Radial domain fourier transform interpolation method for weak reflection information

Abstract

The radial domain Fourier transform interpolation method for weak reflection information provided by the application comprises the following steps: preprocessing collected trace data to obtain a radial domain regularized offset data body and an azimuth equidistant data body; transforming the radial domain regularized offset data body from a time-space domain to a frequency-wave number domain; selecting corresponding maximum spectral energy components in the frequency-wave number domain; completing frequency-space domain prediction data according to the maximum spectral energy components; transforming and iterating the frequency-space domain prediction data to complete data reconstruction interpolation processing and obtain reconstruction processing data; and adding the reconstruction processing data to a three-dimensional grid, reassigning values, and obtaining a complete interpolation data body. Through radial domain Fourier transform interpolation, the data signal-to-noise ratio is further improved, and the uniformity of offset and azimuth is ensured, and finally a complete radial domain Fourier transform interpolation method for weak signals is formed.

Description

Technical Field

[0001] This invention relates to the field of data supervision, and in particular to a radial domain Fourier transform interpolation method for weak reflection information. Background Technology

[0002] With the deepening of oil and gas exploration and development and the continuous improvement of geological requirements, acquisition methods are constantly developing towards richer information, such as high density, high precision, ultra-long offset, and wide azimuth. Due to the differences in excitation and reception methods and acquisition and observation systems at different acquisition stages, as well as factors such as the complexity of the surface, irregularities in local observation systems and data elements have also been caused, posing a great challenge to seismic data fusion processing and migration imaging.

[0003] Specifically, this manifests in significant differences in trace spacing, shot spacing, area, offset, and azimuth between different data blocks. These differences directly lead to substantial variations in coverage frequency, energy, signal-to-noise ratio, frequency, phase, and azimuth between different blocks, resulting in inconsistent illumination of subsurface imaging points, uneven distribution of effective information, and large discrepancies in energy consistency, directly impacting imaging quality. However, migration imaging also demands a high degree of regularity and integrity in seismic data.

[0004] Therefore, in order to obtain better and more accurate imaging results, it is necessary to perform regularization processing on the actual data to obtain relatively regular and complete seismic data from the irregular data. This will improve the quality of the data, reduce the impact of acquisition footprints, spatial frequency distortion, and offset arcs caused by data irregularity, and at the same time, improve the overall signal-to-noise ratio of the data as much as possible while maintaining the amplitude, thus ensuring imaging quality.

[0005] To address the impact of irregular data on migration imaging, scholars both domestically and internationally have proposed numerous regularization interpolation methods. Currently, the main regularization interpolation methods include:

[0006] Five-dimensional interpolation regularization primarily utilizes four spatial dimensions—the x and y coordinates of the shot point and receiver point—plus a time dimension to describe the data, allowing the seismic data to be viewed as a five-dimensional data volume. The first four spatial dimensions can be the x and y coordinates of the CMP point plus the projection of the shot-receiver distance in the x and y directions; or the x and y coordinates of the CMP point plus the absolute shot-receiver distance and shot-receiver azimuth. Five-dimensional interpolation regularization involves all five dimensions in the calculation simultaneously. Depending on the actual seismic data, for data with insufficient sampling density, interpolation is generally performed in the shot-receiver domain. Through interpolation, shot lines and receiver lines can be densified, or the area of ​​the data can be reduced. For wide-azimuth, high-density acquired data, regularization can be performed in the OVT domain, regularizing the shot-receiver midpoint to the center of the CMP area grid. This achieves densification of shot lines and shot points in the shot-receiver domain, increasing spatial sampling density, reconstructing shallow missing data, and increasing the shallow coverage frequency, thereby improving the image quality of the data.

[0007] OVT domain regularized interpolation technology involves sorting data into regularized OVT domain gathers based on azimuth and offset in the gather head, transforming the data to the frequency-wavenumber domain using Fourier transform to obtain the Fourier transform of the original data, and then searching for the maximum energy in the frequency-wavenumber domain by setting a threshold operator through a defined number of iterations. After energy optimization using the threshold operator, the data is transformed to the time-space domain, and regularized interpolation is completed through multiple iterations.

[0008] Existing interpolation methods have two main shortcomings. Firstly, conventional interpolation primarily involves sorting existing data at equal intervals and regularizing the offsets. For missing information, weighting factors are calculated based on the horizontal and vertical distances of the interpolated trace. The amplitude value of the interpolated trace is then calculated as the average amplitude of surrounding reference seismic traces. The drawback is that it doesn't consider the true azimuth and offset information during interpolation; it only assigns values ​​to parameters such as amplitude, frequency, and offset, failing to obtain the true information of the interpolated trace. Secondly, Fourier interpolation mainly involves performing a Fourier transform in three-dimensional spacetime to form unique frequency-wavenumber spectral information in the FK domain. The maximum spectral component is generated through a discrete Fourier transform using spectral estimation, extracted, and added to the estimated spectrum. After weighting the data, an inverse discrete Fourier transform is performed. Finally, the original data is subtracted from the weighted data to complete the interpolation. The disadvantage is that irregularly sampled data is prone to spectral leakage during Fourier transform, resulting in severe spurious frequency information in the interpolated data, failing to obtain high-quality data. Summary of the Invention

[0009] In view of the above problems, the present invention is proposed to provide a radial domain Fourier transform interpolation method for weak reflection information that overcomes or at least partially solves the above problems.

[0010] According to one aspect of the present invention, a radial domain Fourier transform interpolation method for weak reflection information is provided, comprising:

[0011] Preprocess the collected gather data to obtain radial domain regularized offset data volume and azimuth angle equally spaced data volume;

[0012] Transform the radial domain regularized offset data volume from the time-space domain to the frequency-wavenumber domain;

[0013] Select the corresponding maximum spectral energy component in the frequency wavenumber domain;

[0014] The frequency spatial domain prediction data is completed based on the maximum spectral energy component.

[0015] The frequency spatial domain prediction data is transformed and iterated to complete the data reconstruction interpolation process and obtain the reconstructed data.

[0016] After adding the reconstructed data to the three-dimensional mesh, it is reassigned to obtain a complete interpolated data volume.

[0017] Optionally, the preprocessing of the collected gathers to obtain radial domain regularized offset data volumes and azimuth angle equally spaced data volumes specifically includes:

[0018] Obtain Daoist data;

[0019] The gather data is sorted into regularized data volumes of 0-360 degrees at equal intervals along the azimuth and offset;

[0020] The regularized data volume is divided into equal offset distances and azimuth angle gathers and transformed into regularized data volumes within 0-180 degrees along a straight line direction of diameter and / or radius, thus completing the radial domain regularized offset distance and azimuth angle equally spaced data volumes.

[0021] Optionally, transforming the radial domain regularized offset data volume from the time-space domain to the frequency-wavenumber domain specifically includes:

[0022] Select the dimension of the data in the radial domain regularized offset data volume;

[0023] The dimensions include: main survey line, connecting survey line, offset distance, azimuth angle, and time;

[0024] Define the desired mesh volume based on the stated dimensions;

[0025] The radial domain regularized offset data volume is subjected to discrete Fourier transform according to frequency characteristics, transforming the data from the time-space domain to the frequency-wavenumber domain, and discrete Fourier transform is performed on each frequency information.

[0026] Optionally, selecting the corresponding maximum spectral energy component in the frequency wavenumber domain specifically includes:

[0027] Select frequencies free of spurious frequencies in the frequency space domain and calculate prior values.

[0028] The frequency data are weighted using the prior values, and the corresponding maximum spectral energy component is selected.

[0029] Optionally, the step of completing the frequency spatial domain prediction data based on the maximum spectral energy component specifically includes:

[0030] Add the maximum spectral energy component within the frequency data range to the corresponding estimated spectrum;

[0031] After cropping and extrapolating the estimated spectrum for each frequency data, the estimated spectrum is corrected to obtain the corrected estimated spectrum.

[0032] The corrected estimated spectrum is convolved with the original frequency data to obtain frequency spatial domain prediction data.

[0033] Optionally, the step of transforming and iterating the frequency spatial domain prediction data to complete the data reconstruction interpolation process and obtain the reconstructed data specifically includes:

[0034] The frequency spatial domain prediction data is subjected to a discrete Fourier inverse transform to a specified position in the time spatial domain.

[0035] In the time-space domain, the predicted data is subtracted from the original data to complete one round of iterative processing. After multiple rounds of iteration, the maximum spectral energy gradually recovers to the average spectral energy value, completing the data reconstruction interpolation and obtaining the reconstructed data.

[0036] Optionally, the step of adding the reconstructed data to the 3D mesh and then reassigning values ​​to obtain the complete interpolated data volume specifically includes:

[0037] The reconstructed data is added to a three-dimensional mesh;

[0038] The interpolated data, including the main survey line, connecting survey line, offset, azimuth, and time, were re-assigned to obtain the complete interpolated data volume.

[0039] This invention provides a radial domain Fourier transform interpolation method for weak reflection information, comprising: preprocessing collected gather data to obtain radial domain regularized offset data volume and azimuth angle equally spaced data volume; transforming the radial domain regularized offset data volume from the time-space domain to the frequency-wavenumber domain; selecting the corresponding maximum spectral energy component in the frequency-wavenumber domain; completing frequency-space domain prediction data based on the maximum spectral energy component; performing transformation iteration on the frequency-space domain prediction data to complete data reconstruction interpolation processing, obtaining reconstructed data; adding the reconstructed data to a three-dimensional mesh and reassigning values ​​to obtain a complete interpolated data volume. Radial domain Fourier transform interpolation further improves the data signal-to-noise ratio and ensures the uniformity of offset and azimuth angle, ultimately forming a complete radial domain Fourier transform interpolation method for weak signals.

[0040] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 A flowchart of a radial domain Fourier transform interpolation method for weak reflection information provided in an embodiment of the present invention;

[0043] Figure 2 For radial domain interpolation surface element attributes (azimuth angle converted to 180 degrees, interval 15 degrees);

[0044] Figure 3 This is a schematic diagram of the Discrete Fourier Transform.

[0045] Figure 4 This is a schematic diagram of the maximum spectral energy after the discrete Fourier transform.

[0046] Figure 5 This is a schematic diagram of the overlay before interpolation;

[0047] Figure 6 This is a schematic diagram of the superposition of elements after interpolation following a radial domain Fourier transform. Detailed Implementation

[0048] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0049] The terms "comprising" and "having," and any variations thereof, in the specification, embodiments, claims, and drawings of this invention are intended to cover non-exclusive inclusion, such as including a series of steps or units.

[0050] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0051] This invention proposes a radial domain Fourier transform interpolation method for weakly reflected signals. First, dynamic correction data is sorted into a 0-360 degree regularized data volume along the azimuth and offset. Simultaneously, the regularized offset and azimuth gathers are transformed radially to a 0-180 degree regularized data volume, completing the radial domain regularized offset and azimuth equally spaced data volume. Regular sampling of the radial domain regularized azimuth and offset ensures sufficient filling of small-angle gathers, thus defining the target data. Second, after frequency sorting of the radial domain data volume, a Discrete Fourier Transform (DFT) is performed on each frequency to the FK domain. The DFT spectral components of each frequency are calculated, and the component with the highest spectral energy is added to the estimated spectrum. This spectrum is then applied to the original data, weighted, and finally subjected to an Inverse DFT to the TX domain to output the predicted data. Finally, the predicted data after the inverse discrete Fourier transform is subtracted from the original data, and the next iteration is performed simultaneously. After multiple iterations, the effective signal is enhanced, random noise is attenuated, and the signal-to-noise ratio of the interpolated data is significantly improved, while the offset, azimuth, and other attributes are more uniform, and the data band-limiting and sampling properties are excellent. Through radial domain five-dimensional spatial interpolation technology, a complete radial domain Fourier transform interpolation method for weak signals is finally formed.

[0052] This invention is a radial domain Fourier transform interpolation method for weak signals. For example... Figure 1 As shown, the processing flow includes the following steps:

[0053] Step 1: Import the preprocessed gather data. First, divide the data into 0-360 degree regularized data volumes at equal intervals along the azimuth and offset. Simultaneously, transform the regularized offset and azimuth gathers into 0-180 degree regularized data volumes along a straight line of diameter or radius, completing the radial domain regularized offset and azimuth equal-interval data volumes. Figure 2 As shown.

[0054] Step 2: Based on the radial domain regularized data volume, select five dimensions for the data, including: main survey line, connecting survey line, offset, azimuth, and time, to define the desired grid volume. Perform Discrete Fourier Transform (DFT) on the input data according to its frequency characteristics, transforming the data from the time-space domain to the frequency-wavenumber domain. Perform DFT on each frequency information, such as... Figure 3 As shown.

[0055] Step 3: In the frequency space domain, select a lower frequency that does not contain spurious frequencies to calculate the prior value. Then, weight the higher frequency data using the prior value and select the corresponding maximum spectral energy component, such as... Figure 4 As shown.

[0056] Step 4: Add the maximum spectral energy component in the higher frequency data range to the corresponding estimated spectrum. After cutting and extrapolating the estimated spectrum for each frequency data, complete the estimated spectrum correction. Finally, convolve the corrected estimated spectrum with the original data to complete the frequency spatial domain prediction data.

[0057] Step 5: Perform an inverse discrete Fourier transform on the frequency spatial domain prediction data to the specified location in the time spatial domain. In the time spatial domain, subtract the predicted data from the original data to complete one round of iterative processing. After multiple rounds of iteration, the maximum spectral energy gradually recovers to the average spectral energy value, completing the data reconstruction interpolation process. Figure 5 As shown.

[0058] Step 6: After adding the reconstructed data volume to the 3D mesh, reassign values ​​to the main survey line, connecting survey line, offset, azimuth, and time of the interpolated data to obtain the complete interpolated data volume. Interpolation using radial domain Fourier transform improves the uniformity of offset and azimuth, significantly enhances the data signal-to-noise ratio, and demonstrates the advantages of migration imaging. This is beneficial for subsequent seismic interpretation, attribute analysis, and azimuth-specific inversion, such as... Figure 6 As shown.

[0059] Beneficial Effects: This invention has three advantages. First, it utilizes dynamic correction data to perform equal-interval sorting and regularization of the data volume along azimuth and offset. Simultaneously, it transforms the regularized offset and azimuth gathers along the radial direction to a regularized data volume within 0-180 degrees, completing the radial domain regularized data volume and ensuring regular sampling of small-angle and near-offset information. Second, it performs frequency sorting on the radial domain regularized data volume, performs Discrete Fourier Transform, and adds the maximum spectral energy component to the estimated spectrum. After trimming and extrapolating the estimated spectrum for each frequency, it completes the estimated spectrum correction. Finally, it convolves the corrected estimated spectrum with the original data to complete the frequency spatial domain prediction data. In the temporal spatial domain, the predicted data is subtracted from the original data to complete multiple rounds of iterative processing, ensuring high signal-to-noise ratio data while preventing spurious frequency phenomena. A third party adds a three-dimensional mesh to the reconstructed data volume and reassigns values ​​to the interpolated main survey line, connecting survey line, offset, azimuth, and time, ultimately obtaining a complete interpolated data volume. The radial domain Fourier transform interpolation further improves the data signal-to-noise ratio and ensures the uniformity of offset and azimuth angle, ultimately forming a complete radial domain Fourier transform interpolation method for weak signals.

[0060] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method of radial domain Fourier transform interpolation for weakly reflective information, characterized by, The transform interpolation method comprises: The preprocessed collected gather data is obtained, and a radial domain regularized offset data body and an azimuth equidistant data body are obtained; The radial domain regularized offset data body is transformed from a time-space domain to a frequency-wavenumber domain; A corresponding maximum spectral energy component is selected in the frequency-wavenumber domain; Frequency space domain prediction data is completed according to the maximum spectral energy component; The frequency space domain prediction data is transformed and iterated to complete data reconstruction interpolation processing, and reconstruction processing data is obtained; After the reconstruction processing data is added to a three-dimensional grid, reassignment is performed to obtain a complete interpolation data body.

2. The method according to claim 1, wherein, The preprocessed collected gather data is obtained, and a radial domain regularized offset data body and an azimuth equidistant data body are obtained, and specifically comprises: Gather data is obtained; The gather data is sorted into a 0-360 degree regularized data body along the azimuth and offset equidistant; The regularized data body is divided into offset and azimuth gathers along the diameter and / or radius straight line direction to convert to a 0-180 degree internal regularized data body, and a radial domain regularized offset and azimuth equidistant data body is completed.

3. The method of claim 1, wherein, The radial domain regularized offset data body is transformed from a time-space domain to a frequency-wavenumber domain, and specifically comprises: The dimension of the data in the radial domain regularized offset data body is selected; The dimension comprises: a main survey line, a contact survey line, an offset, an azimuth, and a time; A desired grid body is defined according to the dimension; The radial domain regularized offset data body is subjected to discrete Fourier transform according to frequency characteristics, and the data is transformed from a time-space domain to a frequency-wavenumber domain, and discrete Fourier transform is performed on each frequency information.

4. The method of claim 1, wherein, A corresponding maximum spectral energy component is selected in the frequency-wavenumber domain, and specifically comprises: A frequency without aliasing is selected in the frequency space domain, and a prior value is calculated; The frequency data is weighted and calculated using the prior value, and a corresponding maximum spectral energy component is selected.

5. The method of claim 1, wherein, Frequency space domain prediction data is completed according to the maximum spectral energy component, and specifically comprises: The maximum spectral energy component in the frequency data range is added to a corresponding estimated spectrum; After cutting and extrapolation are performed on each frequency data estimated spectrum, estimated spectrum correction is completed, and a corrected estimated spectrum is obtained; The corrected estimated spectrum is convolved with the original frequency data to obtain frequency space domain prediction data.

6. The method of claim 1, wherein, The frequency space domain prediction data is transformed and iterated to complete data reconstruction interpolation processing, and reconstruction processing data is obtained, and specifically comprises: The frequency space domain prediction data is subjected to discrete inverse Fourier transform to a specified position in a time-space domain; In the time-space domain, the prediction data is subtracted from the original data to complete one round of iteration processing, and after multiple rounds of iteration, the maximum spectral energy gradually recovers to the average spectral energy value, data reconstruction interpolation is completed, and reconstruction processing data is obtained.

7. The method of claim 1, wherein, The reconstruction processing data is added to a three-dimensional grid, and reassignment is performed to obtain a complete interpolation data body, and specifically comprises: The reconstruction processing data is added to a three-dimensional grid; The interpolation data main survey line, contact survey line, offset, azimuth, and time are revalued to obtain a complete interpolation data body.

Images