High-precision velocity analysis method and system based on sparse inversion

Abstract

The invention discloses a high-precision velocity analysis method and system based on sparse inversion, and the method comprises the steps: obtaining a convolution matrix of a seismic wave at different velocities based on the parameter information of the seismic wave; obtaining a velocity spectrum of the seismic gather based on the parameter information of the seismic waves; and performing sparse inversion processing on the velocity spectrum of each common midpoint gather in the seismic gather based on the convolution matrix to obtain an analysis result. According to the method, the sparse inversion defuzzification technology is applied to velocity analysis in seismic data processing for the first time, so that the resolution of a velocity spectrum in the velocity direction is improved, and a good effect is achieved in two-dimensional actual seismic data. According to the method, the propagation characteristics of seismic waves can be simulated and processed more accurately, so that the precision of velocity analysis is improved; according to the method, the sparse inversion theory is utilized to perform deconvolution processing on the velocity spectrum at each moment, so that the influence of a fuzzy window function can be effectively removed, and the resolution of the velocity spectrum in the velocity direction is further improved.

Description

Technical Field

[0001] The present invention relates to the field of seismic data processing, and in particular to a high-precision velocity analysis method and system based on sparse inversion. Background Art

[0002] Seismic velocity is one of the important data results of seismic data processing and interpretation. Conventional velocity spectra are obtained by the maximum stacking energy criterion, that is, the seismic data event axis is repeatedly corrected at different constant velocities, and then the corrected gathers are superimposed and the stacking energy is calculated to obtain the stacking velocity spectrum. According to the above velocity scanning strategy, different types of velocity spectra can be obtained by stacking different targets.

[0003] Taner and Koehler proposed a coherence criterion based on similarity coefficient in 1969. The coherence spectrum velocity analysis method based on normal time difference correction (NMO) corrected the hyperbolic phase axis of seismic data through velocity scanning, obtained the spatial coherence of the corrected data, and obtained the peak of the similarity coefficient spectrum to achieve velocity analysis. Neidell and Taner further elaborated this theory in 1971; Vries and Berkhout adopted the velocity analysis method (MEVA) based on minimum entropy theory; Geltinsky et al. used the phase and overall correlation algorithm to obtain velocity information; Biondi and Kostov studied the velocity analysis method based on eigenvalue operation; Toldi first proposed a velocity analysis algorithm that avoids manual participation, using the Dix formula to calculate the corresponding root mean square velocity, and then estimate the stacking velocity. Morozov and Smithson implemented a velocity analysis method based on phase correlation, and used the statistical hypothesis test method to calculate the velocity spectrum, which effectively improved the resolution of the velocity spectrum; Larner and Celis proposed a selective cross-correlation velocity analysis method, which selected part of the data between the minimum and maximum offsets for velocity analysis. This method takes into account the coherent cancellation of the formation phase axis stacking caused by the AVO phenomenon, which weakens the stacking energy. By comparing the local similarity between the gather and the reference trace, the weighting is performed, which improves the resolution of the velocity spectrum and can more effectively analyze the problem of the weakening of the stacking velocity spectrum energy caused by the AVO anomaly. Abbad et al. used a two-step strategy to realize velocity analysis for homogeneous transversely isotropic media (VTI media) with vertical symmetry axes; Khoshnavaz developed a velocity analysis method based on NMO, which achieved good velocity analysis results for seismic data of different AVO types; Wang and Gao used Poynting vectors to extract reverse time migration angle gathers, derive the formula for angle-gather velocity analysis, and use human-computer interaction to realize velocity analysis, providing a more accurate background velocity model for migration imaging.

[0004] Although the existing velocity analysis methods based on the traditional maximum stacking energy criterion have certain application value in seismic wave velocity analysis, they also have some technical and methodological defects and challenges. First, these methods rely too much on manual intervention, especially in the picking of velocity spectra and the selection of velocity windows, which increases the subjectivity and inconsistency of the analysis results. Second, traditional methods usually rely on low-resolution velocity spectra, which limits the accuracy and resolution of velocity analysis. In addition, these methods are sensitive to noise and interference in the data and are easily affected by these factors, resulting in instability and errors in the velocity spectrum. Furthermore, existing velocity analysis techniques also face challenges when dealing with complex geological conditions. For example, in low signal-to-noise ratio seismic data, velocity analysis is very difficult, and it is also difficult to make velocity corrections based on the initial velocity analysis results. The velocity adjustment on the gather lacks reliable data basis, and due to the large changes in adjacent gathers and the lack of lateral consistency, the results of velocity adjustment lack directionality, and the more adjustments are made, the worse the effect. In addition, the accuracy and resolution of velocity analysis are also affected by near-surface and underground geological factors and seismic acquisition factors. If the array length is too short, the velocity spectrum resolution will be reduced. The static correction problem generated near the surface also has a great influence on the velocity analysis. If the static correction is not done well, the velocity resolution will be greatly affected.

[0005] In summary, the existing velocity analysis technology has the following defects in seismic wave velocity analysis: low velocity resolution, sensitivity to noise, difficulty in handling complex geological conditions, dynamic correction problems, and bandwidth limitations. These problems limit the accuracy and reliability of velocity analysis and require further improvement and optimization. Summary of the invention

[0006] Aiming at the problem that the velocity spectrum obtained by conventional velocity analysis has poor energy concentration and low resolution in the velocity direction, the present invention proposes a sparse velocity analysis strategy, that is, a secondary processing of the velocity spectrum based on the sparse inversion method. It is believed that the velocity spectrum at each moment is formed by the convolution of the pulse function and the window function. Due to the existence of the fuzzy window function, the resolution of the velocity spectrum in the velocity direction is reduced. Based on this, the present technology uses the hyperbolic phase axis of the known velocity to extract the deconvolution operator in the velocity direction, and then uses the sparse inversion theory to deconvolve the velocity spectrum at each moment to remove the influence of the fuzzy window function. In the deconvolution process, we apply non-negative constraints and use the existing alternating direction multiplier method (ADMM) to optimize and solve the established target equation. Since the non-negative constraints are added to the equation, the high-resolution velocity spectrum after solution is positive, and the high-resolution velocity spectrum after "window removal" is obtained, that is, the sparse velocity spectrum. The algorithm is a secondary processing of the velocity spectrum with high computational efficiency, and can complete high-resolution velocity spectrum analysis and velocity modeling.

[0007] To achieve the above object, the present invention provides a high-precision velocity analysis method based on sparse inversion, the steps comprising:

[0008] Based on the parameter information of the seismic wave, the convolution matrix of the seismic wave at different speeds is obtained;

[0009] Get the initial velocity spectrum of the seismic gather;

[0010] The initial velocity spectrum is sparsely inverted based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

[0011] Preferably, the parameter information includes: velocity analysis interval, maximum and minimum scanning speeds and offset information; the method for obtaining the convolution matrix includes: based on the acquired parameter information, synthesizing a hyperbola of known specific velocity, conducting conventional velocity analysis, extracting sub-waves at different velocities of seismic waves, and forming the convolution matrix.

[0012] Preferably, the method for obtaining the velocity spectrum comprises: firstly determining a parameter range for performing conventional velocity analysis based on the parameter information, and performing conventional velocity analysis on the seismic trace gather based on the parameter range to obtain the velocity spectrum.

[0013] Preferably, the method for performing the conventional velocity analysis comprises: performing velocity scanning by repeatedly performing dynamic correction on seismic data to obtain the velocity spectrum, the steps comprising:

[0014] For each common center gather in the seismic gather, the NMO time difference Δt is calculated according to its offset distance h and velocity v:

[0015]

[0016] Where t0 is the two-way travel time. For each seismic trace i, apply its offset information h i and the NMO time difference Δt, and the adjusted seismic trace r is obtained. NMO (t,hi ) :

[0017] T NMo (t,h i )=r(t0-△t(h i , v))

[0018] After the dynamic correction, the amplitudes of all seismic traces at the same time t are superimposed to calculate the superposition energy S(t, v). This process is expressed as:

[0019]

[0020] Where N is the total number of seismic traces in the CMP gathers involved in the stacking;

[0021] The velocity analysis method based on the similarity coefficient criterion determines the optimal dynamic correction velocity according to the size of the similarity coefficient. The coherent velocity spectrum S is defined as:

[0022]

[0023] Wherein, f is the analysis window length, r is the sampling point amplitude; by changing the speed v one by one, speed scanning is realized, and the superposition energy or coherence at different speeds is further calculated, and then the initial speed spectrum is obtained.

[0024] Preferably, the method for obtaining the analysis result includes: using a sparse inversion algorithm, combining the convolution matrix and non-negative constraints to perform deconvolution processing on the velocity spectrum of each common center point set in the seismic set, using an alternating direction multiplier method, optimizing and solving the established inversion target equation to obtain the analysis result.

[0025] The present invention also provides a high-precision velocity analysis system based on sparse inversion, the system is used to implement the above method, including: a convolution construction module, a velocity spectrum acquisition module and an analysis module;

[0026] The convolution construction module is used to obtain the convolution matrix of the seismic wave at different speeds based on the parameter information of the seismic wave;

[0027] The velocity spectrum acquisition module is used to acquire the initial velocity spectrum of the seismic gather;

[0028] The analysis module is used to perform sparse inversion processing on the initial velocity spectrum based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

[0029] Preferably, the parameter information includes: velocity analysis interval, maximum and minimum scanning speeds and offset information; the workflow of the convolution construction module includes: based on the acquired parameter information, synthesizing a hyperbola of known specific velocity, conducting conventional velocity analysis, extracting sub-waves at different velocities of seismic waves, and forming the convolution matrix.

[0030] Preferably, the workflow of the velocity spectrum acquisition module includes: firstly determining a parameter range for conventional velocity analysis based on the parameter information, and performing conventional velocity analysis on the seismic trace gather based on the parameter range to obtain the velocity spectrum.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention applies sparse inversion defuzzification technology to the velocity analysis of seismic data for the first time to improve the resolution of the velocity spectrum in the velocity direction, and achieves good results in two-dimensional actual seismic data; by extracting sub-waves at each velocity from the conventional velocity analysis results, a deconvolution operator is established. This method can more accurately simulate and process the propagation characteristics of seismic waves, thereby improving the accuracy of velocity analysis; the present invention uses sparse inversion theory to perform deconvolution processing on the velocity spectrum at each moment, which can effectively remove the influence of the fuzzy window function, thereby improving the resolution of the velocity spectrum in the velocity direction. This processing method not only improves the resolution of the velocity spectrum, but also enhances the ability to indicate velocity anomalies of seismic data; through the "window removal" processing, the algorithm can obtain a high-resolution sparse velocity spectrum. This velocity spectrum has a higher resolution in the velocity direction and can more finely reflect the velocity changes of the strata, which is particularly important for velocity analysis under complex geological conditions and can provide more accurate underground structure information. This technology can realize modular processing and has a high promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0034] Figure 1 A schematic diagram of a method flow of an embodiment of the present invention;

[0035] Figure 2 A high-precision velocity analysis technology processing flow chart based on sparse inversion according to an embodiment of the present invention;

[0036] Figure 3 The following is a schematic diagram of the deconvolution operator manufacturing process according to an embodiment of the present invention; wherein: Figure 3 (a) is the artificially synthesized reflection coefficient model; Figure 3 (b) To perform reaction correction using different speeds; Figure 3 (c) is stacking velocity analysis; Figure 3 (d) is the deconvolution operator;

[0037] Figure 4 Schematic diagram of artificially synthesized common center point gather data of an embodiment of the present invention; Figure 4 (a) To adopt Figure 5 (c) The linear velocity synthesized seismic gather data shown by the dashed arrows; Figure 4 (b) For adoption Figure 5 (c) The variable velocity synthesized seismic gather data is shown by the solid arrow in the middle; Figure 4 (c) Figure 4 (a) with Figure 4 (b) The superposition data;

[0038] Figure 5 is the velocity spectrum corresponding to the artificially synthesized common center point gather data of the embodiment of the present invention; wherein Figure 5 (a) for Figure 4 (c) The stacked velocity spectrum obtained by velocity analysis of the data shown; Figure 5 (b) for Figure 4 (c) Coherent velocity spectrum obtained by velocity analysis of the data shown; Figure 5 (c) for Figure 4 (a) The sparse velocity spectrum obtained after secondary processing of the velocity spectrum shown;

[0039] Figure 6 is the velocity field and post-stack data of an embodiment of the present invention; wherein, Figure 6 (a) is the velocity field after smoothing; Figure 6 (b) is the data after superposition;

[0040] Figure 7 is the 250th CMP gather of the embodiment of the present invention and the velocity spectrum obtained by velocity analysis; wherein, Figure 7 (a) is the CMP gather; Figure 7 (b) is the superposition velocity spectrum; Figure 7 (c) is a sparse velocity spectrum; Figure 7 (d) is the CMP gather after NMO correction;

[0041] Figure 8 is the 750th CMP gather of the embodiment of the present invention and the velocity spectrum obtained by velocity analysis; wherein, Figure 8 (a) is the CMP gather; Figure 8 (b) is the superposition velocity spectrum; Figure 8 (c) is a sparse velocity spectrum; Figure 8 (d) is the CMP gather after NMO correction;

[0042] Fig. 9 This is a superposition diagram of 2110m / s velocity and post-stack data in an embodiment of the present invention; wherein, Fig. 9 (a) is the superposition velocity spectrum; Fig. 9 (b) is the sparse velocity spectrum;

[0043] Fig.10 This is a superposition diagram of the 3220m / s velocity and post-stack data in an embodiment of the present invention; wherein, Fig.10 (a) is the superposition velocity spectrum; Fig.10 (b) is the sparse velocity spectrum. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] 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.

[0046] Embodiment 1

[0047] like Figure 1 , Figure 2 As shown, it is a schematic diagram of the method flow of the present invention, and the steps include:

[0048] S1. Based on the parameter information of the seismic wave, the convolution matrix of the seismic wave at different speeds is obtained.

[0049] The parameter information includes: velocity analysis interval, maximum and minimum scanning velocity and offset information. Based on the acquired seismic wave parameter information, a hyperbola with a known specific velocity is synthesized, conventional velocity analysis is carried out, and sub-waves at different velocities are extracted to form a convolution matrix, such as Figure 3 (d) as shown.

[0050] S2. Obtain the initial velocity spectrum of the seismic gather.

[0051] Conventional velocity analysis achieves velocity scanning by repeatedly performing dynamic correction on seismic wave data to obtain the relationship between velocity and energy, which is the velocity spectrum. Based on the velocity spectrum, stacking velocity or coherence velocity can be obtained, thereby achieving the distinction between multiple waves and velocity modeling.

[0052] The purpose of the dynamic correction is to adjust the difference in the arrival time of seismic waves caused by the change in underground velocity. For horizontal strata, the hyperbolic phase axis is corrected to be horizontal. Then, for each CMP gather (common center gather) in the seismic gather, the dynamic correction time difference Δt is calculated according to its offset distance h and velocity v:

[0053]

[0054] Where t0 is the two-way travel time. For each seismic trace i, apply its offset information h i and the NMO time difference Δt, and the adjusted seismic trace r is obtained. NMO (t,h i ):

[0055] r NMO (t,h i )=r(t0-△t(hi ,v)) (2)

[0056] After the dynamic correction, the amplitudes of all seismic traces at the same time t are superimposed to calculate the superposition energy S(t, v). This process can be expressed as:

[0057]

[0058] Where N is the total number of seismic traces in the CMP gathers involved in the stacking.

[0059] The velocity analysis method based on the similarity coefficient criterion determines the optimal dynamic correction velocity according to the size of the similarity coefficient. The coherent velocity spectrum S is defined as:

[0060]

[0061] Among them, f is the analysis window length, r is the sampling point amplitude; by changing the speed v one by one, the speed scan is realized, and the superposition energy or coherence at different speeds is further calculated, and then the speed spectrum (i.e., the initial speed spectrum) is obtained. However, the initial speed spectrum is formed by the convolution of the pulse function and the window function. Due to the existence of the fuzzy window function, the resolution of the initial speed spectrum in the speed direction is reduced.

[0062] S3. Perform sparse inversion processing on the initial velocity spectrum based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

[0063] The velocity spectrum of each common center point (CMP) in the seismic gathers is deconvolved by using a sparse inversion algorithm combined with a convolution matrix and non-negative constraints. The established inversion target equation is optimized and solved using the existing alternating direction method of multipliers (ADMM) to obtain the analysis results.

[0064] Specifically, for seismic data, the actual velocity spectrum S(t, v) of the hyperbolic phase axis indicated by the CMP gather in the seismic gather at a point will deviate from the ideal high-resolution velocity spectrum A(t, v). The fuzzy phenomenon of the velocity spectrum can be described by the convolution of the high-resolution velocity spectrum A(t, v) and the fuzzy window function W(v) in the velocity direction. The window function W(v) describes the influence of the dynamic correction error on the velocity spectrum. The blurred velocity spectrum S(t, v) can be expressed as:

[0065] S(t,υ)=A(t,υ)*W(v) (5)

[0066] Where, the window function W(v) can be regarded as the response function of the velocity error. If the velocity is very accurate, the window function will be very sharp and mainly concentrated at the ideal position of the velocity spectrum. If the dynamic correction is inaccurate, the energy distribution of the velocity spectrum in the velocity direction becomes less concentrated, resulting in blurring.

[0067] Assuming that the low-resolution velocity spectrum S(t, v) obtained through velocity analysis can be regarded as the convolution of the impulse response function x(v) and the fuzzy function W(v) in the velocity direction at each time point, the modified formula (5) is:

[0068] S(v)=W(v)*x(v) (6)

[0069] Furthermore, the convolution process is converted into a matrix form, and for each time point

[0070]

[0071] in, is a convolution matrix with a Toeplitz structure, is the absolute value of the signal, indicating the amplitude of the signal. is the instantaneous speed curve corresponding to the time. and the window matrix W, then the desired high-resolution velocity spectrum It can be obtained by deconvolution of the instantaneous velocity curve through the sparse inversion algorithm, that is:

[0072]

[0073] Among them, λ is the regularization parameter, which controls the sparsity of the solution. When λ is large, the value in formula (8) The sparser the velocity spectrum, the higher the resolution, but it is also easy to lose weak signals. It is necessary to add a non-negative penalty term to equation (8). When it is less than zero, a penalty is given to increase the value of the objective function, forcing it to search for the best value in the feasible domain greater than or equal to 0.

[0074]

[0075] in, is the time index. The velocity spectrum of each time point is solved by looping Substituting it into equation (9) and solving it, we can obtain a high-resolution velocity spectrum. It is worth noting that equation (9) only needs to input and W can obtain a high-resolution velocity spectrum. It is sparse, so it is called sparse velocity spectrum, and the corresponding solution process is called sparse velocity analysis. In addition to the superposition velocity spectrum used in this article, this idea is also applicable to the coherent velocity spectrum.

[0076] Formula (9) belongs to a constrained optimization problem, and the external penalty function method needs to be considered. By applying a quadratic external penalty constraint, Formula (9) is modified to an unconstrained optimization problem. For each time index, we have:

[0077]

[0078] Where β is a penalty function used to ensure the non-negativity of the velocity spectrum. express For each element in hour will be larger when When , the constraint is 0. Formula (10) is simplified to:

[0079]

[0080] in, Because the velocity spectrum is not necessarily the same at each time point, the regularization parameter λλ is set to c is a constant, that is, when solving equation (11), the signal adaptive regularization parameter corresponding to each time point can ensure the stability of the solution.

[0081] In this embodiment, since the proposed sparse velocity analysis is a secondary processing of the existing velocity spectrum, it is easy to obtain the reference velocity based on the initial velocity spectrum, and the velocity can be used for the variable speed scanning of the next velocity analysis. When the reference velocity is provided in advance, if it is allowed to tolerate the reduction of the resolution in some time directions, then the above-mentioned Gaussian smoothing can be directly added to the sparse inversion solution process. Specifically, the local inclination can be used to obtain the general trend of the reference velocity.

[0082] First, modify equation (11) so that the velocity spectrum at all times can be calculated. At this time, λ is time-varying, that is,

[0083]

[0084] Given the offset h, it is easy to use the inverse tangent function to get the reference velocity υ ref The local inclination angle θ(v, h) is obtained, and then a velocity-offset trajectory set {v ref (h) = θ(v, h)}, and further construct a guided filtering strategy to smooth the data along the trajectory direction to achieve the purpose of suppressing noise, that is,

[0085]

[0086] Where γ is the regularization operator that controls the smoothing strength, For A T The construction of guided filtering.

[0087] like Figure 3The figure shows the process of making a deconvolution operator. In the calculation process, the seismic wavelet can be extracted in advance and the model can be established with the reflection coefficient ( Figure 3 a), and then apply different speeds (V1, V2, V3, V4) to it for reaction correction ( Figure 3 b) and conduct speed analysis ( Figure 3 c), then the deconvolution operator W(v) in the velocity direction can be obtained by data drive ( Figure 3 d).

[0088] In this embodiment, Gaussian smoothing and denoising is also performed on the processed velocity spectrum (ie, the analysis result), but this step can be skipped if a reference velocity or a local inclination angle is used.

[0089] Embodiment 2

[0090] The advantages of the method of the present invention will be described in detail below in conjunction with this embodiment.

[0091] First, artificial synthetic data is used to test the performance of the method of the present invention, and the sampling time is 0-8s. Based on the convolution principle, 30Hz Ricker wavelet is used to synthesize the smoothed data, and then the reaction correction is performed at different speeds to obtain the following: Figure 4 Gather data shown. Figure 4 The true speed of a corresponds to Figure 5 The straight line indicated by the dashed arrow; Figure 4 The real speed of b corresponds to Figure 5 The curve indicated by the solid arrow in the figure. The velocity analysis range is set to 1000-5000m / s, and the stacking velocity analysis, coherent velocity analysis and sparse velocity analysis methods proposed in this patent are used to obtain the corresponding velocity spectra, as shown in Figure 5 As shown, it is obvious that the sparse velocity analysis proposed in this patent has better focus and can more accurately indicate the position of the true velocity.

[0092] The performance of the method of the present invention is further tested by using CMP gathers of actual data collected by marine streamers. The data has been preprocessed by side waves, linear interference, multiple wave suppression, deconvolution, etc., and experienced processing personnel have picked up a relatively accurate velocity spectrum ( Figure 6 a), can better complete the dynamic correction stacking ( Figure 6 b). Based on the existing reference speed ( Figure 6 a), we perform deconvolution processing on the stacked velocity spectrum in the velocity direction and extract the 250th and 750th CMP gathers, as shown in Figure 7 and Figure 8 As shown in the figure, the dotted line at the center of the speed spectrum is the more accurate speed provided by the processing personnel. Figure 7 , Figure 8It can be found that after deconvolution processing, the energy of the original superimposed velocity spectrum in the velocity direction is greatly compressed ( Figure 7 b. Figure 8 b), the energy clusters have better aggregation and are more compatible with the pre-provided speed; Figure 7 c It can be found that at around 2.4s, the wave velocity lower than the effective wave velocity may be the residual multiple waves; then, the profiles at different velocities (2110m / s, 3220m / s) were obtained by using stacking velocity analysis and sparse velocity analysis. Fig. 9 and Fig.10 As shown in the figure, it is a superposition diagram of different velocities and post-stack data. The velocity at this time has been normalized to indicate the possibility of velocity anomaly in the post-stack data. The superposition diagram section shows that the method proposed in this patent has higher resolution and higher analysis accuracy than the traditional stacking velocity spectrum, and can more clearly and accurately indicate the velocity anomaly distribution of the section.

[0093] Embodiment 3

[0094] The present embodiment also provides a high-precision velocity analysis system based on sparse inversion, comprising: a convolution construction module, a velocity spectrum acquisition module and an analysis module; the convolution construction module is used to obtain a convolution matrix of seismic waves at different velocities based on parameter information of seismic waves; the velocity spectrum acquisition module is used to obtain an initial velocity spectrum of a seismic gather; and the analysis module is used to perform sparse inversion processing on the initial velocity spectrum based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

[0095] The parameter information includes: velocity analysis interval, maximum and minimum scanning speeds, and offset information; the workflow of the convolution construction module includes: based on the acquired parameter information, synthesizing a hyperbola with known specific velocity, conducting conventional velocity analysis, extracting sub-waves at different seismic wave velocities, and forming a convolution matrix.

[0096] The workflow of the velocity spectrum acquisition module includes: firstly, determining the parameter range for conventional velocity analysis based on parameter information, performing conventional velocity analysis on the seismic gathers based on the parameter range, and obtaining the velocity spectrum.

[0097] The workflow of the analysis module includes: using the sparse inversion algorithm, combining the convolution matrix and non-negative constraints to deconvolve the velocity spectrum of each common center point in the seismic gather, optimizing and solving the inversion target equation through the alternating direction multiplier method, and obtaining the analysis results.

[0098] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A high-precision velocity analysis method based on sparse inversion, characterized in that the steps include: Based on the parameter information of the seismic wave, the convolution matrix of the seismic wave at different speeds is obtained; Get the initial velocity spectrum of the seismic gather; The initial velocity spectrum is sparsely inverted based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

2. The high-precision velocity analysis method based on sparse inversion according to claim 1 is characterized in that: The parameter information includes: velocity analysis interval, maximum and minimum scanning speeds and offset information; the method for obtaining the convolution matrix includes: based on the acquired parameter information, synthesizing a hyperbola with known specific velocity, conducting conventional velocity analysis, extracting sub-waves at different velocities of seismic waves, and forming the convolution matrix.

3. The high-precision velocity analysis method based on sparse inversion according to claim 1 is characterized in that: The method for obtaining the velocity spectrum includes: firstly determining a parameter range for conventional velocity analysis based on the parameter information, and performing conventional velocity analysis on the seismic trace gather based on the parameter range to obtain the velocity spectrum.

4. The high-precision velocity analysis method based on sparse inversion according to claim 3 is characterized in that: The method for performing the conventional velocity analysis includes: performing velocity scanning by repeatedly performing dynamic correction on seismic data to obtain the velocity spectrum, the steps including: For each common center gather in the seismic gather, the NMO time difference Δt is calculated according to its offset distance h and velocity v: Where t0 is the two-way travel time; for each seismic trace i, its offset information h is applied i and the NMO time difference Δt, and the adjusted seismic trace r is obtained. NMO (t,h i ): r NMO (t,h i )=r(t0-Δt(h i ,v)) After the dynamic correction, the amplitudes of all seismic traces at the same time t are superimposed to calculate the superposition energy S(t, v). This process is expressed as: Where N is the total number of seismic traces in the CMP gathers involved in the stacking; The velocity analysis method based on the similarity coefficient criterion determines the optimal dynamic correction velocity according to the size of the similarity coefficient. The coherent velocity spectrum S is defined as: Wherein, f is the analysis window length, r is the sampling point amplitude; by changing the speed v one by one, speed scanning is realized, and the superposition energy or coherence at different speeds is further calculated, and then the initial speed spectrum is obtained.

5. The high-precision velocity analysis method based on sparse inversion according to claim 1 is characterized in that: The method for obtaining the analysis result comprises: using a sparse inversion algorithm, combining the convolution matrix and the non-negative constraint to perform deconvolution processing on the velocity spectrum of each common center point gather in the seismic gather, using an alternating direction multiplier method, optimizing and solving the established inversion target equation, and obtaining the analysis result.

6. A high-precision velocity analysis system based on sparse inversion, the system being used to implement the method according to any one of claims 1 to 5, characterized in that: include: Convolution building module, velocity spectrum acquisition module and analysis module; The convolution construction module is used to obtain the convolution matrix of the seismic wave at different speeds based on the parameter information of the seismic wave; The velocity spectrum acquisition module is used to acquire the initial velocity spectrum of the seismic gather; The analysis module is used to perform sparse inversion processing on the initial velocity spectrum based on the convolution matrix to obtain a high-precision velocity spectrum of the seismic gather.

7. The high-precision velocity analysis system based on sparse inversion according to claim 6 is characterized in that: The parameter information includes: velocity analysis interval, maximum and minimum scanning speeds and offset information; the workflow of the convolution construction module includes: based on the acquired parameter information, synthesizing a hyperbola with a known specific velocity, conducting conventional velocity analysis, extracting sub-waves at different velocities of seismic waves, and forming the convolution matrix.

8. The high-precision velocity analysis system based on sparse inversion according to claim 6, characterized in that: The workflow of the velocity spectrum acquisition module includes: firstly, determining a parameter range for conventional velocity analysis based on the parameter information, and performing conventional velocity analysis on the seismic trace gather based on the parameter range to obtain the velocity spectrum.

Images