Data-driven pseudo three-dimensional imaging method, device, equipment and storage medium

Abstract

The present invention belongs to the technical field of oil exploration seismic data processing and discloses a data-driven pseudo-3D imaging method. The method performs de-migration, data matching and 3D gridding on 2D post-stack migration data, interpolation between sparse data traces, data-driven dip weight factor-constrained data reconstruction, and data-driven step-by-step iterative interpolation to obtain a pseudo-3D data volume. A pseudo-3D velocity field is established through multi-information constraints. The pseudo-3D velocity field is used to perform post-stack migration on the pseudo-3D data volume to obtain a migration result. The migration result is then post-stack modified and subjected to high-resolution processing to achieve data-driven pseudo-3D imaging. The present invention also discloses a data-driven pseudo-3D imaging device, equipment, and storage medium. The present invention is suitable for pseudo-3D imaging of old 2D data, greatly improving the utilization rate of old 2D data and resolving the closure issue between different 2D data, providing interpreters with a more intuitive and convenient data volume.

Description

Technical Field

[0001] The present invention belongs to the technical field of oil exploration seismic data processing, and relates to a data-driven pseudo three-dimensional imaging method, device, equipment and storage medium. Background Art

[0002] With the continuous deepening of oil and gas exploration, the density of two-dimensional seismic exploration data in key exploration areas at home and abroad is very high. However, due to the different acquisition times and parameters of these two-dimensional seismic exploration data, and the different processing methods for these data, many problems will be encountered when comprehensively interpreting these data, such as the difficulty of closing between work areas and the multi-solution problem of structural interpretation. Therefore, how to tap the potential of old two-dimensional data has become a hot topic and key research topic.

[0003] Pseudo-3D imaging is a method of converting old two-dimensional data into a pseudo-3D data volume through data interpolation. However, current pseudo-3D imaging methods lack corresponding supporting technologies, such as anti-migration technology and sparse data interpolation methods. Moreover, current interpolation methods do not consider the impact of formation dip on data reconstruction. These factors will lead to a large number of abnormal traces and spatial false frequencies in the interpolated pseudo-3D data volume, making it difficult for interpreters to use them. Therefore, it is necessary to study a pseudo-3D imaging technology with strong applicability and higher imaging accuracy to fully tap the utilization potential of old two-dimensional data. Summary of the Invention

[0004] The purpose of the present invention is to provide a data-driven pseudo 3D imaging method, which obtains a pseudo 3D data volume by processing and interpolating 2D post-stack migration data, so as to improve the utilization rate of old 2D data and solve the closure problem of different 2D data.

[0005] Another object of the present invention is to provide a data-driven pseudo three-dimensional imaging device, equipment and storage medium.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:

[0007] A data-driven pseudo three-dimensional imaging method comprises the following steps:

[0008] The 2D post-stack migration data is processed to obtain a pseudo 3D data volume. A pseudo 3D velocity field is established through multi-information constraints. The pseudo 3D velocity field is used to perform post-stack migration on the pseudo 3D data volume to obtain the migration results. The migration results are then post-stack modified and processed with high resolution to achieve data-driven pseudo 3D imaging.

[0009] The process of processing the two-dimensional post-stack migration data to obtain a pseudo three-dimensional data volume comprises the following steps:

[0010] S1, 2D post-stack migration data de-migration

[0011] Obtain a 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain a self-excited and self-received stack section;

[0012] S2. Data matching and 3D gridding

[0013] The obtained self-excited and self-received superimposed sections are subjected to high signal-to-noise ratio processing and Fourier transform to obtain a superimposed section after data matching; the superimposed section after data matching is subjected to three-dimensional grid processing to obtain a three-dimensional grid model of the superimposed section;

[0014] S3, sparse data interpolation

[0015] For the stacked section 3D grid model, if the line spacing value of the sparse data area is greater than the set value, the internal and external inter-trace interpolation boundaries of the stacked section 3D grid model are determined according to the line spacing, and the grids within the internal inter-trace interpolation boundary area are interpolated to obtain a pseudo 3D stacked data volume 3D grid model;

[0016] The line distance is the distance between two parallel two-dimensional lines;

[0017] S4. Data-driven tilt weight factor constrains data reconstruction

[0018] A 3D grid model of a pseudo 3D stacked data volume is subjected to denoising to obtain pseudo 3D post-stack seismic data, a discrete Fourier transform is performed on the 3D post-stack seismic data to obtain an initial Fourier spectrum, a dip factor weight of the Fourier spectrum is iteratively calculated, a Fourier spectrum component with maximum energy after weighting according to the dip factor weight is selected, the Fourier spectrum component is added to an estimated spectrum, and then multiplied with a spatial sampling operator, and the product result is subtracted from the Fourier spectrum to update the Fourier spectrum, and the iteration is repeated. Finally, an inverse Fourier transform is performed on the estimated spectrum and output to a desired position, thereby completing the reconstruction of the dip factor reflection and obtaining a 3D grid model of the pseudo 3D reconstructed data volume;

[0019] S5. Data-driven step-by-step iterative interpolation

[0020] The grid in the three-dimensional grid model of the pseudo three-dimensional superimposed data volume obtained in step S3 is used as the first-level grid. The size of the first-level grid is iterated step by step in the three-dimensional grid model of the pseudo three-dimensional reconstructed data volume obtained in step S4, so that the size of the next-level grid is adjusted to half of the previous level until it is reduced to the target grid. The three-dimensional grid model of the pseudo three-dimensional reconstructed data volume after each iteration is interpolated to obtain a pseudo three-dimensional data volume.

[0021] As a limitation, the multi-information in the multi-information constraint includes a 2D velocity file, a velocity field of an adjacent 3D work area, well logging data, and a structural steering velocity smoothing parameter.

[0022] As a further limitation, in step S1, the finite difference method is used to perform post-stack de-migration on the two-dimensional post-stack time migration section.

[0023] As a further limitation, in step S2, when the superimposed cross-section after data matching is subjected to three-dimensional gridding processing, the size of the grid is:

[0024]

[0025] Wherein, Line gap represents the line distance between adjacent two-dimensional lines.

[0026] As a further limitation, in step S5, discrete Fourier transform is performed on the three-dimensional post-stack seismic data to obtain the initial Fourier spectrum calculation formula:

[0027]

[0028] in, is the initial amplitude spectrum, the initial Fourier spectrum, D in is the amplitude spectrum of the input data, F is the positive Fourier transform, T is the matrix mask data, γ is the tilt weight factor, and p is a non-negative exponential correction term used to control the size of γ;

[0029] The calculation formula for the tilt factor weight of the iterative calculation of the Fourier spectrum is:

[0030] γ(k x ,k y ,f)=A(θ x ,θ y )=∫|D in (θ x ,θ y ,r)|dr

[0031]

[0032] Among them, k x k is the data of the main survey line direction after discrete Fourier transform of 3D post-stack seismic data, y is the data of the tie line direction after the three-dimensional post-stack seismic data is subjected to discrete Fourier transform, f is the frequency domain part after the three-dimensional post-stack seismic data is subjected to discrete Fourier transform; θ x is the effective apparent inclination angle of radial Fourier transform along the main survey line, θ y is the effective apparent inclination angle of the radial Fourier transform along the tie line direction, and A is θ x and θ y The sum of the amplitude spectra.

[0033] The present invention also provides a data-driven pseudo three-dimensional imaging device, comprising:

[0034] A pseudo 3D data volume module is used to process 2D post-stack migration data to obtain a pseudo 3D data volume;

[0035] Multi-information constraint pseudo 3D velocity field module, used to establish pseudo 3D velocity field through multi-information constraint;

[0036] Migration module, used to perform post-stack migration on pseudo 3D data volume using pseudo 3D velocity field to obtain migration results;

[0037] Data-driven pseudo 3D imaging is used to perform post-stack modification and high-resolution processing on the migration results to achieve data-driven pseudo 3D imaging.

[0038] As a limitation, the pseudo three-dimensional data volume module further includes:

[0039] The 2D post-stack migration data de-migration module is used to obtain the 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain the self-excited and self-collected stack section;

[0040] The data matching and three-dimensional gridding module is used to perform high signal-to-noise ratio processing and Fourier transform on the obtained superimposed sections of self-excited and self-received data to obtain the superimposed sections after data matching; the superimposed sections after data matching are subjected to three-dimensional gridding processing to obtain a three-dimensional gridding model of the superimposed sections;

[0041] The sparse data interpolation module is used to determine the internal and external interpolation boundaries of the stacked section 3D grid model based on the line spacing if the line spacing value of the sparse data area is greater than the set value. The interpolation is then performed on the grids within the internal interpolation boundary area to obtain a pseudo 3D stacked data volume 3D grid model.

[0042] A data-driven dip weight factor constrained data reconstruction module is used to perform noise reduction processing on a three-dimensional grid model of a pseudo three-dimensional stacked data volume to obtain pseudo three-dimensional post-stack seismic data, perform discrete Fourier transform on the three-dimensional post-stack seismic data to obtain an initial Fourier spectrum, iteratively calculate the dip factor weight of the Fourier spectrum, select the Fourier spectrum component with the maximum energy after weighting according to the dip factor weight, add the Fourier spectrum component to the estimated spectrum, then multiply it with the spatial sampling operator, and subtract the product result from the Fourier spectrum to update the Fourier spectrum. The iteration is repeated, and finally the estimated spectrum is inverse Fourier transformed and output to the desired position, completing the reconstruction of the dip factor reflection and obtaining a three-dimensional grid model of the pseudo three-dimensional reconstructed data volume;

[0043] A data-driven step-by-step iterative interpolation module is used to use the grid in the obtained pseudo three-dimensional superimposed data volume three-dimensional grid model as the first-level grid, and to iterate the size of the first-level grid in the obtained pseudo three-dimensional reconstructed data volume three-dimensional grid model step by step so that the size of the next-level grid is adjusted to half of the previous level, and to interpolate the pseudo three-dimensional reconstructed data volume three-dimensional grid model after each iteration to obtain a pseudo three-dimensional data volume.

[0044] The present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store at least one computer program, and the at least one computer program is loaded by the processor and executes the above-mentioned data-driven pseudo three-dimensional imaging method.

[0045] The present invention also provides a storage medium, which is used to store at least one computer program, and the at least one computer program is used to execute the above-mentioned data-driven pseudo three-dimensional imaging method.

[0046] Due to the adoption of the above technical solution, the present invention has achieved the following technical advancements compared with the prior art:

[0047] (1) The present invention processes and interpolates two-dimensional post-stack migration data to obtain a pseudo three-dimensional data volume, which greatly improves the utilization rate of old two-dimensional data and solves the closure problem of different two-dimensional data, providing interpreters with a more intuitive and convenient data volume;

[0048] (2) The present invention establishes a pseudo-3D velocity field through multi-information constraints, providing a reasonable velocity field for the next step of post-stack migration of the pseudo-3D data volume. After post-stack modification and high-resolution processing of the migration results, data-driven pseudo-3D imaging is achieved, providing interpreters with more efficient and reliable pseudo-3D data, greatly improving the utilization potential of old 2D data.

[0049] (3) The present invention performs post-stack de-migration on the two-dimensional post-stack time migration section, thereby restoring the two-dimensional post-stack time migration section to a self-excited and self-retracted stack section, thereby improving the closure accuracy;

[0050] (4) The present invention performs interpolation between sparse data traces for the three-dimensional grid model of the stacked profile, solving the problem of interpolation anomalies between large spacings of two-dimensional seismic exploration lines; and constrains data reconstruction by driving the dip weight factor with data, thereby obtaining an interpolation result with higher accuracy and fewer anomalies.

[0051] In summary, the present invention is applicable to pseudo three-dimensional imaging of old two-dimensional data, which greatly improves the utilization rate of old two-dimensional data, solves the closure problem of different two-dimensional data, and provides interpreters with more intuitive and convenient data bodies. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 Shown is a flow chart of the method of Example 1 of the present invention;

[0053] FIG2( a ) shows a two-dimensional post-stack time migration section in Example 1 of the present invention;

[0054] FIG2( b ) shows the stacked cross section of the self-excitation and self-collection obtained after post-stack demigration in Example 1 of the present invention;

[0055] FIG3( a ) shows an arbitrary line profile before data matching time difference correction in Example 1 of the present invention;

[0056] FIG3( b ) shows an arbitrary line profile after data matching time difference correction in Example 1 of the present invention;

[0057] FIG4( a ) shows a three-dimensional gridded model of a superimposed cross section obtained in Example 1 of the present invention;

[0058] FIG4( b ) shows a 3D gridded model of a pseudo 3D superimposed data volume after interpolation between sparse data channels in Example 1 of the present invention;

[0059] Figure 5 1. A comparison diagram of data reconstruction with data-driven tilt weight factor constraints in Example 1 of the present invention;

[0060] Figure 6 This is an example diagram of a grid for data-driven step-by-step iterative interpolation in Example 1 of the present invention;

[0061] Figure 7 Schematic diagram showing cross-sectional comparison of a three-dimensional grid model of a pseudo three-dimensional reconstructed data volume before interpolation, conventional pseudo three-dimensional interpolation, and a pseudo three-dimensional data volume by data-driven step-by-step iterative interpolation in Example 1 of the present invention;

[0062] Figure 8 Shown are slice comparison diagrams of a 3D grid model of a pseudo 3D reconstructed data volume before interpolation, conventional pseudo 3D interpolation, and a pseudo 3D data volume by data-driven step-by-step iterative interpolation in Example 1 of the present invention;

[0063] FIG9( a ) shows the pseudo 3D velocity field before smoothing in Example 1 of the present invention;

[0064] FIG9( b ) shows the smoothed pseudo 3D velocity field in Example 1 of the present invention;

[0065] FIG10( a ) is a cross-sectional view of a pseudo 3D data volume before post-stack migration in Example 1 of the present invention;

[0066] FIG10( b ) shows a cross-sectional view of the pseudo 3D data volume after post-stack migration in Example 1 of the present invention;

[0067] Figure 11The figure shows a comparison between the original 3D imaging and the data-driven pseudo 3D imaging of a certain block in Example 1 of the present invention;

[0068] Figure 12 The figure shows a superimposed display of a pseudo 3D imaging time slice driven by data of a certain block in Example 1 of the present invention and an uncertainty analysis diagram;

[0069] Figure 13 FIG2 is a block diagram of an apparatus according to Embodiment 2 of the present invention;

[0070] Figure 14 FIG2 is a block diagram of a pseudo three-dimensional data volume module in Example 2 of the present invention;

[0071] Figure 15 FIG2 is a schematic diagram showing the structure of a computer device in Embodiment 2 of the present invention;

[0072] Figure 16 The figure shows a schematic diagram of the structure of the computer storage medium in embodiment 2 of the present invention. DETAILED DESCRIPTION

[0073] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0074] Example 1 A data-driven pseudo-three-dimensional imaging method

[0075] like Figure 1 As shown, this embodiment is a data-driven pseudo three-dimensional imaging method, that is, a method of simulating three-dimensional seismic data imaging processing based on processed two-dimensional stacked seismic data. Since it is not true three-dimensional seismic data, it is called a pseudo three-dimensional imaging method.

[0076] The processing flow of this embodiment includes the following steps:

[0077] S1, 2D post-stack migration data de-migration

[0078] Obtain a 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain a self-excited and self-received stack section;

[0079] In this step, the finite difference method is used to perform post-stack demigration on the 2D post-stack time migration section. In the data preparation stage of data-driven pseudo-3D imaging, after obtaining the 2D post-stack time migration section of the 2D work area, post-stack demigration of the data is necessary because the closure of the seismic data must be performed on the basis of self-excitation and self-collection.

[0080] Figures 2(a) and 2(b) show the two-dimensional post-stack time migration section obtained in this step and the superimposed section of self-excitation and self-receiving obtained after post-stack demigration. As shown in Figure 2(b), the diffraction wave demigration is successful after the two-dimensional post-stack time migration section is demigrated.

[0081] S2. Data matching and 3D gridding

[0082] The obtained self-excited and self-received superimposed sections are subjected to high signal-to-noise ratio processing and Fourier transform to obtain a superimposed section after data matching; the superimposed section after data matching is subjected to three-dimensional grid processing to obtain a three-dimensional grid model of the superimposed section;

[0083] When the superimposed sections after data matching are processed into three-dimensional grids, the grid size is:

[0084]

[0085] Among them, Line gap represents the distance between adjacent two-dimensional lines;

[0086] In this step, high signal-to-noise ratio processing is performed on the obtained stacked sections of self-excitation and self-collection to remove random noise in the stacked sections, and Fourier transform can significantly reduce the generation of abnormal channels. Since the acquisition time and method of different two-dimensional work areas are different, and the processing methods of different data are also different, high signal-to-noise ratio processing and Fourier transform are performed on the obtained stacked sections of self-excitation and self-collection, and after data matching, the phase, time difference, frequency and amplitude between different two-dimensional work areas can be adjusted to gradually reduce the differences between them. Therefore, data matching is the key to ultimately achieving data-driven pseudo-3D imaging;

[0087] As shown in Figure 3(a) and Figure 3(b), these are the arbitrary line profiles before and after the time difference correction for data matching in this step. It can be seen that after the time difference correction, the time difference between different two-dimensional lines in different working areas is eliminated, reaching a closed state.

[0088] After data matching, the stacked sections after data matching are subjected to 3D gridding. That is, all 2D data in the stacked sections after data matching need to be applied to a 3D grid to prepare for subsequent 3D interpolation.

[0089] S3, sparse data interpolation

[0090] For the stacked section 3D grid model, if the line distance value in the sparse data area is greater than the set value, the internal and external inter-trace interpolation boundaries of the stacked section 3D grid model are determined based on the line distance. Inter-trace interpolation is performed on the grids within the internal inter-trace interpolation boundary area to obtain a pseudo 3D stacked data volume 3D grid model. The line distance is the distance between two parallel 2D lines.

[0091] In this step, the set value is two kilometers. For sparse data areas in the stacked profile 3D grid model, the line spacing can be reduced through interpolation. The biggest problem that interpolation solves is that interpolation of overly sparse data may cause instability. Therefore, when the line spacing value of the sparse data area is greater than two kilometers, interpolation is performed on the grid within the internal interpolation boundary area to reduce the abnormal traces generated by interpolation. In addition, by determining the internal and external interpolation boundaries of the stacked profile 3D grid model based on the line spacing, the range of interpolation can be controlled. Although interpolation can improve the accuracy of interpolation in the main structural direction, it may damage the details of the interpolated layer.

[0092] Figures 4(a) and 4(b) show the 3D grid model of the stacked section and the 3D grid model of the pseudo 3D stacked data volume after interpolation of sparse data traces. It can be seen from the figures that interpolation significantly reduces the number of abnormal traces generated by interpolation and improves the accuracy of interpolation in the main structural direction.

[0093] S4. Data-driven tilt weight factor constrains data reconstruction

[0094] A 3D grid model of a pseudo 3D stacked data volume is subjected to denoising to obtain pseudo 3D post-stack seismic data, a discrete Fourier transform is performed on the 3D post-stack seismic data to obtain an initial Fourier spectrum, a dip factor weight of the Fourier spectrum is iteratively calculated, a Fourier spectrum component with maximum energy after weighting according to the dip factor weight is selected, the Fourier spectrum component is added to an estimated spectrum, and then multiplied with a spatial sampling operator, and the product result is subtracted from the Fourier spectrum to update the Fourier spectrum, and the iteration is repeated. Finally, an inverse Fourier transform is performed on the estimated spectrum and output to a desired position, thereby completing the reconstruction of the dip factor reflection and obtaining a 3D grid model of the pseudo 3D reconstructed data volume;

[0095] This step can significantly reduce the anomalies of pseudo-3D interpolation. The first round of estimated spectrum is the initial Fourier spectrum of the data. The estimated spectrum after iteration is the Fourier spectrum of the seismic data after enhanced dip interpolation. Since the Fourier spectrum before interpolation is sparse, the spatial sampling rate after interpolation must be considered, so it is multiplied by the spatial sampling operator to eliminate aliasing.

[0096] In this step, discrete Fourier transform is performed on the three-dimensional post-stack seismic data to obtain the initial Fourier spectrum calculation formula:

[0097]

[0098] in, is the initial amplitude spectrum, that is, the initial Fourier spectrum, D inis the amplitude spectrum of the input data, F is the positive Fourier transform, T is the matrix mask data, γ is the tilt weight factor, and p is a non-negative exponential correction term used to control the size of γ;

[0099] The calculation formula for the tilt factor weight of the iterative calculation of the Fourier spectrum is:

[0100] γ(k x ,k y ,f)=A(θ x ,θ y )=∫|D in (θ x ,θ y ,r)|dr

[0101]

[0102] Among them, k x k is the data of the main survey line direction after discrete Fourier transform of 3D post-stack seismic data, y is the data of the tie line direction after the three-dimensional post-stack seismic data is subjected to discrete Fourier transform, f is the frequency domain part after the three-dimensional post-stack seismic data is subjected to discrete Fourier transform; θ x is the effective apparent inclination angle of radial Fourier transform along the main survey line, θ y is the effective apparent inclination angle of the radial Fourier transform along the tie line direction, and A is θ x and θ y Sum of amplitude spectra;

[0103] like Figure 5 The figure shows a comparison of data-driven dip weight factor-constrained data reconstruction. It can be seen from the figure that the seismic traces reconstructed by constraining the data with dip weight factor are more reasonable.

[0104] S5. Data-driven step-by-step iterative interpolation

[0105] The grid in the three-dimensional gridded model of the pseudo three-dimensional superimposed data volume obtained in step S3 is used as a first-level grid, and the size of the first-level grid is iterated step by step in the three-dimensional gridded model of the pseudo three-dimensional reconstructed data volume obtained in step S4, so that the size of the next-level grid is adjusted to half of the previous level until it is reduced to the target grid, and the three-dimensional gridded model of the pseudo three-dimensional reconstructed data volume after each iteration is interpolated to obtain a pseudo three-dimensional data volume;

[0106] like Figure 6 The figure shows an example of the data-driven iterative interpolation grid in this step. In the figure, during the iteration process, the size of the next level grid is adjusted to half of the previous level, and the interpolated grid is gradually reduced until it reaches the size of the target grid.

[0107] like Figure 7The figure shows a cross-sectional comparison of the 3D mesh model of the pseudo-3D reconstructed data volume before interpolation, conventional pseudo-3D interpolation, and the pseudo-3D data volume obtained through data-driven, step-by-step iterative interpolation. Conventional pseudo-3D interpolation directly interpolates without performing step-by-step processing. The figure shows that the pseudo-3D data volume obtained through data-driven, step-by-step iterative interpolation has fewer anomalies, a more reasonable structure, and richer details.

[0108] like Figure 8 The figure shows a comparison of the slices of the pseudo-3D reconstructed data volume before interpolation, the 3D grid model, the conventional pseudo-3D interpolation, and the pseudo-3D data volume obtained by data-driven step-by-step iterative interpolation. The figure shows that the pseudo-3D data slices obtained by data-driven step-by-step iterative interpolation have richer details, especially in the sparse data area at the work area boundary, and the structure is more reasonable.

[0109] S6. Establish a pseudo 3D velocity field through multi-information constraints, perform post-stack migration on the pseudo 3D data volume using the pseudo 3D velocity field to obtain a migration result, perform post-stack modification and high-resolution processing on the migration result, and realize data-driven pseudo 3D imaging;

[0110] When establishing a pseudo 3D velocity field through multi-information constraints, the multi-information constraints include 2D velocity files, velocity fields of adjacent 3D work areas, logging data, and structural guidance velocity smoothing parameters. Figures 9(a) and 9(b) show the pseudo 3D velocity fields before and after smoothing using the structural guidance velocity smoothing parameters in this step. As can be seen from the figures, after velocity smoothing, the velocity differences between different 2D lines are reduced, and the velocity changes reasonably along the structural direction, which is more beneficial for later post-stack migration.

[0111] Post-stack migration of the pseudo-3D data volume using the pseudo-3D velocity field can restore the reflection waves to their original positions and converge the diffraction waves, ultimately restoring the seismic data to their true spatial positions for subsequent interpretation and analysis. Figures 10(a) and 10(b) show the cross-sections of the pseudo-3D data volume before and after post-stack migration. As can be seen from the figures, post-stack migration solves the problem of reflection wave homing and converges the diffraction waves, which is more conducive to subsequent interpretation and analysis.

[0112] After obtaining data-driven pseudo-3D imaging through the above steps, this embodiment can perform uncertainty analysis on the obtained data-driven pseudo-3D imaging to evaluate the reliability and accuracy of the seismic data processing results, determine the impact of various errors and uncertainties on the processing results, and provide corresponding improvement plans. In the evaluation of data-driven pseudo-3D imaging, uncertainty analysis technology is added to analyze the interpolation results to determine whether the interpolated results are reasonable, which can improve the accuracy and reliability of the final results.

[0113] Taking the two-dimensional post-stack time migration section obtained in a block in Indonesia as an example, the obtained two-dimensional post-stack time migration section is processed using the method of this embodiment, and the following is obtained: Figure 11 The following figure shows a comparison of the original 3D image of the block and the data-driven pseudo-3D image obtained using this embodiment. The left figure shows the original 3D image of the block, and the right figure shows the data-driven pseudo-3D image obtained using this embodiment. Comparing the data-driven pseudo-3D image obtained using this embodiment with the original 3D image of the block shows that the 2D work area data in the original 3D image of the block is very sparse, while the data-driven pseudo-3D image obtained using this embodiment has clear structure, rich details, and high accuracy, fully meeting the needs of interpreters for structural interpretation.

[0114] like Figure 12 The figure shows the superposition of the time slice and uncertainty analysis diagram of the data-driven pseudo-3D imaging in this block. It can be seen from the figure that the uncertainty analysis value of the data-driven pseudo-3D imaging is higher in areas with higher data density, proving that the pseudo-3D imaging results in this area are reliable, but the credibility of the interpolation results is lower for areas with sparse data.

[0115] Example 2: A data-driven pseudo-three-dimensional imaging device

[0116] like Figure 13 FIG. 1 is a block diagram of a data-driven pseudo 3D imaging device according to Example 1. The device includes:

[0117] A pseudo 3D data volume module is used to process 2D post-stack migration data to obtain a pseudo 3D data volume;

[0118] Multi-information constraint pseudo 3D velocity field module, used to establish pseudo 3D velocity field through multi-information constraint;

[0119] Migration module, used to perform post-stack migration on pseudo 3D data volume using pseudo 3D velocity field to obtain migration results;

[0120] Data-driven pseudo 3D imaging is used to perform post-stack modification and high-resolution processing on the migration results to achieve data-driven pseudo 3D imaging.

[0121] like Figure 14 As shown, the pseudo three-dimensional data volume module also includes:

[0122] The 2D post-stack migration data de-migration module is used to obtain the 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain the self-excited and self-collected stack section;

[0123] The data matching and three-dimensional gridding module is used to perform high signal-to-noise ratio processing and Fourier transform on the obtained superimposed sections of self-excited and self-received data to obtain the superimposed sections after data matching; the superimposed sections after data matching are subjected to three-dimensional gridding processing to obtain a three-dimensional gridding model of the superimposed sections;

[0124] The sparse data interpolation module is used to determine the internal and external interpolation boundaries of the stacked section 3D grid model based on the line spacing if the line spacing value of the sparse data area is greater than the set value. The interpolation is then performed on the grids within the internal interpolation boundary area to obtain a pseudo 3D stacked data volume 3D grid model.

[0125] A data-driven dip weight factor constrained data reconstruction module is used to perform noise reduction processing on a three-dimensional grid model of a pseudo three-dimensional stacked data volume to obtain pseudo three-dimensional post-stack seismic data, perform discrete Fourier transform on the three-dimensional post-stack seismic data to obtain an initial Fourier spectrum, iteratively calculate the dip factor weight of the Fourier spectrum, select the Fourier spectrum component with the maximum energy after weighting according to the dip factor weight, add the Fourier spectrum component to the estimated spectrum, then multiply it with the spatial sampling operator, and subtract the product result from the Fourier spectrum to update the Fourier spectrum. The iteration is repeated, and finally the estimated spectrum is inverse Fourier transformed and output to the desired position, completing the reconstruction of the dip factor reflection and obtaining a three-dimensional grid model of the pseudo three-dimensional reconstructed data volume;

[0126] A data-driven step-by-step iterative interpolation module is used to use the grid in the obtained pseudo three-dimensional superimposed data volume three-dimensional grid model as the first-level grid, and to iterate the size of the first-level grid in the obtained pseudo three-dimensional reconstructed data volume three-dimensional grid model step by step so that the size of the next-level grid is adjusted to half of the previous level, and to interpolate the pseudo three-dimensional reconstructed data volume three-dimensional grid model after each iteration to obtain a pseudo three-dimensional data volume.

[0127] The data-driven pseudo 3D imaging device provided in this embodiment is provided with the division of the above functional modules as an example for explanation when performing data processing. In actual applications, the above functions can be assigned to different functional modules as needed.

[0128] Based on the same inventive concept, Figure 15 As shown, this embodiment further provides a computer device, including: at least a processor and a memory, the memory is used to store at least one computer program, and the at least one computer program is loaded by the processor and executes the data-driven pseudo three-dimensional imaging method of embodiment 1.

[0129] Based on the same inventive concept, Figure 16As shown, this embodiment further provides a computer-readable storage medium, which is used to store at least one computer program, and the at least one computer program is used to execute the data-driven pseudo three-dimensional imaging method of embodiment 1.

[0130] Finally, it should be noted that a person skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods.

[0131] Furthermore, it should be appreciated that the computer-readable storage media (eg, memory) herein can be either volatile memory or nonvolatile memory, or can include both volatile and nonvolatile memory.

[0132] It will also be appreciated by those skilled in the art that the various exemplary logic blocks, modules, circuits and algorithmic steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software or a combination of the two. In order to clearly illustrate this interchangeability of hardware and software, a general description has been given of the functions of various schematic components, blocks, modules, circuits and steps. Whether this function is implemented as software or hardware depends on specific applications and the design constraints imposed on the entire system. Those skilled in the art can implement the function in various ways for each specific application, but this implementation decision should not be interpreted as causing a departure from the disclosed scope of the embodiments of the present invention.

[0133] It should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art may still modify the technical solutions described in the above embodiments or replace some of the technical features therein with equivalents. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A data-driven pseudo-3D imaging method, characterized in that: The following steps are involved: The 2D post-stack migration data is processed to obtain a pseudo 3D data volume. A pseudo 3D velocity field is established through multi-information constraints. The pseudo 3D velocity field is used to perform post-stack migration on the pseudo 3D data volume to obtain the migration results. The migration results are then post-stack modified and processed with high resolution to achieve data-driven pseudo 3D imaging. The process of processing the two-dimensional post-stack migration data to obtain a pseudo three-dimensional data volume comprises the following steps: S1, 2D post-stack migration data de-migration Obtain a 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain a self-excited and self-received stack section; S2. Data matching and 3D gridding The obtained self-excited and self-received superimposed sections are subjected to high signal-to-noise ratio processing and Fourier transform to obtain a superimposed section after data matching; the superimposed section after data matching is subjected to three-dimensional grid processing to obtain a three-dimensional grid model of the superimposed section; S3, sparse data interpolation For the stacked section 3D grid model, if the line spacing value of the sparse data area is greater than the set value, the internal and external inter-trace interpolation boundaries of the stacked section 3D grid model are determined according to the line spacing, and the grids within the internal inter-trace interpolation boundary area are interpolated to obtain a pseudo 3D stacked data volume 3D grid model; The line distance is the distance between two parallel two-dimensional lines; S4. Data-driven tilt factor weight constraint data reconstruction A 3D grid model of a pseudo 3D stacked data volume is subjected to denoising to obtain pseudo 3D post-stack seismic data, a discrete Fourier transform is performed on the 3D post-stack seismic data to obtain an initial Fourier spectrum, a dip factor weight of the Fourier spectrum is iteratively calculated, a Fourier spectrum component with maximum energy after weighting according to the dip factor weight is selected, the Fourier spectrum component is added to an estimated spectrum, and then multiplied with a spatial sampling operator, and the product result is subtracted from the Fourier spectrum to update the Fourier spectrum, and the iteration is repeated. Finally, an inverse Fourier transform is performed on the estimated spectrum and output to a desired position, thereby completing the reconstruction of the dip factor reflection and obtaining a 3D grid model of the pseudo 3D reconstructed data volume; S5. Data-driven step-by-step iterative interpolation The grid in the three-dimensional grid model of the pseudo three-dimensional superimposed data volume obtained in step S3 is used as the first-level grid. The size of the first-level grid is iterated step by step in the three-dimensional grid model of the pseudo three-dimensional reconstructed data volume obtained in step S4, so that the size of the next-level grid is adjusted to half of the previous level until it is reduced to the target grid. The three-dimensional grid model of the pseudo three-dimensional reconstructed data volume after each iteration is interpolated to obtain a pseudo three-dimensional data volume.

2. The data-driven pseudo 3D imaging method according to claim 1, characterized in that: The multi-information in the multi-information constraint includes two-dimensional velocity files, velocity fields of adjacent three-dimensional work areas, well logging data and structural steering velocity smoothing parameters.

3. The data-driven pseudo 3D imaging method according to claim 1, wherein: In step S1, a finite difference method is used to perform post-stack de-migration on the two-dimensional post-stack time migration section.

4. The data-driven pseudo 3D imaging method according to claim 1, wherein: In step S2, when the superimposed cross section after data matching is subjected to three-dimensional gridding, the size of the grid is: in, Indicates the distance between adjacent 2D lines.

5. The data-driven pseudo 3D imaging method according to claim 1, wherein: In step S5, discrete Fourier transform is performed on the three-dimensional post-stack seismic data to obtain the initial Fourier spectrum calculation formula: in, is the initial amplitude spectrum, that is, the initial Fourier spectrum, is the amplitude spectrum of the input data, is the positive Fourier transform, is the matrix mask data, is the tilt factor weight, is a non-negative exponential correction term used to control size; The calculation formula for the tilt factor weight of the iterative calculation of the Fourier spectrum is: in, The data of the main survey line direction after the discrete Fourier transform of the 3D post-stack seismic data is performed. The data of tie line direction after discrete Fourier transform of 3D post-stack seismic data, The frequency domain portion of the 3D post-stack seismic data after discrete Fourier transform; is the effective apparent dip angle of radial Fourier transform along the main survey line, is the effective apparent dip angle of radial Fourier transform along the tie line direction, for and The sum of the amplitude spectra.

6. A data-driven pseudo-3D imaging device, characterized in that: include: A pseudo 3D data volume module is used to process 2D post-stack migration data to obtain a pseudo 3D data volume; Multi-information constraint pseudo 3D velocity field module, used to establish pseudo 3D velocity field through multi-information constraint; Migration module, used to perform post-stack migration on pseudo 3D data volume using pseudo 3D velocity field to obtain migration results; Data-driven pseudo 3D imaging, used for post-stack modification and high-resolution processing of migration results, to achieve data-driven pseudo 3D imaging; The pseudo three-dimensional data volume module further includes: The 2D post-stack migration data de-migration module is used to obtain the 2D post-stack time migration section of the 2D work area, perform post-stack de-migration on the 2D post-stack time migration section, and obtain the self-excited and self-collected stack section; The data matching and three-dimensional gridding module is used to perform high signal-to-noise ratio processing and Fourier transform on the obtained superimposed sections of self-excited and self-received data to obtain the superimposed sections after data matching; the superimposed sections after data matching are subjected to three-dimensional gridding processing to obtain a three-dimensional gridding model of the superimposed sections; The sparse data interpolation module is used to determine the internal and external interpolation boundaries of the stacked section 3D grid model based on the line spacing if the line spacing value of the sparse data area is greater than the set value. The interpolation is then performed on the grids within the internal interpolation boundary area to obtain a pseudo 3D stacked data volume 3D grid model. A data-driven dip factor weight-constrained data reconstruction module is used to perform noise reduction processing on a three-dimensional grid model of a pseudo three-dimensional stacked data volume to obtain pseudo three-dimensional post-stack seismic data, perform discrete Fourier transform on the three-dimensional post-stack seismic data to obtain an initial Fourier spectrum, iteratively calculate the dip factor weight of the Fourier spectrum, select the Fourier spectrum component with the maximum energy after weighting according to the dip factor weight, add the Fourier spectrum component to the estimated spectrum, then multiply it with the spatial sampling operator, and subtract the product result from the Fourier spectrum to update the Fourier spectrum. The iteration is repeated, and finally the estimated spectrum is inverse Fourier transformed and output to the desired position, thereby completing the reconstruction of the dip factor reflection and obtaining a three-dimensional grid model of the pseudo three-dimensional reconstructed data volume; A data-driven step-by-step iterative interpolation module is used to use the grid in the obtained pseudo three-dimensional superimposed data volume three-dimensional grid model as the first-level grid, and to iterate the size of the first-level grid in the obtained pseudo three-dimensional reconstructed data volume three-dimensional grid model step by step so that the size of the next-level grid is adjusted to half of the previous level, and to interpolate the pseudo three-dimensional reconstructed data volume three-dimensional grid model after each iteration to obtain a pseudo three-dimensional data volume.

7. A computer device, characterized in that: The computer device includes a processor and a memory, the memory is used to store at least one computer program, and the at least one computer program is loaded by the processor and executes the data-driven pseudo three-dimensional imaging method according to any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium is used to store at least one computer program, and the at least one computer program is used to execute the data-driven pseudo three-dimensional imaging method according to any one of claims 1 to 5.

Images