Three-dimensional earthquake variable-speed mapping method

Through the three-dimensional seismic variable speed map method, including seismic data preprocessing, improved offset processing and machine learning algorithms, the problem of low analysis accuracy during seismic wave transmission in the existing technology is solved, and more accurate seismic imaging and underground are achieved Structural explanation.

CN119001841APending Publication Date: 2024-11-22SHENHUA XINJIE ENERGY
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411046975.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing seismic data interpretation methods have low analysis accuracy during complex seismic wave transmission, especially when the velocity changes greatly, there are large errors.

Method used

The three-dimensional seismic speed change graph method is adopted, including seismic data preprocessing, construction of initial velocity model, improved offset processing, optimized time-depth conversion, machine learning algorithm for velocity reanalysis and secondary offset processing to generate three-dimensional seismic speed change picture.

Benefits of technology

Improves the accuracy and accuracy of seismic data, provides clearer underground structural information, suitable for precise imaging in complex geological structures and multiple reflections.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119001841A_ABST
    Figure CN119001841A_ABST
Patent Text Reader

Abstract

The invention provides a three-dimensional earthquake variable-speed mapping method. The method comprises the following steps: acquiring earthquake data; constructing an initial velocity model, and performing velocity analysis and correction; performing migration processing on the seismic data by using an improved migration processing method and an initial velocity model to obtain the seismic data after migration; converting the offset seismic data to a depth domain through an optimized time-depth conversion method, and obtaining depth domain seismic data; in combination with a machine learning algorithm, carrying out speed re-analysis and re-correction on the depth domain seismic data to obtain a corrected speed optimization model; and according to the speed optimization model, carrying out secondary migration processing on seismic data to obtain an accurate seismic imaging result, and generating a three-dimensional seismic variable-speed map. Based on the provided method, the seismic imaging result is more accurate, and clearer underground structure information is provided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of seismic data interpretation, and in particular to a three-dimensional seismic variable velocity mapping method. Background Art

[0002] Two crucial parameters in seismic physics are reflection time and velocity. As exploration continues to deepen, higher requirements are placed on the development of seismic technology. From imaging accuracy to lithology analysis, from acquisition to data processing and interpretation, velocity information analysis and utilization are particularly important.

[0003] In seismic exploration, the concept of velocity is more often used in seismic data processing and interpretation. There are many concepts of velocity due to different application links. For example, in the process of seismic data processing, the dynamic correction velocity is the stacking velocity, the offset processing uses the offset velocity, and the time-to-depth conversion in the data interpretation process involves the average velocity, and the layer velocity concept is involved in the interpretation of stratum lithology. At present, the most commonly used methods are the average velocity method, the curved line method and the refraction line method. In addition, in data processing, the offset velocity field can be directly used to perform pre-stack depth migration to obtain depth domain data, and through depth domain interpretation, the depth domain structural map can be directly obtained. The average velocity method is the most commonly used time-to-depth conversion method. This method directly extracts the average velocity of a certain reflection interface from the stacking velocity field, and uses the time-to-depth conversion method to obtain the depth map of the reflection interface. The time-to-depth conversion using the velocity field formed by the stacking velocity is based on the assumption that the horizontal layered medium and the CMP gather reflection wave time-distance curve are hyperbolic to achieve velocity analysis. When the velocity changes greatly laterally, the above method will cause a large error. Therefore, it is necessary to provide a new seismic data interpretation method to solve the problems existing in the prior art. Summary of the invention

[0004] The purpose of the present invention is to provide a three-dimensional seismic variable velocity mapping method, aiming to solve the problems of complex analysis and low accuracy in the complex seismic wave transmission process.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A three-dimensional seismic variable velocity mapping method, the method comprising the following steps:

[0007] S1. Collecting seismic data and preprocessing the seismic data, wherein the preprocessing includes noise removal, filtering, and profile stitching;

[0008] S2. Construct an initial velocity model and perform velocity analysis and correction;

[0009] S3. Performing migration processing on the seismic data using an improved migration processing method and an initial velocity model to obtain migrated seismic data;

[0010] S4. Convert the migrated seismic data into a depth domain by an optimized time-depth conversion method, and obtain depth-domain seismic data;

[0011] S5. Combining a machine learning algorithm, reanalyzing and recalibrating the velocity of the depth domain seismic data to obtain a corrected velocity optimization model;

[0012] S6. Perform secondary migration processing of seismic data according to the velocity optimization model to obtain accurate seismic imaging results;

[0013] S7. Generate a three-dimensional seismic velocity map based on the seismic imaging results.

[0014] Optionally, the S1 further includes:

[0015] discretizing the seismic data into a three-dimensional voxel grid, each voxel representing a data point;

[0016] The data in the voxel grid is interpolated using a cubic spline interpolation method to convert discrete data points into a continuous seismic volume model.

[0017] Optionally, the initial velocity model is:

[0018]

[0019] Where v(x,y,z) represents the velocity of the underground medium at the coordinate point (x,y,z), x,y,z represent the horizontal, vertical and depth coordinates of the seismic record respectively; v 0 (x, y, z) represents the initial velocity distribution of the underground medium. In rocks, the longitudinal wave velocity is 5000-8000 m / s, the shear wave velocity is 3000-5000 m / s, and the density is 2500-3000 kg / m 3 ; An represents the coefficient for controlling amplitude; Bn represents the coefficient for controlling exponential decay; Cn represents the coefficient for controlling horizontal changes; Dn represents the coefficient for controlling horizontal phase; En represents the coefficient for controlling vertical changes; Fn represents the coefficient for controlling vertical phase.

[0020] Optionally, the offset processing is specifically:

[0021]

[0022]

[0023] Wherein, P(·) is the seismic data after migration, R(·) is the reflection coefficient, which represents the reflection characteristics of the underground medium at the position (x, y), S(·) is the original seismic data, T represents the time range of the seismic record, t represents the reflection time, w is the angular velocity, and f is the main frequency.

[0024] Optionally, S4 specifically includes:

[0025] S41. Calculate the depth value h using the optimized time-to-depth conversion method, the formula is as follows:

[0026] h=∫f 0 ·e (-α·z) ·v(z)·sinθdt

[0027] Among them, h represents the depth, f 0 is the influence factor at the surface, α is the attenuation coefficient, the value range of α is 0.1-0.75, v(z) represents the velocity at depth z, θ refers to the angle between the seismic wave and the bottom surface when propagating in the underground medium, and dt represents the time interval; the distance influence factor is introduced into the time-to-depth conversion method to ensure the conversion accuracy in complex geological structures or multiple reflections;

[0028] S42. According to the calculated depth value h, the offset seismic data is interpolated to the corresponding depth position to obtain the depth domain seismic data, specifically including:

[0029] S421. Determine the vertical sampling interval Δz of the depth domain seismic data;

[0030] S422. For each depth value h, find the two nearest depth sample points h 1 and h 2 ,h 1 ≤h≤h 2 ;

[0031] S423. Calculate the interpolation weight w of the depth domain seismic data according to the distance relationship between the depth sample points 1 and w 2 ;

[0032] S424. According to the interpolation weight w 1 and w 2 , and the corresponding seismic data d of the depth sample point 1 and d 2 , calculate the value d of the depth domain seismic data,

[0033] d=w 1 *d 1 +w 2 *d 2

[0034] Among them, w 1 +w 2 =1

[0035] S425. Repeat steps S421-S424 to interpolate all depth values ​​to obtain complete depth domain seismic data.

[0036] Optionally, in S5, the inversion velocity is corrected using a gradient descent method to obtain the velocity optimization model, and the specific steps are as follows:

[0037] S51. Initialize the initial velocity model v(x, y, z);

[0038] S52. Based on the initial velocity model, a synthetic seismic record d is generated by forward simulation. syn ;

[0039] S53. Calculate the synthetic seismic record d syn Observational seismic records obs The residual I between

[0040] I=d obs -d syn

[0041] S54. Calculate the gradient G of the residual I with respect to the v(x, y, z), and introduce a regularization constraint term J at the same time.

[0042]

[0043] Among them, λ is the regularization parameter,

[0044] S55. Update the velocity model v(x, y, z) to obtain an optimized velocity model v opt ,

[0045] v opt =v(x,y,z)+β*G

[0046] Where β is the learning rate

[0047] S56. Repeat steps S52-S55 until convergence conditions are reached and output the speed optimization model.

[0048] Optionally, in S6, the seismic data P after secondary migration s (x,y,z,t) is:

[0049]

[0050] Optionally, the S7 specifically includes:

[0051] S71. Calculate the seismic imaging result Q(x,y,z) at the coordinate point (x,y,z)

[0052]

[0053] S72. Display Q(x, y, z) in a three-dimensional volume rendering to intuitively display the velocity distribution of the underground medium.

[0054] Optionally, the three-dimensional volume rendering adopts at least one rendering method selected from voxel ray projection, isosurface rendering, and switching rendering.

[0055] Optionally, the interpreting using the depth domain seismic data further comprises:

[0056] By analyzing the reflection characteristics, reflection amplitude, and reflection coefficient in the seismic data, the underground lithology distribution, structural morphology, fault location, and fold structure are inferred; the reflection characteristics include reflection morphology, continuity, intensity, and spectrum.

[0057] The present invention has the following beneficial technical effects: the present invention provides a three-dimensional seismic variable velocity mapping method. By constructing an initial velocity model and performing velocity analysis and correction, a basic velocity reference is provided to provide accurate velocity information for subsequent processing steps; by introducing a reflection coefficient, the offset processing method is improved to reduce the offset error and improve the accuracy of seismic data; at the same time, a distance influence factor is introduced into the time-depth conversion method and a sine function fitting operation is performed on the seismic data to convert the multipath problem into a sine function frequency problem, thereby ensuring the conversion accuracy in complex geological structures or multiple reflections; the velocity re-analysis and re-correction of the depth domain seismic data is performed in combination with the gradient descent method to further improve the accuracy and precision of the velocity model; at the same time, the secondary offset processing operation makes the seismic imaging results more accurate and provides clearer underground structure information; three-dimensional seismic variable velocity mapping is generated based on the seismic imaging results, which can intuitively display the underground velocity changes and provide important references for geological interpretation and exploration decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A flow chart of a three-dimensional seismic velocity mapping method provided by an embodiment of the present invention.

[0060] Figure 2A cross-sectional diagram of the amplitude of the seismic imaging at the vertical interface provided for an embodiment of the present invention.

[0061] Figure 3 A three-dimensional view of peaks in seismic imaging is provided for embodiments of the present invention.

[0062] Figure 4 A seismic image drawn using voxel ray projection method is provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0064] The following describes the implementation methods of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, in the absence of conflict, the following embodiments and the features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work belong to the scope of protection of the present application.

[0065] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structures and / or functions described herein are merely illustrative. Based on the present application, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspect set forth herein can be used to implement an apparatus and / or practice a method.

[0066] It should also be noted that the illustrations provided in the following embodiments are only used to schematically illustrate the basic concept of the present application.

[0067] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the examples can be practiced without these specific details.

[0068] The purpose of the present invention is to provide an image-assisted MBR process control method, system, device and medium, aiming to solve the problems of low efficiency and low precision in MBR process control.

[0069] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0070] Reference Figure 1 , showing a three-dimensional seismic velocity mapping method according to an embodiment of the present application, the method comprising the following steps:

[0071] A three-dimensional seismic variable velocity mapping method, the method comprising the following steps:

[0072] S1. Collecting seismic data and preprocessing the seismic data, wherein the preprocessing includes noise removal, filtering, and profile stitching;

[0073] S2. Construct an initial velocity model and perform velocity analysis and correction;

[0074] S3. Performing migration processing on the seismic data using an improved migration processing method and an initial velocity model to obtain migrated seismic data;

[0075] S4. Convert the migrated seismic data into a depth domain by an optimized time-depth conversion method, and obtain depth-domain seismic data;

[0076] S5. Combining a machine learning algorithm, reanalyzing and recalibrating the velocity of the depth domain seismic data to obtain a corrected velocity optimization model;

[0077] S6. Perform secondary migration processing of seismic data according to the velocity optimization model to obtain accurate seismic imaging results;

[0078] S7. Generate a three-dimensional seismic velocity map based on the seismic imaging results.

[0079] Preferably, in S1, the preprocessing specifically includes:

[0080] Noise removal: First, the original seismic data is quality checked to eliminate abnormal data caused by possible instrument failure or human interference. Then, the data is denoised using filters or wavelet denoising methods to remove high-frequency noise and low-frequency background noise.

[0081] Filtering: Filtering seismic data can remove some irrelevant frequency components and improve the clarity of seismic signals. Common filtering methods include low-pass filtering, high-pass filtering, band-pass filtering, etc., and appropriate filter parameters are set at the same time.

[0082] Section stitching: In 3D seismic data processing, multiple seismic sections are usually collected and need to be stitched together to form a complete 3D data set. First, the different sections collected are time-aligned to ensure that the arrival time of the same seismic event on different sections is consistent. Then, according to the acquisition geometry and the overlapping parts of the seismic data, the sections are stitched together to form a continuous 3D seismic data set.

[0083] Preferably, the S1 includes:

[0084] discretizing the seismic data into a three-dimensional voxel grid, each voxel representing a data point;

[0085] The data in the voxel grid is interpolated using a cubic spline interpolation method to convert discrete data points into a continuous seismic volume model.

[0086] Exemplarily, cubic spline interpolation is used to interpolate discretized three-dimensional voxel grid data.

[0087] The specific process is as follows:

[0088] Suppose there is a 3D voxel grid, where the coordinates of each voxel are (x, y, z) and the corresponding value is f(x, y, z). To perform cubic spline interpolation, you need to calculate the interpolation function inside each voxel.

[0089] First, define an interpolation function L(x,y,z), which is composed of the interpolation function inside each voxel. For each voxel, the interpolation function can be expressed as:

[0090] L(x,y,z)=a+bx+cy+dz+ex 2 +fy 2 +gz 2 +hxy+iyz+jxz+kx 3 +ly 3 +mz 3

[0091] Among them, a, b, c, d, e, f, g, h, i, j, k, l, and m are coefficients to be determined.

[0092] At the eight vertices of each voxel, the values ​​of the interpolation function are consistent with the original data:

[0093] L(x1,y1,z1)=f(x1,y1,z1)L(x1,y1,z2)=f(x1,y1,z2)L(x1,y2,z1)=f(x1,y2,z1)L(x1,y2,z2)=f(x1,y2,z2) L(x2,y1,z1)=f(x2,y1,z1)L(x2,y1,z2)=f(x2,y1,z2)L(x2,y2,z1)=f(x2,y2,z1)L(x2,y2,z2)=f(x2,y2,z2)

[0094] On the six faces of each voxel, the first-order partial derivatives of the interpolation function agree with the original data:

[0095]

[0096] On the twelve sides of each voxel, the second-order partial derivatives of the interpolation function agree with the original data:

[0097]

[0098] By solving these equations, the interpolation function inside each voxel can be obtained, thus realizing cubic spline interpolation.

[0099] Optionally, the initial velocity model is:

[0100]

[0101] Where v(x,y,z) represents the velocity of the underground medium at the coordinate point (x,y,z), x,y,z represent the horizontal, vertical and depth coordinates of the seismic record respectively; v 0 (x, y, z) represents the initial velocity distribution of the underground medium. In rocks, the longitudinal wave velocity is 5000-8000 m / s, the shear wave velocity is 3000-5000 m / s, and the density is 2500-3000 kg / m 3; An represents the coefficient for controlling amplitude. A larger An value can enhance the amplitude of the reflected signal, and a smaller An value can reduce the amplitude. The general value is between 0.5-2; Bn represents the coefficient for controlling exponential attenuation. A larger Bn value can accelerate the attenuation of speed, and a smaller Bn value can reduce the attenuation. The general value is between 0.1-0.5; Cn represents the coefficient for controlling horizontal changes. A larger Cn value can enhance the lateral change of speed, and a smaller Cn value can reduce the change. The general value is between 0.1-1; Dn represents the coefficient for controlling horizontal phase. A larger Dn value can enhance the change of phase, and a smaller Dn value can reduce the change. The general value is between 0.1-1; En represents the coefficient for controlling vertical changes. A larger En value can enhance the longitudinal change of speed, and a smaller En value can reduce the change. The general value is between 0.1-1; Fn represents the coefficient for controlling vertical phase. A larger Fn value can enhance the change of phase, and a smaller Fn value can reduce the change. The general value is between 0.1-1. Among them, An, Bn, Cn, Dn, En, and Fn can be flexibly adjusted according to the properties of the underground medium and the expected velocity variation range, and the optimal values ​​can also be determined through experiments and inversion.

[0102] Preferably, the reflection characteristics of the underground medium are considered: the seismic data migration processing is to restore the reflection characteristics of the underground medium, and the introduction of the reflection coefficient influencing factor can more accurately simulate the reflection process of the underground medium, so that the processing result is more in line with the actual situation. Therefore, the migration processing is specifically as follows:

[0103]

[0104]

[0105] Wherein, P(·) is the seismic data after migration, R(·) is the reflection coefficient, which represents the reflection characteristics of the underground medium at the position (x, y), S(·) is the original seismic data, T represents the time range of the seismic record, t represents the reflection time, w is the angular velocity, and f is the main frequency.

[0106] Preferably, the reflection coefficient influencing factor can perform weighted processing on the seismic data, making the reflection signal more prominent, thereby improving the resolution of the seismic data. This is very important for the interpretation of underground structures and the identification of exploration targets. The reflection coefficient influencing factor can suppress multiple reflections and noise in the seismic data, making the processing results clearer and improving the quality of the seismic data. The introduction of the reflection coefficient influencing factor can better constrain the inversion problem in the seismic imaging process and improve the accuracy and reliability of the imaging results.

[0107] Preferably, in the seismic velocity analysis, the stratification factor is taken into consideration, and the present invention optimizes and deforms the conventional time-to-depth conversion formula, and introduces the influencing factor of distance conversion; at the same time, the role of introducing sin optimization in the time-to-depth conversion method of the present invention is to reduce the multipath problem in the depth domain seismic data. The multipath problem refers to the fact that when seismic waves propagate in the underground medium, they will be reflected, refracted and scattered through different paths, resulting in multiple repeated reflection waveforms in the seismic record, making seismic interpretation and analysis difficult. The purpose of introducing sin optimization is to convert the multipath problem into a frequency problem of the sine function by fitting the seismic data with a sine function. By analyzing the frequency of the sine function, the propagation speed and path of the seismic wave underground can be determined, thereby accurately inferring the underground geological information such as lithology, structure and fault.

[0108] Therefore, the S4 specifically includes:

[0109] S41. Calculate the depth value h using the optimized time-to-depth conversion method, the formula is as follows:

[0110] h=∫f 0 ·e (-α·z) ·v(z)·sinθdt

[0111] Among them, h represents the depth, f 0 is the influencing factor at the surface, α is the attenuation coefficient, and the value range of α is 0.1-0.75. Generally, harder strata have smaller attenuation coefficients, while softer strata have larger attenuation coefficients. v(z) represents the velocity at depth z, θ refers to the angle between the seismic wave and the underlying surface when it propagates in the underground medium, and dt represents the time interval.

[0112] S42. According to the calculated depth value h, the offset seismic data is interpolated to the corresponding depth position to obtain the depth domain seismic data, specifically including:

[0113] S421. Determine the vertical sampling interval Δz of the depth domain seismic data;

[0114] S422. For each depth value h, find the two nearest depth sample points h 1 and h 2 ,h 1 ≤h≤h 2 ;

[0115] S423. Calculate the interpolation weight w of the depth domain seismic data according to the distance relationship between the depth sample points 1 and w 2 ;

[0116] S424. According to the interpolation weight w 1 and w 2, and the corresponding seismic data d of the depth sample point 1 and d 2 , calculate the value d of the depth domain seismic data,

[0117] d=w 1 *d 1 +w 2 *d 2

[0118] Among them, w 1 +w 2 =1

[0119] S425. Repeat steps S421-S424 to interpolate all depth values ​​to obtain complete depth domain seismic data.

[0120] Optionally, by analyzing the reflection characteristics, reflection amplitude and reflection coefficient in the seismic data, the underground lithology, structure, faults and folds can be inferred. Reflection characteristic analysis refers to observing the reflection characteristics in the seismic data, including the morphology, continuity, intensity and spectrum of the reflection. Different lithologies and structures will produce different reflection characteristics. For example, carbonate rocks usually have uniform reflections, while sandstones and shales may show irregular reflections. Reflection amplitude analysis refers to analyzing the reflection amplitude changes in seismic data, which can infer the changes in underground lithology. For example, high-amplitude reflections may indicate the presence of harder lithology, while low-amplitude reflections may indicate the presence of softer lithology. Reflection coefficient analysis refers to calculating the reflection coefficient in the seismic data to infer the changes in underground lithology. The reflection coefficient is the amplitude ratio between the reflected wave and the incident wave. The reflection coefficients of different lithologies have different numerical ranges. By comparing the changes in the reflection coefficient, the changes in the underground lithology can be inferred. Seismic geological interpretation refers to seismic geological interpretation based on the analysis results of reflection characteristics, reflection amplitude and reflection coefficient, combined with geological knowledge and the geological background of the seismic profile. Based on different reflection characteristics and rock type changes, geological features such as underground structures, faults and folds can be inferred.

[0121] Usually, the velocity model is modified by the following methods:

[0122] Inversion method: Through seismic data inversion technology, the velocity distribution of the underground medium is inferred based on the characteristics of seismic wave propagation and the received seismic records. Common inversion methods include full waveform inversion, tomography, etc.

[0123] Prestack depth migration: When the velocity model is inaccurate, prestack depth migration can be performed to correct the velocity model in an iterative manner. By continuously adjusting the velocity model, the prestack depth migration results are more consistent with the actual observed data.

[0124] Post-stack depth migration: When the velocity model is inaccurate, post-stack depth migration can be performed to correct the velocity model in an iterative manner. By continuously adjusting the velocity model, the post-stack depth migration results are more consistent with the actual observed data.

[0125] Earthquake source coding: By changing the coding method of the earthquake source, more information can be provided to constrain the velocity model. Common earthquake source coding methods include multi-frequency coding, multi-angle coding, etc.

[0126] Joint inversion of seismic data: Joint inversion of different types of seismic data (such as seismic records, seismic waveforms, seismic reflections, etc.) can provide more constraints to correct the velocity model.

[0127] Preferably, the present invention uses a machine learning algorithm to invert and correct the velocity model. The present invention specifically uses a gradient descent method to invert and correct the velocity model, and introduces a regularization term constraint to improve the optimization effect of the velocity model. Therefore, in S5, the gradient descent method is used to correct the inversion velocity to obtain the velocity optimization model, and the specific steps are as follows:

[0128] S51. Initialize the initial velocity model v(x, y, z);

[0129] S52. Based on the initial velocity model, a synthetic seismic record d is generated by forward simulation. syn ;

[0130] S53. Calculate the synthetic seismic record d syn Observational seismic records obs The residual I between

[0131] I=d obs -d syn

[0132] S54. Calculate the gradient G of the residual I with respect to the v(x, y, z), and introduce a regularization constraint term J at the same time.

[0133]

[0134] Among them, λ is the regularization parameter,

[0135] S55. Update the velocity model v(x, y, z) to obtain an optimized velocity model v opt ,

[0136] v opt =v(x,y,z)+β*G

[0137] Where β is the learning rate

[0138] Preferably, β can be set to 0.00001;

[0139] S56. Repeat steps S52-S55 until convergence conditions are reached and output the speed optimization model.

[0140] Optionally, in S6, the seismic data P after secondary migration s (x,y,z,t) is:

[0141]

[0142] In the three-dimensional seismic variable velocity imaging analysis of the present invention, secondary migration processing is an important step. Due to the velocity changes in the three-dimensional seismic data, the propagation paths of seismic waves are different at different speeds, so the data needs to be velocity corrected. The secondary migration processing can more accurately locate the underground reflection interface and improve the accuracy and resolution of imaging. At the same time, the secondary migration processing can effectively suppress multiple reflections and migration illusions and improve the quality of the seismic profile. Through the secondary migration processing, the morphology and geometric characteristics of the underground structure can be better restored. The secondary migration processing can provide a more accurate underground velocity model, which provides a basis for subsequent seismic interpretation and quantitative analysis.

[0143] Optionally, the S7 specifically includes:

[0144] S71. Calculate the seismic imaging result Q(x,y,z) at the coordinate point (x,y,z)

[0145]

[0146] S72. Display Q(x, y, z) in a three-dimensional volume rendering to intuitively display the velocity distribution of the underground medium.

[0147] Optionally, the three-dimensional volume rendering adopts at least one rendering method selected from voxel ray projection, isosurface rendering, and switching rendering.

[0148] Reference Figure 4 , shows a seismic image drawn using voxel ray projection method according to an embodiment of the present application.

[0149] Optionally, various visualization methods, such as volume rendering, isosurface rendering, slice display, etc., may be used to generate three-dimensional seismic velocity maps.

[0150] Volume rendering: Displaying seismic velocity maps in the form of three-dimensional volume rendering can intuitively display the velocity distribution of the underground medium. Volume rendering algorithms such as voxel ray projection can be used for visualization.

[0151] Isosurface drawing: Displaying the seismic velocity map in the form of isosurfaces can clearly show the spatial variation of velocity distribution. Isosurface drawing algorithms such as Marching Cubes can be used for visualization.

[0152] Slice display: The seismic velocity map is displayed in slices, and the velocity distribution profile can be displayed on a two-dimensional plane. You can choose the appropriate slice direction and position for display, such as horizontal slice, vertical slice, etc.

[0153] In this specification, the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the embodiments described later, the description is relatively simple, and the relevant parts can be referred to the partial description of the previous embodiments.

[0154] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A three-dimensional seismic velocity mapping method, characterized in that: The method comprises the following steps: S1. Collecting seismic data and preprocessing the seismic data, wherein the preprocessing includes noise removal, filtering, and profile stitching; S2. Construct an initial velocity model and perform velocity analysis and correction; S3. Performing migration processing on the seismic data using an improved migration processing method and an initial velocity model to obtain migrated seismic data; S4. Convert the migrated seismic data into a depth domain by an optimized time-depth conversion method, and obtain depth-domain seismic data; S5. Combining a machine learning algorithm, reanalyzing and recalibrating the velocity of the depth domain seismic data to obtain a corrected velocity optimization model; S6. Perform secondary migration processing of seismic data according to the velocity optimization model to obtain accurate seismic imaging results; S7. Generate a three-dimensional seismic velocity map based on the seismic imaging results.

2. A three-dimensional seismic velocity mapping method according to claim 1, characterized in that: The S1 also includes: discretizing the seismic data into a three-dimensional voxel grid, each voxel representing a data point; The data in the voxel grid is interpolated using a cubic spline interpolation method to convert discrete data points into a continuous seismic volume model.

3. A three-dimensional seismic velocity mapping method according to claim 1, characterized in that: In S2, the initial velocity model is: Where v(x,y,z) represents the velocity of the underground medium at the coordinate point (x,y,z), where x,y,z represent the horizontal, vertical and depth coordinates of the seismic record respectively; v0(x,y,z) represents the initial velocity distribution of the underground medium. In rocks, the longitudinal wave velocity is between 5000-8000m / s, the shear wave velocity is between 3000-5000m / s, and the density is between 2500-3000kg / m 3 ; An represents the coefficient for controlling amplitude; Bn represents the coefficient for controlling exponential decay; Cn represents the coefficient for controlling horizontal changes; Dn represents the coefficient for controlling horizontal phase; En represents the coefficient for controlling vertical changes; Fn represents the coefficient for controlling vertical phase.

4. A three-dimensional seismic velocity mapping method according to claim 3, characterized in that: In S3, the offset processing is specifically as follows: Wherein, P(·) is the seismic data after migration, R(·) is the reflection coefficient, which represents the reflection characteristics of the underground medium at the position (x, y), S(·) is the original seismic data, T represents the time range of the seismic record, t represents the reflection time, w is the angular velocity, and f is the main frequency.

5. A three-dimensional seismic velocity mapping method according to claim 4, characterized in that: Specifically, S4 includes: S41. Calculate the depth value h using the optimized time-to-depth conversion method, the formula is as follows: h=∫f0·e (-α·z) ·v(z)·sinθdt Wherein, h represents the depth, f0 is the influence factor at the surface, α is the attenuation coefficient, the value range of α is 0.1-0.75, v(z) represents the velocity at the depth z, θ refers to the angle between the seismic wave and the bottom surface when propagating in the underground medium, and dt represents the time interval; the distance influence factor is introduced into the time-to-depth conversion method to ensure the conversion accuracy in complex geological structures or multiple reflections; S42. According to the calculated depth value h, the offset seismic data is interpolated to the corresponding depth position to obtain the depth domain seismic data, specifically including: S421. Determine the vertical sampling interval Δz of the depth domain seismic data; S422. For each depth value h, find the two nearest depth sample points h1 and h2, where h1≤h≤h2; S423. Calculate interpolation weights w1 and w2 of depth domain seismic data according to the distance relationship between depth sample points; S424. Calculate the value d of the depth domain seismic data according to the interpolation weights w1 and w2, and the seismic data d1 and d2 of the corresponding depth sample points, d=w1*d1+w2*d2 Among them, w1+w2=1 S425. Repeat steps S421-S424 to interpolate all depth values ​​to obtain complete depth domain seismic data.

6. A three-dimensional seismic velocity mapping method according to claim 5, characterized in that: In S5, the inversion velocity is corrected using the gradient descent method to obtain the velocity optimization model, and the specific steps are as follows: S51. Initialize the initial velocity model v(x, y, z); S52. Based on the initial velocity model, a synthetic seismic record d is generated by forward simulation. syn ; S53. Calculate the synthetic seismic record d syn Observational seismic records obs The residual I between I=d obs -d syn S54. Calculate the gradient G of the residual I with respect to the v(x, y, z), and introduce a regularization constraint term J at the same time. Among them, λ is the regularization parameter, S55. Update the velocity model v(x, y, z) to obtain an optimized velocity model v opt , v opt =v(x,y,z)+β*G Where β is the learning rate S56. Repeat steps S52-S55 until convergence conditions are reached and output the speed optimization model.

7. A three-dimensional seismic velocity mapping method according to claim 6, characterized in that: In S6, the seismic data P after secondary migration is s (x,y,z,t) is:

8. A three-dimensional seismic velocity mapping method according to claim 7, characterized in that: The S7 specifically includes: S71. Calculate the seismic imaging result Q(x,y,z) at the coordinate point (x,y,z) S72. Display Q(x, y, z) in a three-dimensional volume rendering to intuitively display the velocity distribution of the underground medium.

9. A three-dimensional seismic velocity mapping method according to claim 8, characterized in that: The three-dimensional volume rendering adopts at least one rendering method of voxel ray projection, isosurface rendering, and switching rendering.

10. A three-dimensional seismic velocity mapping method according to claim 1, characterized in that: The interpretation using the depth domain seismic data also includes: By analyzing the reflection characteristics, reflection amplitude, and reflection coefficient in the seismic data, the underground lithology distribution, structural morphology, fault location, and fold structure are inferred; the reflection characteristics include reflection morphology, continuity, intensity, and spectrum.

Citation Information

Cited By

  • Three-dimensional intelligent seismic velocity modeling method based on two-dimensional self-attention U-shaped network and low-rank adaptation

    CN120802347A