A stratum dip angle calculation method and system based on gradient tensor characteristic decomposition

Abstract

The present application provides a stratum dip angle calculation method and system based on gradient tensor feature decomposition, relating to the technical field of geophysical exploration. The method comprises: obtaining structural contour line grid data and well point dip angle measured data; constructing a structural trend surface by spatial interpolation on the structural contour line grid data; calculating a first-order gradient vector based on the structural trend surface to determine the maximum tilt direction and dip angle value of the stratum under the constraint of a fault-free boundary; setting a buffer zone near the fault interpretation position, constructing a weight function based on the shortest distance from the calculation point to the fault to attenuate the gradient structural tensor in the buffer zone, and calculating the local dip angle value of the stratum after the influence of the fault is weakened; constructing a well control weight function based on the distance between the well point and the calculation position, weighting and fusing the measured dip angle of the well point and the weakened dip angle value, applying strong constraint at the well point, correcting the maximum tilt direction and dip angle size, and generating the stratum dip angle distribution result in the fault development area.

Description

Technical Field

[0001] This invention relates to the field of geophysical exploration technology, and in particular to a method and system for calculating stratigraphic dip angle based on gradient tensor eigenvalue decomposition. Background Technology

[0002] In the fields of geophysical exploration and geological engineering, stratigraphic dip angle is a crucial fundamental parameter characterizing the spatial attitude of strata, directly impacting structural interpretation, reservoir evaluation, and engineering deployment. Existing methods for calculating stratigraphic dip angle mainly include geometric calculation-based methods and methods based on seismic data attributes. Among geometric calculation-based methods, the three-point method calculates dip angle by constructing a local plane using three control points on the same stratum; this method is highly sensitive to the location of the control points and the accuracy of the data. The plane fitting method uses multiple stratigraphic points for least-squares plane fitting; this method can only reflect the average dip angle of structural units and cannot characterize the local variations in stratigraphic dip angle in fault-prone areas. Methods based on structural contour lines determine the stratigraphic dip direction by analyzing the density and direction of contour lines; however, this method is affected by fault displacement and contour line discontinuities in fault-dense areas, leading to unstable dip angle calculations. Methods based on seismic reflection phase axes estimate local stratigraphic dip angles by calculating the spatial gradient or phase change of the seismic data volume; however, in fault-prone areas, discontinuities and breaks in the seismic reflection phase axes result in outliers in the calculated dip angles. The methods described above do not fully consider the control effect of faults on the attitude of strata in fault-developed areas, and lack effective integration of structural contour trend information and well control data, resulting in insufficient reliability and geological rationality of the calculated strata dip angles in fault-developed areas.

[0003] The technical problem to be solved by this invention is to provide a method for calculating the dip angle of strata based on gradient tensor eigenvalue decomposition, so as to solve the technical problems of unstable calculation results caused by fault interference and lack of organic integration mechanism between structural trend constraints and well control constraints in the process of calculating the dip angle of strata in fault-developed areas. Summary of the Invention

[0004] To address the aforementioned problems in existing technologies, this invention proposes a method for calculating formation dip angle based on gradient tensor eigenvalue decomposition, comprising: S1: Obtain structural contour grid data of the target strata and measured dip angle data at well points; S2: Based on the structural contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. S3: Based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal eigenvector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal eigenvector. S4: Set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. S5: Construct a well control weighting function based on the spatial distance between the well point and the calculation location. Then, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened. Apply strong dip angle constraints at the well point to correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution results of the formation in the fault development zone.

[0005] This invention proposes a system for calculating formation dip angle based on gradient tensor eigenvalue decomposition. The system employs the method described above and includes: The data acquisition module is used to perform step S1: acquire structural contour grid data of the target strata and measured dip angle data at well points; The trend surface construction module is used to execute step S2: based on the constructed contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. The gradient tensor calculation and principal feature extraction module is used to execute step S3: based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal feature vector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal feature vector. The fault buffer processing module is used to execute step S4: set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. The well control fusion correction module is used to execute step S5: construct a well control weight function based on the spatial distance between the well point and the calculation location, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened, apply a strong dip angle constraint at the well point, correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution result of the fault development zone.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: The data acquisition step obtains structural contour grid data and measured dip angle data at well points, providing seismic structural trend information input and well control constraint information input for subsequent calculations, enabling structural trend information and well control information to be used together in the same technical solution.

[0007] The construction of the tectonic trend surface involves spatial interpolation based on tectonic contour grid data to generate a continuous tectonic trend surface. This trend surface preserves the overall undulation of the stratigraphy reflected in seismic interpretation, providing a continuous and differentiable spatial geometric representation for gradient calculation.

[0008] The gradient tensor calculation and principal feature extraction steps are based on constructing a trend surface to calculate the first-order gradient vector. A gradient structure tensor is statistically constructed within a local window, and the principal feature vector is extracted through eigenvalue decomposition to determine the maximum dip direction and dip angle of the formation. Local window statistics reduce the impact of data noise and local anomalies on the gradient direction.

[0009] The fault buffer processing steps construct a distance-related weight function based on the shortest distance from the calculation point to the fault. This function performs weight attenuation on the gradient structure tensor within the fault buffer, and then performs eigenvalue decomposition again on the weighted gradient structure tensor to calculate the local dip angle of the strata after the fault influence has been weakened. The fault buffer uses the fault location as a spatial constraint boundary, and the distance-related weight function effectively attenuates the contribution of gradient tensors near the fault, avoiding direct interference from fault displacement on dip angle calculation.

[0010] The well-controlled fusion correction step constructs a well-controlled weighting function based on the spatial distance between the well point and the calculation location. It then weights and fuses the measured dip angle data at the well point with the local dip angle values ​​of the formation after the fault influence has been weakened. A strong dip angle constraint is applied at the well point to correct the maximum dip direction and dip angle magnitude of the formation. The well-controlled weighting function ensures that the fusion result fully matches the measured data near the well point, and gradually transitions to the structural trend calculation results in areas far from the well point.

[0011] The five steps are sequentially linked: the output of the data acquisition step serves as the input for the structural trend surface construction step; the output of the structural trend surface construction step serves as the input for the gradient tensor calculation step; the dip angle value without fault boundary constraints output from the gradient tensor calculation step serves as the comparison benchmark for the fault buffer processing step; and the output of the fault buffer processing step serves as the input for the well control fusion correction step, ultimately generating the formation dip angle distribution results for the fault-developed zone. Through the sequential execution of the above steps, the formation dip angle calculation for the fault-developed zone is simultaneously constrained by both the structural contour trend and the measured dip angle data from well points. Furthermore, the fault buffer effectively isolates the interference of fault displacement on the dip angle calculation, thereby obtaining stable, reliable, and geologically reasonable formation dip angle calculation results for the fault-developed zone. Attached Figure Description

[0012] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0013] Figure 1 The diagram shown is a flowchart illustrating a method for calculating the dip angle of a formation based on gradient tensor eigenvalue decomposition according to an embodiment of the present invention. Figure 2 The diagram shown is a schematic diagram of a formation dip angle calculation system based on gradient tensor eigenvalue decomposition according to an embodiment of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0015] The specific embodiments of the present invention will be described below.

[0016] Example 1 like Figure 1 As shown, this invention proposes a method for calculating the dip angle of strata based on gradient tensor eigenvalue decomposition, comprising: S1: Obtain structural contour grid data of the target strata and measured dip angle data at well points; S2: Based on the structural contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. S3: Based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal eigenvector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal eigenvector. S4: Set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. S5: Construct a well control weighting function based on the spatial distance between the well point and the calculation location. Then, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened. Apply strong dip angle constraints at the well point to correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution results of the formation in the fault development zone.

[0017] This method calculates the dip angle of strata in fault-developed areas through five sequential steps: data acquisition, construction of trend surfaces, gradient tensor calculation and principal feature extraction, fault buffer processing, and well-controlled fusion correction. The data acquisition step provides both structural contour trend information and measured wellpoint dip angle information for subsequent calculations. The structural trend surface construction step transforms discrete contour data into a continuous, differentiable spatial surface, providing a foundation for gradient calculation. The gradient tensor calculation and principal feature extraction step utilizes local window statistics to construct a gradient structure tensor and extracts principal feature vectors through eigenvalue decomposition, making the determination of the maximum dip direction of the strata noise-resistant. The fault buffer processing step attenuates the gradient tensor near the fault based on distance-related weights, effectively isolating the interference of fault displacement on dip angle calculation. The well-controlled fusion correction step applies strong constraints to the calculation results using measured wellpoint data, ensuring that the results at the wellpoints fully conform to actual geological conditions. The five steps are linked together, with each step's output serving as the input for the next step. Ultimately, the calculated stratigraphic dip angle in the fault-developed area is simultaneously constrained by both the structural contour trend and the well point measured data, and fault interference is effectively suppressed, thus obtaining a stable, reliable, and geologically reasonable stratigraphic dip angle distribution.

[0018] In some implementations, spatial interpolation is performed on the target stratigraphic structural surface based on the constructed contour grid data to construct a continuous structural trend surface that describes the overall undulation of the strata. Specifically, this includes: Obtain 3D seismic interpretation data to determine the target stratigraphic horizon; Obtain formation depth data at the well point and use the formation depth data at the well point as a hard constraint condition; A spatial interpolation algorithm is used, with the constructed contour grid data as the main variable and the formation depth data at the well point as the constraint variable, to perform interpolation calculations on the target formation structure surface. Generate a continuous structural trend surface covering the entire study area.

[0019] This method further restricts the construction of the structural trend surface to use formation depth data at well points as a hard constraint, and uses structural contour grid data as the main variable and well point depth data as the constraint variable for spatial interpolation. This scheme ensures that the generated structural trend surface not only retains the overall undulation pattern reflected by the seismic structural contours, but also accurately matches the actual formation depth at the well point location, improving the spatial accuracy of the structural trend surface in the well-controlled area and providing more accurate basic data input for subsequent gradient calculations.

[0020] In some implementations, based on constructing a trend surface, the first-order gradient vector in the horizontal direction is calculated, and the gradient vector is statistically analyzed within a local window to construct a contour gradient structure tensor. Specifically, this includes: Extract the depth value z of the tectonic trend surface at each grid node, and construct the spatial geometric expression function z=f(x,y) of the stratigraphic interface with spatial plane coordinates x and y as independent variables; Calculate the first-order gradient vector of the function in the horizontal direction, ▽z=[ z / x, z / y]^T; Within a predefined local window, tensor products are performed on the gradient vectors of each grid node within the window to construct a gradient structure tensor, the expression of which is: , Where G is the gradient structure tensor. The first-order gradient vector of the depth function. and These are the partial derivatives of the depth along the x and y directions, respectively.

[0021] This method further defines the specific construction method of the gradient structure tensor, namely, extracting the depth value z of each grid node of the constructed trend surface, constructing a spatial geometric expression function z=f(x,y), calculating the first-order gradient vector, and performing tensor product operation on the gradient vectors of each grid node within a preset local window. This scheme constructs the gradient structure tensor through statistical operations within a local window, so that the calculation of the gradient direction no longer depends on the gradient value of a single grid point, but integrates the gradient information of multiple points within the window, thereby enhancing the stability and noise resistance of the gradient direction.

[0022] In some implementations, eigenvalue decomposition is performed on the gradient structure tensor to extract the principal eigenvectors. Based on these principal eigenvectors, the maximum dip direction and corresponding dip angle of the strata under fault-free boundary constraints are determined. Specifically, this includes: Perform eigenvalue decomposition on the gradient structure tensor G and solve the characteristic equation. Where i = 1, 2, ; Extracting the largest eigenvalue Corresponding main feature vector The direction of the principal feature vector is taken as the maximum dip direction of the stratum at the corresponding location. The dip angle of the formation under fault-free boundary constraints is calculated based on the maximum eigenvalue. The formula is as follows: , in, This represents the dip angle of the strata. The largest eigenvalue of the gradient structure tensor; By traversing all grid nodes within the study area, a dip scalar field without fault boundary constraints is generated. , where x and y are spatial plane coordinates.

[0023] This method further defines the specific formula for eigenvalue decomposition of the gradient structure tensor, extracting the principal eigenvector corresponding to the largest eigenvalue as the maximum tilt direction, and calculating the dip angle using the largest eigenvalue. This scheme separates the direction and magnitude information of the gradient tensor through eigenvalue decomposition, ensuring mathematical consistency between determining the maximum tilt direction and calculating the dip angle. It traverses all grid nodes to generate a dip angle scalar field, providing spatially continuous, fault-free boundary-constrained dip angle data for subsequent fault buffer processing and well control fusion.

[0024] In some implementations, a fault buffer is set up near the fault interpretation location. A distance-related weight function is constructed based on the shortest distance from the computation point to the fault. Weight decay processing is then applied to the gradient structure tensor within the fault buffer. Specifically, this includes: Obtain spatial coordinate data of the fault interpretation location to determine the fault's distribution location F on the plane. f ; Set the fault buffer radius r f For the study area For any point (x, y) within the fault zone, calculate the shortest distance d from that point to the fault. ,in This is a function representing the shortest distance from a point to the fault line. The shortest distance d is not greater than r. f The area was designated as a fault buffer zone. ; In the fault buffer zone Within the fault buffer, a distance-related weighting function w(d) is constructed. This weighting function is then used to weight the gradient structure tensor within the fault buffer, resulting in a weighted gradient structure tensor. ,in This is the original gradient structure tensor.

[0025] This method further defines the specific steps for setting up the fault buffer and weight attenuation processing, namely, determining the fault distribution location, setting the buffer radius, delineating the fault buffer, constructing a distance-related weight function, and weighting the gradient structure tensors within the buffer. This scheme uses the fault location as a spatial constraint boundary, causing the gradient tensor weights within the buffer to vary with distance. The contribution of gradient tensors closer to the fault is attenuated, avoiding the participation of spurious gradient information generated by fault faulting in dip angle calculation, thereby obtaining a more realistic local dip angle value near the fault.

[0026] In some implementations, the distance-related weight function w(d) is constructed as follows: Define the radius d0 of the fault core influence zone and the radius r of the outer boundary of the fault buffer zone, where d0 <r; When the shortest distance d from the calculation point to the fault is not greater than d0, the weight value is set to 0; when d is not less than r, the weight value is set to 1; when d is between d0 and r, the weight value transitions from 0 to 1 according to a power function, and the expression of the weight function is: , Where w(d) is the distance-related weight value, d is the shortest distance from the calculation point to the fault, d0 is the radius of the fault core influence zone, r is the radius of the outer boundary of the fault buffer zone, and n is the attenuation exponent and n≥1.

[0027] This method further defines the specific segmentation of the distance-related weighting function, including setting the radius of the fault core influence zone and the outer boundary radius, and assigning weight values ​​of 0, power function transition, and 1 to different distance intervals. This scheme controls the smoothness of the transition segment through a power function, ensuring a continuous and smooth transition of weights from complete suppression of the fault core zone to complete preservation at the outer boundary. This avoids discontinuities in the dip angle calculation results at the buffer zone boundary caused by abrupt weight changes.

[0028] In some implementations, a well control weighting function is constructed based on the spatial distance between the well point and the calculation location. This function then weights and fuses the measured dip angle data at the well point with the local dip angle values ​​of the formation after the fault influence has been weakened. Specifically, this includes: Obtain the location x of each well point k =(x k y k and the corresponding measured tilt angle data The measured tilt angle data is converted into a unit vector in the direction of maximum tilt. Where k is the well point number, (x k y k Let be the plane coordinates of the k-th well. This is the measured dip angle value at the well location; For any calculation position Calculate the spatial distance between this location and each well point. ,in Let Euclidean distance norm be denoted as . Set the strong constraint radius d of the well point w and well control influence radius R w , where d w <R w Construct the well control weight function, the expression of which is: , Among them, w k (d) represents the weight value of the calculated location of the k-th well pair at a distance of d, where d is the spatial distance from the calculation point to the well point. w R is the radius of the strong constraint on the well point (within which the well control weight is 1). w denoted as the well control influence radius (the well control weight outside this radius is 0), and m is the attenuation exponent with m≥1.

[0029] This method further defines the construction steps of the well control weight function. It converts the measured dip angle of the well point into a unit vector of the maximum dip direction, calculates the spatial distance between the well point and the calculation location, sets the strong constraint radius and the influence radius, and constructs a piecewise well control weight function. This scheme ensures that the well control weight is 1 within the strong constraint radius, 0 outside the influence radius, and smoothly transitions according to a power function in the intermediate region. This achieves the local control effect of the measured well point data on the calculation results, providing a reasonable weight allocation basis for subsequent weighted fusion.

[0030] In some implementations, the measured dip angle data at the well point is weighted and fused with the local dip angle value of the formation after the fault influence is weakened. A strong dip angle constraint is applied at the well point to correct the maximum dip direction and dip angle magnitude of the formation. Specifically, this includes: Extract the dip angle value at the calculated location after the fault influence has been weakened. and its maximum tilt direction unit vector ; The tilt angles are weighted and fused using the following formula: , in, To calculate the tilt angle at the location after fusion, The dip angle is the calculated value after the fault effect has been weakened, and N is the total number of well points. Let k be the weight value of the k-th well at the calculation location. Let be the measured dip angle value of the k-th well. To calculate the spatial distance from the location to the k-th well; at well point x k Satisfying Strong constraints; The maximum tilt direction is then weighted and fused using a vector-based fusion formula: , Where u(x) is the unit vector of the maximum tilt direction after fusion, u c (x) is the unit vector of the maximum dip direction after the fault effect is weakened, u k Let x be the unit vector of the maximum dip direction corresponding to the measured dip angle of the k-th well; the meanings of the other symbols are the same as above. k Satisfying Strong constraints; Based on the maximum tilt direction vector after fusion Calculate the corrected azimuth dip of the strata. and with the fused tilt angle value As the final stratigraphic dip output, where (u x u y ) are the x-component and y-component of the fused unit vector, respectively.

[0031] This method further defines the specific formula for weighted fusion, namely, performing vector weighted fusion on dip angle magnitude and maximum dip direction separately, while ensuring strong constraints are met at the well point. This scheme ensures that the fused result is exactly equal to the measured dip angle at the well point location, and smoothly transitions to the structural trend calculation result after the fault influence is weakened in areas far from the well point. At the same time, it processes dip angle magnitude and dip direction separately, ensuring the accuracy of the corrected dip azimuth angle, and finally outputs the formation dip angle distribution result that simultaneously satisfies the absolute constraints of the well point and the relative constraints of the structural trend.

[0032] Example 2 like Figure 2 As shown, this invention proposes a system for calculating stratigraphic dip angle based on gradient tensor eigenvalue decomposition. The system employs the method described above and includes: The data acquisition module is used to perform step S1: acquire structural contour grid data of the target strata and measured dip angle data at well points; The trend surface construction module is used to execute step S2: based on the constructed contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. The gradient tensor calculation and principal feature extraction module is used to execute step S3: based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal feature vector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal feature vector. The fault buffer processing module is used to execute step S4: set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. The well control fusion correction module is used to execute step S5: construct a well control weight function based on the spatial distance between the well point and the calculation location, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened, apply a strong dip angle constraint at the well point, correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution result of the fault development zone.

[0033] This system corresponds to the method proposed in Example 1, and will not be described in detail here.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for calculating the dip angle of strata based on gradient tensor eigenvalue decomposition, characterized in that, include: S1: Obtain structural contour grid data of the target strata and measured dip angle data at well points; S2: Based on the constructed contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. S3: Based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal eigenvector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal eigenvector. S4: Set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. S5: Construct a well control weighting function based on the spatial distance between the well point and the calculation location. Then, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened. Apply strong dip angle constraints at the well point to correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution results of the formation in the fault development zone.

2. The method according to claim 1, characterized in that, Based on the constructed contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface describing the overall undulation of the strata, specifically including: Obtain 3D seismic interpretation data to determine the target stratigraphic horizon; Obtain formation depth data at the well point and use the formation depth data at the well point as a hard constraint condition; A spatial interpolation algorithm is used, with the constructed contour grid data as the main variable and the formation depth data at the well point as the constraint variable, to perform interpolation calculations on the target formation structure surface. Generate a continuous structural trend surface covering the entire study area.

3. The method according to claim 1, characterized in that, Based on the constructed trend surface, the first-order gradient vector in the horizontal direction is calculated. The gradient vector is statistically analyzed within a local window to construct the contour gradient structure tensor, specifically including: Extract the depth value z of the tectonic trend surface at each grid node, and construct the spatial geometric expression function z=f(x,y) of the stratigraphic interface with spatial plane coordinates x and y as independent variables; Calculate the first-order gradient vector of the function in the horizontal direction, ▽z=[ z / x, z / y]^T; Within a predefined local window, tensor products are performed on the gradient vectors of each grid node within the window to construct a gradient structure tensor, the expression of which is: ， Where G is the gradient structure tensor. The first-order gradient vector of the depth function. and These are the partial derivatives of the depth along the x and y directions, respectively.

4. The method according to claim 3, characterized in that, Eigenvalue decomposition is performed on the gradient structure tensor to extract the principal eigenvectors. Based on the principal eigenvectors, the maximum dip direction and corresponding dip angle of the strata under fault-free boundary constraints are determined. Specifically, this includes: Perform eigenvalue decomposition on the gradient structure tensor G and solve the characteristic equation. Where i = 1, 2, ; Extracting the largest eigenvalue Corresponding main feature vector The direction of the principal feature vector is taken as the maximum dip direction of the stratum at the corresponding location. The dip angle of the formation under fault-free boundary constraints is calculated based on the maximum eigenvalue. The formula is as follows: ， in, This represents the dip angle of the strata. The largest eigenvalue of the gradient structure tensor; By traversing all grid nodes within the study area, a dip scalar field without fault boundary constraints is generated. , where x and y are spatial plane coordinates.

5. The method according to claim 1, characterized in that, A fault buffer is set up near the fault interpretation location. A distance-related weight function is constructed based on the shortest distance from the calculation point to the fault. The gradient structure tensor within the fault buffer is subjected to weight decay processing, specifically including: Obtain the spatial coordinate data of the fault interpretation location to determine the fault's distribution location F on the plane. f ; Set the fault buffer radius r f For the study area For any point (x, y) within the fault zone, calculate the shortest distance d from that point to the fault. ,in This is a function representing the shortest distance from a point to the fault line. The shortest distance d is not greater than r. f The area was designated as a fault buffer zone. ; In the fault buffer zone Within the fault buffer, a distance-related weighting function w(d) is constructed. This weighting function is then used to weight the gradient structure tensor within the fault buffer, resulting in a weighted gradient structure tensor. ,in This is the original gradient structure tensor.

6. The method according to claim 5, characterized in that, The distance-related weight function w(d) is constructed as follows: Define the radius d0 of the fault core influence zone and the radius r of the outer boundary of the fault buffer zone, where d0 <r; When the shortest distance d from the calculation point to the fault is not greater than d0, the weight value is set to 0; when d is not less than r, the weight value is set to 1; when d is between d0 and r, the weight value transitions from 0 to 1 according to a power function, and the expression of the weight function is: ， Where w(d) is the distance-related weight value, d is the shortest distance from the calculation point to the fault, d0 is the radius of the fault core influence zone, r is the radius of the outer boundary of the fault buffer zone, and n is the attenuation exponent and n≥1.

7. The method according to claim 1, characterized in that, A well control weighting function is constructed based on the spatial distance between the well point and the calculation location. This function weights and fuses the measured dip angle data at the well point with the local dip angle values ​​of the formation after the fault influence has been weakened. Specifically, this includes: Obtain the location x of each well point k =(x k y k and corresponding measured tilt angle data The measured tilt angle data is converted into a unit vector in the direction of maximum tilt. Where k is the well point number, (x k y k Let be the plane coordinates of the k-th well. This is the measured dip angle value at the well location; For any calculation position Calculate the spatial distance between this location and each well point. ,in Let Euclidean distance norm be denoted as . Set the strong constraint radius d of the well point w and well control influence radius R w , where d w <R w Construct the well control weight function, the expression of which is: ， Among them, w k (d) represents the weight value of the calculated location of the k-th well pair at a distance of d, where d is the spatial distance from the calculation point to the well point. w R is the radius of the strong constraint on the well point (within which the well control weight is 1). w denoted as the well control influence radius (the well control weight outside this radius is 0), and m is the attenuation exponent with m≥1.

8. The method according to claim 7, characterized in that, The measured dip angle data at the well point are weighted and fused with the local dip angle values ​​of the formation after the fault influence has been weakened. A strong dip angle constraint is applied at the well point to correct the maximum dip direction and dip angle magnitude of the formation. Specifically, this includes: Extract the dip angle value at the calculated location after the fault influence is weakened. and its maximum tilt direction unit vector ; The tilt angles are weighted and fused using the following formula: ， in, To calculate the tilt angle at the location after fusion, The dip angle is the calculated value after the fault effect has been weakened, and N is the total number of well points. Let k be the weight value of the k-th well at the calculation location. Let be the measured dip angle value of the k-th well. To calculate the spatial distance from the location to the k-th well; at well point x k Satisfying Strong constraints; The maximum tilt direction is then weighted and fused using a vector-based fusion formula: ， Where u(x) is the unit vector of the maximum tilt direction after fusion, u c (x) is the unit vector of the maximum dip direction after the fault effect is weakened, u k Let x be the unit vector of the maximum dip direction corresponding to the measured dip angle of the k-th well; the meanings of the other symbols are the same as above. k Satisfying Strong constraints; Based on the maximum tilt direction vector after fusion Calculate the corrected azimuth dip of the strata. and with the fused tilt angle value As the final stratigraphic dip output, where (u x u y ) are the x-component and y-component of the fused unit vector, respectively.

9. A system for calculating stratigraphic dip angle based on gradient tensor eigenvalue decomposition, characterized in that, The system employs the method described in any one of claims 1 to 8, the system comprising: The data acquisition module is used to perform step S1: acquire structural contour grid data of the target strata and measured dip angle data at well points; The trend surface construction module is used to execute step S2: based on the constructed contour grid data, spatial interpolation is performed on the target stratigraphic structural surface to construct a continuous structural trend surface that describes the overall undulation of the strata. The gradient tensor calculation and principal feature extraction module is used to execute step S3: based on the constructed trend surface, calculate the first-order gradient vector in the horizontal direction, perform statistics on the gradient vector within a local window, construct the contour gradient structure tensor, perform eigenvalue decomposition on the gradient structure tensor, extract the principal feature vector, and determine the maximum dip direction and corresponding dip angle of the strata under the condition of no fault boundary constraint based on the principal feature vector. The fault buffer processing module is used to execute step S4: set up a fault buffer near the fault interpretation location, construct a distance-related weight function based on the shortest distance from the calculation point to the fault, perform weight decay processing on the gradient structure tensor in the fault buffer, perform eigenvalue decomposition on the weighted gradient structure tensor again, and calculate the local dip angle of the strata after the fault influence is weakened based on the decomposition results. The well control fusion correction module is used to execute step S5: construct a well control weight function based on the spatial distance between the well point and the calculation location, perform weighted fusion of the measured dip angle data at the well point and the local dip angle value of the formation after the fault influence is weakened, apply a strong dip angle constraint at the well point, correct the maximum dip direction and dip angle of the formation, and generate the dip angle distribution result of the fault development zone.

Images