Random sampling and optimized reconstruction method of seismic data
Through random sampling of seismic data and optimized reconstruction methods, the problem of high density seismic acquisition cost in the old oil area in the eastern region is solved, and high-quality data acquisition and exploration under complex surface conditions is achieved, reducing exploration costs.
Patent Information
- Application Number
- CN202110188392.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-02-18
- Publication Date
- 2024-12-27
- Estimated Expiration
- 2041-02-18
AI Technical Summary
In the eastern old oil areas and dense urban areas, there are many surface obstacles, resulting in high cost of high-density earthquake collection, and regular earthquake collection is difficult to ensure, limiting the exploration and development of favorable oil and gas areas.
The random sampling and optimization reconstruction method of seismic data is used to design a regular observation system for the source and detector, and the sparse sampling coefficient is determined based on the expected exploration cost and data quality, random sampling is performed, and the data is optimized and reconstruction is achieved through sparse inverse transformation.
It effectively reduces the cost of seismic exploration, improves the quality of seismic data, maintains the randomness of the source point and the detection point, avoids large sections of blank collection, and simplifies the difficulty of data reconstruction.
Smart Images
Figure CN114966824B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of oil and gas exploration seismic data acquisition and processing, and in particular to a seismic data random sampling and optimization reconstruction method. Background Art
[0002] At present, the commonly used data acquisition methods for seismic exploration are conventional two-dimensional and three-dimensional regular seismic, which require that the spatial sampling intervals of the shot points and the detector points and the time sampling intervals of the seismic records must be uniform, and there are also strict requirements on the number of sampling points, that is, the source and the detector are arranged regularly at equal intervals to achieve alias-free sampling of seismic data. In recent years, the gradually developed high-density seismic has achieved good application results. However, in the eastern old oil fields with a high degree of exploration, towns are densely populated, there are many surface obstacles, and the relationship between workers and farmers is complex. The construction cost of high-density seismic is getting higher and higher. Even in many areas, due to the limitations of surface conditions, regular seismic acquisition is becoming more and more difficult to guarantee, which limits the exploration and development of favorable oil and gas areas. Based on the above considerations, breaking the conventional regular seismic acquisition and processing framework and inventing irregular seismic data acquisition design and processing methods have important practical significance and practical value.
[0003] In the Chinese patent application with application number: CN201910572070.7, a seismic data reconstruction method is involved, which involves randomly sampling the original signal and performing a sparse transformation; obtaining the coefficient matrix; data reconstruction; error analysis, and judging whether the parameters meet the termination conditions. If they do, it is the reconstruction result, otherwise the data reconstruction step is restarted through iteration.
[0004] In the Chinese patent application with application number: CN201610679857.X, a method and device for reconstructing seismic data are involved, wherein the method includes: obtaining a time-space domain seismic data volume to be reconstructed; performing fast Fourier transform on the time-space domain seismic data volume along the time direction to obtain a time-frequency space domain data volume; performing non-uniform Fourier transform on the time-frequency space domain data volume along the spatial direction according to the actual spatial position to obtain a time-frequency space wavenumber domain data volume; calculating a Gram matrix corresponding to the time-frequency space wavenumber domain data volume; calculating a constraint matrix based on the amplitude spectrum of the time-frequency space wavenumber domain data volume; calculating the wavenumber spectrum components of each time-frequency slice based on the constraint matrix, and storing them in an output matrix; inversely transforming the output matrix to the time-space domain to obtain reconstructed seismic data.
[0005] In the Chinese patent application with application number: CN201510885713.5, a method and device for reconstructing seismic data is involved, which belongs to the field of seismic exploration. The method includes: collecting seismic data; obtaining a preset seismic data threshold; reconstructing the seismic data in the wave atom domain by a convex set projection algorithm according to the preset seismic data threshold, and obtaining the reconstructed seismic data.
[0006] The above existing technologies are significantly different from the present invention and fail to solve the technical problem we want to solve. Therefore, we have invented a new random sampling and optimization reconstruction method for seismic data. Summary of the invention
[0007] The purpose of the present invention is to provide a seismic data acquisition design and optimization reconstruction processing method under the framework of random sparse sampling, which can effectively reduce the cost of seismic exploration and improve the quality of seismic data.
[0008] The purpose of the present invention can be achieved by the following technical measures: a random sampling and optimization reconstruction method for seismic data, which comprises:
[0009] Step 1: Design a seismic observation system with regular sources and receivers;
[0010] Step 2: Determine the sparse sampling coefficient based on the expected exploration cost and seismic data quality;
[0011] Step 3: Select the random sampling step length, determine the random sampling segment, and perform random sampling of detection points;
[0012] Step 4: Select the random sampling step length, determine the random sampling segment, and perform random sampling of the source point;
[0013] Step 5: Output the random sampling design of seismic data;
[0014] Step 6: Use the regular observation system as the sampling matrix to perform sparse inverse transformation on the randomly sampled seismic data to complete the optimized reconstruction.
[0015] The purpose of the present invention can also be achieved by the following technical measures:
[0016] In step 1, a fully complete and regular observation system R of seismic sources and detectors is designed according to the actual surface and underground geological conditions of the exploration area.
[0017] In step 2, the determined sparse sampling is denoted by P.
[0018] In step 3, for a regular detector matrix with N detectors, a reasonable random sampling step size L is selected, and the number of detectors in each step size is int represents the rounding operation. These detectors are used as a random sampling segment. The number of sampling points in the segment is determined according to the sparse sampling coefficient P.
[0019] In step 3, the detectors in each random sampling segment are assigned A random numbers, and the random numbers are sorted from large to small, and the detectors corresponding to the first F random numbers are selected to complete the random sampling of the detectors.
[0020] In step 4, for a regular source matrix with M source points, a reasonable random sampling step size K is selected, and the number of source points in each step size is int represents the rounding operation. These source points are regarded as a random sampling segment. The number of sampling points in the segment is determined according to the sparse sampling coefficient P as G=B*P.
[0021] In step 4, B random numbers are assigned to the focal points in each random sampling segment, and the random numbers are sorted from large to small, and the focal points corresponding to the first G random numbers are screened out to complete the random sampling of the focal points.
[0022] In step 5, the sampling point outputs of the source and the detector are used as a random sampling design for seismic data.
[0023] In step 6, after acquiring seismic data according to the random sampling design scheme, the seismic data is first transformed from the time-space domain to the frequency domain, and data reconstruction is achieved by removing noise and restoring data interpolation. After iterative inversion, it is finally transformed back to the time-space domain according to the sampling matrix of the regular observation system R to complete the optimized reconstruction of the seismic data.
[0024] The random sampling and optimized reconstruction method of seismic data in the present invention is a supporting method integrating random sampling and optimized reconstruction processing of seismic data. The random sampling method of seismic data is used to realize data collection in surface obstacle areas, thereby reducing exploration costs. The optimized reconstruction processing method of seismic data is used to realize accurate data recovery and improve data quality. Compared with the prior art, the present invention has the following advantages:
[0025] First, the randomness of the layout of the source points and the detection points is fully maintained during the sparse sampling process;
[0026] Second, the random sampling segment is flexibly selected according to the random sampling step length, which ensures that the data will not have large blank acquisitions while being random, thus reducing the difficulty of data reconstruction;
[0027] Third, the combination of random sampling of seismic data and optimized reconstruction processing can effectively reduce exploration costs while ensuring data quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1A flowchart of a specific embodiment of the method for random sampling and optimized reconstruction of seismic data of the present invention;
[0029] Figure 2 It is a schematic diagram of a three-dimensional observation system with regular sources and detection points in a specific embodiment 1 of the present invention;
[0030] Figure 3 (a) and (b) are schematic diagrams of random sampling in which the sampling step length of the detection point is 4 and the sampling step length of the shot point is 3 in a specific embodiment 1 of the present invention;
[0031] Figure 4 It is a schematic diagram of a three-dimensional observation system for random sampling of seismic sources and detection points in a specific embodiment 1 of the present invention;
[0032] Figure 5 It is a schematic diagram of an original single shot of randomly sampled data in a specific embodiment 1 of the present invention;
[0033] Figure 6 It is a schematic diagram of a single shot after optimized reconstruction of random sampling data in a specific embodiment 1 of the present invention;
[0034] Figure 7 It is a schematic diagram of a three-dimensional observation system with regular sources and detection points in a specific embodiment 2 of the present invention;
[0035] Figure 8 (a) and (b) are schematic diagrams of random sampling in which the sampling step lengths of the detection points and the shot points are both 6 in a specific embodiment 2 of the present invention;
[0036] Fig. 9 It is a schematic diagram of a three-dimensional observation system for random sampling of seismic sources and detection points in a specific embodiment 2 of the present invention;
[0037] Fig.10 It is a schematic diagram of an original single shot of randomly sampled data in a specific embodiment 2 of the present invention;
[0038] Fig.11 It is a schematic diagram of a single shot after optimized reconstruction of random sampling data in a specific embodiment 2 of the present invention;
[0039] Fig.12 It is a schematic diagram of a three-dimensional observation system with regular sources and detection points in a specific embodiment 3 of the present invention;
[0040] Fig.13 It is a sampling schematic diagram of a shot point sampling step length of 5 in a specific embodiment 3 of the present invention;
[0041] Fig.14 It is a schematic diagram of a three-dimensional observation system for random sampling of seismic sources and detection points in a specific embodiment 3 of the present invention;
[0042] Fig.15 A schematic diagram of a single shot at a missing shot point position of the randomly sampled data in Example 3 of the present invention;
[0043] Fig.16 It is a schematic diagram of a single shot after optimized reconstruction of randomly sampled data in a specific embodiment 3 of the present invention. DETAILED DESCRIPTION
[0044] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.
[0046] like Figure 1 As shown, Figure 1 The flowchart of the random sampling and optimization reconstruction method of seismic data of the present invention is as follows. The random sampling and optimization reconstruction method of seismic data comprises the following steps:
[0047] Step 101, design an observation system R with fully complete and regular sources and detectors according to the actual surface and underground geological conditions of the exploration area.
[0048] Step 102, determining a sparse sampling coefficient P according to expected exploration costs and seismic data quality.
[0049] Step 103, for a regular detector matrix with N detectors, select a reasonable random sampling step size L, and the number of detectors in each step size is int represents the rounding operation. These detectors are regarded as a random sampling segment. The number of sampling points in the segment is determined as F = A*P according to the sparse sampling coefficient P. A random number is assigned to each detector in each random sampling segment, and the random numbers are sorted from large to small. The detectors corresponding to the first F random numbers are selected to complete the random sampling of the detectors.
[0050] Step 104: for a regular source matrix with M source points, select a reasonable random sampling step length K, and the number of source points in each step length is int represents the rounding operation. These focal points are regarded as a random sampling segment. The number of sampling points in the segment is determined as G = B*P according to the sparse sampling coefficient P. B random numbers are assigned to the focal points in each random sampling segment, and the random numbers are sorted from large to small. The focal points corresponding to the first G random numbers are selected to complete the random sampling of the focal points.
[0051] Step 105, outputting the source point and the sampling point of the detector as a random sampling design scheme for seismic data.
[0052] Step 106, after acquiring seismic data according to the seismic data random sampling design scheme, the seismic data is first transformed from the time-space domain to the frequency domain, and data reconstruction is achieved by removing noise and restoring data interpolation. After iterative inversion, it is finally transformed back to the time-space domain according to the sampling matrix of the regular observation system R to achieve optimal reconstruction of the seismic data.
[0053] In a specific embodiment 1 of the present invention, for a certain three-dimensional seismic exploration area, the random sampling and optimization reconstruction method of seismic data is implemented according to the operating steps of the present invention to illustrate the effect of the invention.
[0054] Step 1: According to the actual situation of the work area, design a set of observation systems with source and detector rules, such as Figure 2 As shown, the detection points and sources are arranged regularly, the black dots represent the detection points, and the white triangles represent the source points.
[0055] Step 2: According to the expected exploration cost and seismic data quality, the sparse sampling coefficient is determined to be 0.5, that is, random sampling is performed according to 33.3% of the original source points and 50% of the original detection points.
[0056] Step 3, select a random sampling step size of 4, that is, every 4 detection points are used as a random sampling segment to implement random sampling of the detector, and 4*50%=2 sampling points are randomly collected in each segment, such as Figure 3 As shown in (a), the black solid circles represent the selected sampling points, and the white dotted circles represent the unselected sampling points.
[0057] Step 4, determine the random sampling step length to be 3, that is, every 3 source points are used as a random sampling segment to achieve random sampling of the source points, and 3*33.3%=1 sampling point is randomly collected in each segment, such as Figure 3 As shown in (b), the black solid triangles represent the selected sampling points, and the white dotted triangles represent the unselected sampling points.
[0058] Step 5: Output the sampling points of the source point and the detector point as the random sampling observation system of seismic data, such as Figure 4 As shown, the black dots represent the detection points and the white triangles represent the source points.
[0059] Step 6, after collecting data according to the random sampling design scheme of the source point and the receiver point, the single shot is obtained, such as Figure 5 As shown in the figure, it can be seen that since the random sparse sampling coefficient of the detection point is 0.5, it is equivalent to the data missing half of the seismic traces. After implementing sparse transformation and data recovery on the seismic data, the regular observation system is used as the sampling matrix to implement the inverse transformation and complete the optimization reconstruction. The reconstructed single shot is as follows Figure 6 As shown in the figure, we can see that the missing half of the seismic traces have been effectively reconstructed with a high signal-to-noise ratio. Due to the use of random sparse sampling, the number of samples of source points and receiver points is reduced, effectively reducing the exploration cost.
[0060] In a specific embodiment 2 of the present invention, for certain seismic data, the random sampling and optimization reconstruction method of seismic data is implemented according to the operating steps of the present invention to illustrate the effect of the invention.
[0061] Step 1: According to the actual situation of the work area, design a set of observation systems with source and detector rules, such as Figure 7 As shown, the detection points and sources are arranged regularly, the black dots represent the detection points, and the white triangles represent the source points.
[0062] Step 2: According to the expected exploration cost and seismic data quality, the sparse sampling coefficient is determined to be 0.5, that is, random sampling is performed according to 50% of the original source points and original detection points.
[0063] Step 3, select a random sampling step of 6, that is, every 6 detection points are used as a random sampling segment to implement random sampling of the detector, and 6*50%=3 sampling points are randomly collected in each segment, such as Figure 8 As shown in (a), the black solid circles represent the selected sampling points, and the white dotted circles represent the unselected sampling points.
[0064] Step 4, determine the random sampling step length to be 6, that is, every 6 source points are used as a random sampling segment to achieve random sampling of the source points, and 6*50%=3 sampling points are randomly collected in each segment, such as Figure 8 As shown in (b), the black solid triangles represent the selected sampling points, and the white dotted triangles represent the unselected sampling points.
[0065] Step 5: Output the sampling points of the source point and the detector point as the random sampling observation system of seismic data, such as Fig. 9 As shown, the black dots represent the detection points and the white triangles represent the source points.
[0066] Step 6, after collecting data according to the random sampling design scheme of the source point and the receiver point, the single shot is obtained, such as Fig.10As shown in the figure, it can be seen that since the random sparse sampling coefficient of the detection point is 0.5, it is equivalent to the data missing half of the seismic traces. After implementing sparse transformation and data recovery on the seismic data, the regular observation system is used as the sampling matrix to implement the inverse transformation and complete the optimization reconstruction. The reconstructed single shot is as follows Fig.11 As shown, we can see that the missing half of the seismic trace has been effectively reconstructed with a high signal-to-noise ratio.
[0067] In a specific embodiment 3 of the present invention, for certain seismic data, the random sampling and optimization reconstruction method of seismic data is implemented according to the operating steps of the present invention to illustrate the effect of the invention.
[0068] Step 1: According to the actual situation of the work area, design a set of observation systems with source and detector rules, such as Fig.12 As shown, the detection points and sources are arranged regularly, the black dots represent the detection points, and the white triangles represent the source points.
[0069] Step 2: According to the expected exploration cost and seismic data quality, the sparse sampling coefficient is determined to be 0.6, that is, random sampling is performed according to 60% of the original source points, and the detection points remain unchanged.
[0070] Step 3: No random sparse sampling is performed on the detection points.
[0071] Step 4, determine the random sampling step length to be 5, that is, every 5 source points are used as a random sampling segment to achieve random sampling of the source points, and 5*60%=3 sampling points are randomly collected in each segment, such as Fig.13 As shown, the black solid triangle represents the selected sampling point, and the white dotted triangle represents the unselected sampling point.
[0072] Step 5: Output the sampling points of the source point and the detector point as the random sampling observation system of seismic data, such as Fig.14 As shown, the black dots represent the detection points and the white triangles represent the source points.
[0073] Step 6: After collecting data according to the random sampling design of the source point and the receiver point, select the single shot at the missing position and display it in Fig.15 In the figure, we can see that the entire shot set is missing due to the missing shot points. After implementing sparse transformation and data recovery on the seismic data, we use the regular observation system as the sampling matrix to implement inverse transformation and complete the optimized reconstruction. The reconstructed single shot is as follows: Fig.16 As shown, it can be seen that the missing single shot is effectively reconstructed with a high signal-to-noise ratio.
[0074] Finally, it should be noted that the above is only 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 aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
[0075] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.
Claims
1. A method for random sampling and optimization reconstruction of seismic data, characterized in that: The seismic data random sampling and optimization reconstruction method includes: Step 1: Design a seismic observation system with regular sources and receivers; Step 2: Determine the sparse sampling coefficient P based on the expected exploration cost and seismic data quality; Step 3: Select the random sampling step length, determine the random sampling segment, and perform random sampling of detection points; Step 4: Select the random sampling step length, determine the random sampling segment, and perform random sampling of the source point; Step 5: Output the random sampling design of seismic data; Step 6: Use the regular observation system as the sampling matrix to perform sparse inverse transformation on the randomly sampled seismic data to complete the optimization reconstruction; In step 3, for a regular detector matrix with N detectors, a reasonable random sampling step size L is selected, and the number of detectors in each step size is int represents the rounding operation. These detectors are used as a random sampling segment. The number of sampling points in the segment is determined by the sparse sampling coefficient P as F = A*P; Assign A random numbers to the detectors in each random sampling segment, sort them from large to small, select the detectors corresponding to the first F random numbers, and complete the random sampling of the detectors; In step 4, for a regular source matrix with M source points, a reasonable random sampling step size K is selected, and the number of source points in each step size is int represents the rounding operation. These source points are regarded as a random sampling segment. The number of sampling points in the segment is determined by the sparse sampling coefficient P as G = B*P; Assign B random numbers to the focal points in each random sampling segment, sort them from large to small, select the focal points corresponding to the first G random numbers, and complete the random sampling of focal points.
2. The seismic data random sampling and optimization reconstruction method according to claim 1, characterized in that: In step 1, a fully complete and regular observation system R of seismic sources and detectors is designed according to the actual surface and underground geological conditions of the exploration area.
3. The seismic data random sampling and optimization reconstruction method according to claim 1, characterized in that: In step 5, the sampling point outputs of the source and the detector are used as a random sampling design for seismic data.
4. The method for random sampling and optimization reconstruction of seismic data according to claim 1, characterized in that: In step 6, after acquiring seismic data according to the random sampling design scheme, the seismic data is first transformed from the time-space domain to the frequency domain, and data reconstruction is achieved by removing noise and restoring data interpolation. After iterative inversion, it is finally transformed back to the time-space domain according to the sampling matrix of the regular observation system R to complete the optimized reconstruction of the seismic data.