineation and identification, and is a key step in accurately identifying the inner and outer zones of the fault zone.

[0027] Step S6: using an iterative optimization algorithm with a dynamic weight adjustment function to solve the energy functional model, and performing contour segmentation operations on the multi-scale fault attributes based on the solution results to achieve the identification and determination of the outer and inner zones of the strike-slip fault fracture zone; Specifically, step S6 uses an iterative optimization algorithm with a dynamic weight adjustment function to solve the energy functional model constructed in step S5, and contour divides the multi-scale fault attributes based on the solution results to realize the identification and judgment of the outer and inner zones of the strike-slip fault fracture zone. During the iteration process, the weight will be adjusted according to the energy change of each iteration. When the energy change is large, the weight adjustment amplitude will increase accordingly to converge to the optimal solution faster; when the energy change is small, the weight adjustment amplitude will decrease to avoid over-adjustment. At the same time, an energy change threshold is introduced. When the total energy change is less than the threshold, the algorithm is considered to have converged and the iterative process is terminated. In addition, during the iteration process, the weight adjustment coefficient will be feedback-adjusted according to the weight adjustment amplitude.

[0028] Through this dynamically adjusted iterative optimization algorithm, the minimum value of the energy functional can be solved more efficiently, so that the model can better fit the actual fracture zone characteristics. According to the solution results, the multi-scale fracture attributes are divided into contours, which can clearly define the outer and inner zones of the fracture zone, such as Figure 4 As shown in the figure, (a) is the outer zone and (b) is the inner zone, which accurately identifies the location and range of the deep strike-slip fault zone, providing key identification results for subsequent in-depth research and practical application of the fault zone.

[0029] Step S7: For the identified strike-slip fault zones, a model is constructed using a topological relationship based on the distance-angle dual constraint condition, and the topological relationship between the fault zones is analyzed in the three-dimensional space dimension to construct a three-dimensional topological structure. Specifically, step S7 uses a topological relationship construction model based on the distance-angle dual constraint conditions for the identified strike-slip fault zone, analyzes the topological association between each fault zone area in the three-dimensional space dimension, and completes the construction of the three-dimensional space topological structure. In the topological relationship construction model based on the distance-angle constraint, for any two fault zone areas, the topological relationship coefficient is calculated by the minimum distance between the two areas, the average angle difference, and the set distance and angle weight coefficients. The minimum distance is obtained by calculating the Euclidean distance between the boundary points of the two areas, and the average angle difference is obtained by calculating the angle between the main directions of the two areas and taking the average. At the same time, the regional area weight is introduced to correct the topological relationship coefficient.

[0030] Deep strike-slip fault zones often exist not in isolation underground; complex spatial relationships exist between them. By constructing three-dimensional spatial topological relationships, we can clarify the relative positions, orientations, and interconnectedness of individual fault zones. This helps us understand the overall distribution pattern and interactions of these zones, providing an important basis for analyzing the evolution of geological structures and the migration of oil and gas. It also provides more comprehensive information for assessing the comprehensive impact of fault zones on subsequent projects, such as identifying sweet spots in fractured reservoirs and drilling operations.

[0031] Step S8: Based on the constructed three-dimensional spatial topological relationship and combined with oil and gas accumulation analysis, the favorable areas of oil and gas reservoirs in the strike-slip fault zone are characterized to complete the prediction analysis of the distribution trend of the strike-slip fault zone.

[0032] Specifically, deep strike-slip fault zones often do not exist in isolation underground; rather, complex spatial relationships exist between them. By constructing a three-dimensional spatial topological relationship, we can clarify the relative position, orientation, and degree of connection between the various fault zones.

[0033] Preferably, in step S1, among the multiple radial filters of the cascade architecture, the first The output data of the radial filter is taken as the The input data of the radial filter, the cutoff frequency of the two adjacent radial filters and satisfy relationship, in which is the frequency attenuation coefficient, and its value range is ; During the downsampling process, The sampling interval of scale data and scale number follow The law, among which, is the initial sampling interval, The sampling interval growth coefficient is greater than 1, and the downsampling operation adopts a sampling strategy with variable row and column intervals, and the row and column sampling interval ratio satisfies ,in, For the The row and column sampling ratio coefficient of each scale.

[0034] Preferably, in step S2, based on the angle-weighted directional basis filter combination model, scale data in any direction Filtered data , its calculation expression is ,in, is the total number of basis filters; For the direction Next The weighting function of the basis filter satisfies ; For the The scale corresponding to Basis filter, weighting function According to direction With the base filter direction Angle The specific formula is , and introduce the angle correction factor , adjust the weighting function, the adjusted weighting function ,in, For the Angle correction coefficient of each basis filter.

[0035] Preferably, in step S3, the optimization algorithm adopts a gradient search model based on adaptive step size and direction dynamic correction. During the search process, the step size is According to the gradient modulus of the current point Perform dynamic adjustment, the adjustment formula is ,in, is the step size adjustment coefficient, In order to prevent the denominator from being a very small positive number of zero, a direction correction factor is introduced , when the search direction changes twice in a row with an angle greater than the set threshold When the search direction is corrected, the corrected search direction for ,in, To correct the search direction, is the estimated value of the optimal direction of the target; and in the search process, according to the rate of change of the gradient direction Step size adjustment coefficient Perform secondary adjustment, the adjustment formula is ,in, is the step size coefficient adjustment factor.

[0036] Preferably, in step S4, the weighted coefficient based on neighborhood differences To calculate, first determine scale data points Neighborhood range , by calculating the attribute variance of the data points in the neighborhood , combined with the attribute mean of the neighborhood data points , the weighted coefficient calculation formula is ,in, is the number of scales considered, the characterization properties of the strike-slip fault zone Expressed as ,in, Respectively The sum of the absolute values of the amplitudes of the filtering results of the scales in the directions parallel and perpendicular to the seismic phase axis, and the neighborhood gradient weight is introduced at the same time Correct the representation attributes, and the corrected representation attributes ,in, For the scale data points The attribute gradient of .

[0037] Preferably, in step S5, the energy functional based on the multiphase level set Expressed as ,in, is the total number of level set functions; is the total number of attributes; For the level set function; for point The seismic attribute value at ; For the level set region The mean of the attributes, the attribute weight matrix Elements According to the attributes In the level set Coefficient of variation within a region Calculation, the calculation formula is ,in, For the level set region The standard deviation of the attributes, the attribute association weight matrix is introduced , the energy functional is corrected, and the corrected energy functional ,in, For attributes With attributes The association weight coefficient.

[0038] Preferably, in step S6, the iterative optimization algorithm based on dynamic weight adjustment is used in the iterative process. The energy change at each iteration Make adjustments, and the adjustment formula is ,in, is the number of iterations, is the weight adjustment coefficient, and the energy change threshold is introduced ,when When , the iteration process is terminated; and during the iteration process, the amplitude is adjusted according to the weight Weight adjustment coefficient Feedback adjustment is performed, and the adjustment formula is ,in, is the feedback regulation factor.

[0039] Preferably, in step S7, a model is constructed based on the topological relationship of distance-angle constraints, and any two fractured zones are and , its topological relationship coefficient Expressed as ,in, is the minimum distance between two regions, obtained by calculating the Euclidean distance between the boundary points of the two regions; is the maximum distance threshold set; is the average angle difference between the two regions, which is obtained by calculating the angle between the main directions of the two regions and taking the average; is the distance and angle weight coefficient, and the area weight is introduced The topological relationship coefficient is corrected, and the corrected topological relationship coefficient ,in, Respectively for regions and The area, is the regional area weight coefficient.

[0040] The present invention proposes a method for identifying deep strike-slip fault zones. In terms of data processing, traditional technologies use a single fixed filtering strategy, which makes it difficult to effectively separate the different frequency components of seismic data, resulting in the loss of key information. This method uses Fourier transform to convert the original seismic data into the frequency-wavenumber domain, and uses multiple radial filter cascade architectures for multi-scale decomposition. It can gradually extract data features from high frequency to low frequency, and obtain multiple groups of scale data with different frequency characteristics. In the downsampling process, by setting a sampling interval rule related to the scale number and adopting a strategy with variable row and column intervals, while reducing the amount of data, it retains effective information to the maximum extent, providing a high-quality data foundation for subsequent analysis, and effectively solving the problem of imprecise data processing in traditional methods.

[0041] At the model construction level, traditional identification models rely on a single or small number of attributes and cannot accurately depict the complex morphology of fracture zones. This method introduces multiple level set functions, constructs an attribute weight matrix, and fuses multiple attributes to construct an energy functional model. This method fully considers the varying importance of different attributes in the identification process and also introduces an attribute association weight matrix to modify the energy functional. This method can accurately describe the characteristics of fracture zones, transforming the identification problem into solving the energy functional minimum, laying a theoretical foundation for accurate identification.

[0042] Existing technologies lack effective analytical capabilities for analyzing the three-dimensional topological relationships and distribution trends of fault zones. This method constructs a model based on topological relationships based on distance-angle constraints. This method comprehensively considers factors such as distance, angle, and area between regions, clarifies the relative position, orientation, and degree of correlation between fault zones, and constructs an accurate three-dimensional topological structure. This method enables precise prediction of distribution trends, comprehensively improving the analysis of deep strike-slip fault zones and meeting the needs of practical engineering applications.

[0043] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0044] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying deep strike-slip fault zones, characterized in that: The following steps are involved: Step S1: The original seismic data is mapped to the frequency-wavenumber domain using the Fourier transform mechanism. A cascaded architecture consisting of multiple radial filters with frequency attenuation characteristics is used to perform multi-scale decomposition processing on the seismic data in the frequency-wavenumber domain. In the process of increasing the scale step by step, the data volume is reduced using the alternate row and column downsampling technique to obtain multiple sets of scale data covering different frequency characteristics. Step S2: Convert the data of each scale from the frequency-wavenumber domain to the time-space domain through a directional basis filter. Based on the theoretical system of controllable directional filters, the arbitrary directional filter is characterized by a combination of basis filters and weighting functions, and multi-directional filtering operations are performed on the data of each scale. Step S3: For the data set after direction filtering, use the optimization solution strategy to perform a calculation process on each data point to obtain the direction that can make the filtered amplitude reach the peak as the optimal direction, and simultaneously record the weighting function parameters corresponding to the optimal direction; Step S4: Based on the sum of the absolute values of the amplitudes of the directional filtered data in the directions parallel to and perpendicular to the seismic event, potential correlations between the sum and geological features are explored. A weighted coefficient calculation mechanism based on neighborhood attribute differences is introduced to construct an attribute index system for characterizing strike-slip fault zones, and multi-scale fault attribute extraction is performed. Step S5: Introduce multiple level set function elements, extend the signed pressure function to multi-classification problem scenarios, consider the importance of different attributes in the recognition process, construct an attribute weight matrix, and build an energy functional model suitable for multi-phase level set based on the above elements; Step S6: using an iterative optimization algorithm with a dynamic weight adjustment function to solve the energy functional model, and performing contour segmentation operations on the multi-scale fault attributes based on the solution results to achieve the identification and determination of the outer and inner zones of the strike-slip fault fracture zone; Step S7: For the identified strike-slip fault zones, a model is constructed using a topological relationship based on the distance-angle dual constraint condition, and the topological relationship between the fault zones is analyzed in the three-dimensional space dimension to construct a three-dimensional topological structure. Step S8: Based on the constructed three-dimensional spatial topological relationship, using oil and gas accumulation analysis, the favorable area of oil and gas reservoirs in the strike-slip fault fracture zone is characterized to complete the prediction analysis of the distribution trend of the strike-slip fault fracture zone.

2. A method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: In step S1, among the multiple radial filters of the cascade architecture, The output data of the radial filter is taken as the The input data of the radial filter, the cutoff frequency of the two adjacent radial filters and satisfy relationship, in which is the frequency attenuation coefficient, and its value range is ; During the downsampling process, The sampling interval of scale data and scale number follow The law, among which, is the initial sampling interval, The sampling interval growth coefficient is greater than 1, and the downsampling operation adopts a sampling strategy with variable row and column intervals, and the row and column sampling interval ratio satisfies ,in, For the The row and column sampling ratio coefficient of each scale.

3. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: In step S2, based on the angle-weighted directional basis filter combination model, scale data in any direction Filtered data , its calculation expression is ,in, is the total number of basis filters; For the direction Next The weighting function of the basis filter satisfies ; For the The scale corresponding to Basis filter, weighting function According to direction With the base filter direction Angle The specific formula is , and introduce the angle correction factor , adjust the weighting function, the adjusted weighting function ,in, For the Angle correction coefficient of each basis filter.

4. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: In step S3, the optimization algorithm adopts a gradient search model based on adaptive step size and direction dynamic correction. During the search process, the step size According to the gradient modulus of the current point Perform dynamic adjustment, the adjustment formula is ,in, is the step size adjustment coefficient, In order to prevent the denominator from being a very small positive number of zero, a direction correction factor is introduced , when the search direction changes twice in a row with an angle greater than the set threshold When the search direction is corrected, the corrected search direction for ,in, To correct the search direction, is the estimated value of the optimal direction of the target; and in the search process, according to the rate of change of the gradient direction Step size adjustment coefficient Perform secondary adjustment, the adjustment formula is ,in, is the step size coefficient adjustment factor.

5. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: The step S4 is based on the weighted coefficient of neighborhood difference To calculate, first determine scale data points Neighborhood range , by calculating the attribute variance of the data points in the neighborhood , combined with the attribute mean of the neighborhood data points , the weighted coefficient calculation formula is ,in, is the number of scales considered, the characterization properties of the strike-slip fault zone Expressed as ,in, Respectively The sum of the absolute values of the amplitudes of the filtering results of the scales in the directions parallel and perpendicular to the seismic phase axis, and the neighborhood gradient weight is introduced at the same time Correct the representation attributes, and the corrected representation attributes ,in, For the scale data points The attribute gradient of .

6. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: The step S5 is based on the energy functional of the multiphase level set. Expressed as ,in, is the total number of level set functions; is the total number of attributes; For the level set function; for point The seismic attribute value at ; For the level set region The mean of the attributes, the attribute weight matrix Elements According to the attributes In the level set Coefficient of variation within a region Calculation, the calculation formula is ,in, For the level set region The standard deviation of the attributes, the attribute association weight matrix is introduced , the energy functional is corrected, and the corrected energy functional ,in, For attributes With attributes The association weight coefficient.

7. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: In step S6, the iterative optimization algorithm based on dynamic weight adjustment is used in the iterative process. The energy change at each iteration Make adjustments, and the adjustment formula is ,in, is the number of iterations, is the weight adjustment coefficient, and the energy change threshold is introduced ,when When , the iteration process is terminated; and during the iteration process, the amplitude is adjusted according to the weight Weight adjustment coefficient Feedback adjustment is performed, and the adjustment formula is ,in, is the feedback regulation factor.

8. The method for identifying a deep strike-slip fault zone according to claim 1, characterized in that: In step S7, a model is constructed based on the topological relationship of distance-angle constraints, and any two fractured zones are and , its topological relationship coefficient Expressed as ,in, is the minimum distance between two regions, obtained by calculating the Euclidean distance between the boundary points of the two regions; is the maximum distance threshold set; is the average angle difference between the two regions, which is obtained by calculating the angle between the main directions of the two regions and taking the average; is the distance and angle weight coefficient, and the area weight is introduced The topological relationship coefficient is corrected, and the corrected topological relationship coefficient ,in, Respectively for regions and The area, is the regional area weight coefficient.

Images