Broadband wave impedance modeling method based on multi-information fusion, medium and equipment

Abstract

The invention relates to the technical field of geophysical exploration, in particular to a broadband wave impedance modeling method based on multi-information fusion, a medium and equipment. The method comprises the following steps: establishing a constraint model of a broadband reflection coefficient and broadband wave impedance, and inputting background wave impedance and the broadband reflection coefficient after magnitude calibration in the model; the reflection coefficients are sparsely increased, and the broadband reflection coefficients from large to small are gradually utilized to invert broadband wave impedance from large to small; and outputting a broadband wave impedance result. Wavelets do not need to be extracted, and the broadband wave impedance is formed by iteratively fusing the background wave impedance and the broadband reflection coefficient. The low-frequency information of the broadband wave impedance is from the background wave impedance, the high-frequency information is from the broadband reflection coefficient, and the intermediate-frequency information is from the background wave impedance and the broadband reflection coefficient at the same time, so that the estimation of seismic wavelets is avoided, and the broadband wave impedance modeling precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geophysical exploration technology, in particular to a wideband wave impedance modeling method based on multi-information fusion, medium and equipment. BACKGROUND

[0002] The core problem of oil and gas seismic exploration can be theoretically summarized as follows: according to the measured pre-stack seismic data and the collected prior information about the elastic parameters to be inverted, the elastic parameters in the wide wave number band are estimated by using the wave equation and its various variants under the Bayes estimation theory. The elastic parameters in the wide wave number band are converted into lithology parameters by using the knowledge of rock physics, and the reservoir oil and gas bearing property is evaluated, and finally a reasonable drilling decision is made.

[0003] The elastic parameter estimation in the wide wave number band, i.e. the generalized seismic wave imaging, is the core technical link of the oil and gas industry. Modern seismic wave imaging is summarized as a high-dimensional elastic parameter estimation problem under the Bayes estimation theory, which can also be called the FWI (full waveform inversion) imaging problem. The FWI is an application example of the Bayes estimation theory in exploration seismology.

[0004] Based on the Bayes estimation theory, the wideband elastic parameter estimation, i.e. the wideband wave impedance estimation, is a strong nonlinear inverse problem with multiple solutions, which is based on the pre-stack seismic data and the prior information by using the FWI method. Under the actual data condition, it is difficult to obtain a result that can be used for reservoir description, and it is necessary to use as much information as possible to improve the accuracy of the wideband wave impedance estimation under the logic of multi-information fusion.

[0005] At present, the development of seismic wave inversion imaging technology has gradually shifted from the research on the inversion algorithm (Newton gradient iteration algorithm) to the research on the ideas and methods for improving the inversion accuracy under the constraint of multiple sources. Therefore, there is an urgent need for a wideband wave impedance modeling method based on multi-information fusion to improve the accuracy of wideband wave impedance modeling. SUMMARY

[0006] In order to avoid the above-mentioned problems existing in the prior art, the purpose of the present application is to provide a wideband wave impedance modeling method based on multi-information fusion, medium and equipment.

[0007] In order to achieve the above-mentioned purpose, the present application provides the following technical scheme: a wideband wave impedance modeling method based on multi-information fusion, comprising the following steps:

[0008] S1: establishing a constraint model of wideband reflection coefficient and wideband wave impedance, and inputting the background wave impedance and the wideband reflection coefficient after magnitude calibration in the constraint model;

[0009] S2: sparse lifting of the wideband reflection coefficient, using the wideband reflection coefficient from large to small to inverse the wideband wave impedance from large to small;

[0010] S3: when the wideband reflection coefficient and the wideband wave impedance reach the preset iteration stop condition, output the wideband wave impedance modeling result.

[0011] The application is further provided that the constraint model of the wideband reflection coefficient and the wideband wave impedance in step S1 is as follows:

[0012]

[0013]

[0014] Wherein, R (k) is the sparse reflection coefficient information extracted from the input wideband reflection coefficient r, R is the ideal wideband reflection coefficient, r is the input wideband reflection coefficient; β, γ, μ, η, λ are all hyperparameters, which are artificially given scalars; L is a low-pass filter operator, v bg is the input background wave impedance, v is the expected output wideband semi-log wave impedance; v (k) is the wideband wave impedance obtained after iteration, k is the iteration number, TV(v) is the total variation algorithm, β0 is the hyperparameter for controlling the sparsity of the reflection coefficient, D z is the difference operator of the wideband wave impedance along the z direction, q k <1 is the sparsity relaxation factor.

[0015] The application is further provided that the input background wave impedance v bg is three-dimensional data, and the input wideband reflection coefficient r is three-dimensional data after fidelity imaging, resolution improvement processing and magnitude calibration.

[0016] The application is further provided that the calculation formula of TV(v) is as follows:

[0017]

[0018] (D z v) m =[v i,j,m -v i,j,m-1 ],m=1,2,…,nz

[0019] (D x v) i =[v i,j,m -v i-1,j,m ],i=1,2,…,nx

[0020] (D y v) j =[v i,j,m -vi,j-1,m j=1,2,…,ny

[0021] Wherein, i, j, m are the subscripts of the expected output of the wideband semi-logarithmic wave impedance v in x, y, z directions, nx, ny, nz are the total number of sampling points of the expected output of the wideband semi-logarithmic wave impedance v in x, y, z directions, D x , D y , D z are the difference operators of the wideband wave impedance along x, y, z directions. z v m is the vector difference of the sampling points in the z direction. x v i is the vector difference of the sampling points in the x direction. y v j is the vector difference of the sampling points in the y direction.

[0022] The application is further provided that the step S2 is specifically that, after the sparse promotion of the wideband reflection coefficient r, the wideband reflection coefficient and the wideband wave impedance constraint model established in the step S1 are substituted, the constraint model is solved, the wideband reflection coefficient r and the wideband wave impedance x are iteratively updated and constrained to each other, the wideband wave impedance v from large to small is inverted from the wideband reflection coefficient r from large to small (k) .

[0023] The application also relates to an electronic device, which comprises:

[0024] a memory storing executable instructions;

[0025] a processor running the executable instructions in the memory to realize the above-mentioned wideband wave impedance modeling method based on multi-information fusion.

[0026] A computer readable storage medium stores a computer program, which is executed by a processor to realize the above-mentioned wideband wave impedance modeling method based on multi-information fusion.

[0027] The application aims to fuse multiple information to establish a wideband wave impedance model. The background wave impedance and the wideband reflection coefficient after magnitude calibration are fused, the reflection coefficient is sparse promoted in order to reduce the false image introduced by the reflection coefficient sidelobe to the wave impedance, the reflection coefficient and the wave impedance are iteratively updated and constrained to each other, and finally the wideband wave impedance is formed. Without extracting a wavelet, the obtained wideband wave impedance model is helpful for lithologic oil and gas reservoir exploration.

[0028] In summary, the beneficial effects of the above technical solutions of the application are as follows:

[0029] Compared with the conventional wave impedance modeling method, the method of the present application does not need to extract a wavelet, and forms a wideband wave impedance by iteratively fusing a background wave impedance and a wideband reflection coefficient. The low-frequency information of the wideband wave impedance comes from the background wave impedance, the high-frequency information comes from the wideband reflection coefficient, and the medium-frequency information comes from both the background wave impedance and the wideband reflection coefficient. The present application avoids the estimation of a seismic wavelet, and improves the precision of wideband wave impedance modeling. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed for the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0031] Figure 1 The method flowchart of the present application.

[0032] Figure 2 The wideband reflection coefficient of the input model data in step 1 of embodiment 1 of the present application.

[0033] Figure 3 The background wave impedance of the input model data in step 1 of embodiment 1 of the present application.

[0034] Figure 4 The wideband wave impedance modeling result obtained in step 3 of embodiment 1 of the present application.

[0035] Figure 5 The real wave impedance of the model data in step 1 of embodiment 1 of the present application.

[0036] Figure 6 The wideband reflection coefficient of the input actual data in step 1 of embodiment 2 of the present application.

[0037] Figure 7 The background wave impedance of the input actual data in step 1 of embodiment 2 of the present application.

[0038] Figure 8 The wideband wave impedance modeling result obtained in step 3 of embodiment 2 of the present application.

[0039] Figure 9 The three-dimensional display of the background wave impedance of the input actual data in step 1 of embodiment 3 of the present application.

[0040] Figure 10 The three-dimensional display of the wideband wave impedance modeling result obtained in step 3 of embodiment 3 of the present application.

[0041] Figure 11This is an 8 km depth slice of the broadband reflection coefficient of the actual data input in step 1 of Example 4 of the present invention.

[0042] Figure 12 This is an 8 km depth slice of the background wave impedance of the actual data input in step 1 of Example 4 of the present invention.

[0043] Figure 13 This is an 8 km depth slice of the broadband wave impedance modeling result obtained in step 3 of Example 4 of the present invention. DETAILED DESCRIPTION

[0044] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings of the present invention. Based on the embodiments of the present invention, other similar embodiments obtained by ordinary technicians in this field without making any creative work should fall within the scope of protection of the present invention.

[0045] The present invention will be further described below with reference to the accompanying drawings and preferred embodiments.

[0046] Example 1:

[0047] like Figure 1 As shown in FIG, a preferred embodiment of the present invention is a broadband wave impedance modeling method based on multi-information fusion, comprising the following steps:

[0048] S1: Establish a constraint model of broadband reflection coefficient and broadband wave impedance, input the background wave impedance and the broadband reflection coefficient after magnitude calibration into the constraint model; the broadband reflection coefficient of the model data input in this embodiment is as follows Figure 3 As shown, the background wave impedance of the model data is as follows Figure 2 shown. Figure 2 、 Figure 3 Here, z represents the depth direction of the formation, and x represents the lateral length of the formation.

[0049] The input background wave impedance v bg It is three-dimensional data, and the input broadband reflection coefficient r is three-dimensional data after fidelity imaging, resolution improvement processing, and magnitude calibration.

[0050] It should be noted that background wave impedance is derived from imaging velocity and density: background wave impedance = imaging velocity * density. The calibrated broadband reflection coefficient refers to an imaging result whose magnitude is consistent with the true subsurface reflection coefficient, for example, the reflection coefficient range is between [-1, 1]. However, most reflection coefficients are actually very small, on the order of 0.005. Therefore, the broadband reflection coefficient is an imaging result calibrated using the well logging reflection coefficient.

[0051] The constraint model of the wideband reflection coefficient and the wideband wave impedance is as follows:

[0052]

[0053]

[0054] wherein R (k) is sparse reflection coefficient information extracted from the input wideband reflection coefficient r, R is an ideal wideband reflection coefficient, r is the input wideband reflection coefficient; β, γ, μ, η, λ are all hyperparameters, which are artificially given scalars; L is a low-pass filtering operator, v bg is the input background wave impedance, and v is the expected output wideband semi-log wave impedance; v (k) is the wideband wave impedance obtained after iteration, k is the number of iterations, TV(v) is a total variation algorithm, β0 is a hyperparameter for controlling the sparsity of the reflection coefficient, D z is a difference operator of the wideband wave impedance along the z direction, q k <1 is a sparsity relaxation factor.

[0055]

[0056] (D z v) m =[v i,j,m -v i,j,m-1 ],m=1,2,…,nz

[0057] (D x v) i =[v i,j,m -v i-1,j,m ],i=1,2,…,nx

[0058] (D y v) j =[v i,j,m -v i,j-1,m ],j=1,2,…,ny

[0059] wherein i, j, m are subscripts of the expected output wideband semi-log wave impedance v in the x, y, z directions, respectively, nx, ny, nz are total numbers of sampling points of the expected output wideband semi-log wave impedance v in the x, y, z directions, respectively, D x , D y , D z are difference operators of the wideband wave impedance along the x, y, z directions, respectively.(D z v) m is a vector difference of the sampling points in the z direction, (D x v) i is a vector difference of the sampling points in the x direction, (D y v)j What is sought is the vector difference of the sampling point in the y direction.

[0060] S2: Sparsely improve the broadband reflection coefficient and use the broadband reflection coefficient from large to small to invert the broadband wave impedance from large to small; after sparsely improving r, substitute it into the constraint model of broadband reflection coefficient and broadband wave impedance established in step S1, solve the constraint model, make r and x constrain each other, iteratively update, and use r from large to small to invert x from large to small.

[0061] S3: When the broadband reflection coefficient and broadband wave impedance reach the preset iteration stop condition, the broadband wave impedance result is output. Figure 4 The actual wave impedance of the model data is shown as Figure 5 It can be seen that the broadband wave impedance result output by the embodiment of the present invention has a high similarity with the actual wave impedance of the model data, which verifies that the broadband wave impedance modeling method based on multi-information fusion of the present invention has high accuracy. Figure 4 、 Figure 5 Here z represents the depth of the formation, and x represents the lateral length of the formation.

[0062] Example 2:

[0063] A broadband wave impedance modeling method based on multi-information fusion. The difference between this embodiment and embodiment 1 is that the input in step S1 of this embodiment is Figure 6 The broadband reflection coefficient of the actual data shown, and Figure 7 The background wave impedance of the actual data is shown. The output broadband wave impedance result is as follows Figure 8 shown. Figure 6 、 Figure 7 、 Figure 8 Here z represents the depth of the formation, and x represents the lateral length of the formation.

[0064] Example 3:

[0065] A broadband wave impedance modeling method based on multi-information fusion. The difference between this embodiment and embodiment 1 is that the input in step S1 of this embodiment is Figure 9 The three-dimensional data of the background wave impedance of the actual data shown is as follows. Figure 10 As shown, Figure 9 、 Figure 10 Where z represents the formation depth, Imp represents the wave impedance, line represents the line number, and cdp represents the common depth point number.

[0066] Example 4:

[0067] A broadband wave impedance modeling method based on multi-information fusion. The difference between this embodiment and embodiment 1 is that the input in step S1 of this embodiment is Figure 113D data of 8km depth slice of wideband reflection coefficient of the actual data shown and Figure 12 3D data of 8km depth slice of background wave impedance of the actual data shown. The results of this embodiment are shown in Figure 13 Figure 11 Figure 12 Figure 13 In the above, Imp represents wave impedance, line represents line number, and cdp represents common depth point number.

[0068] Embodiments 3 and 4 prove that the wideband wave impedance modeling method based on multi-information fusion described in the present application is also applicable to 3D data.

[0069] Embodiment 5:

[0070] An electronic device, comprising:

[0071] a memory storing executable instructions;

[0072] a processor running the executable instructions in the memory to implement a wideband wave impedance modeling method based on multi-information fusion.

[0073] Embodiment 6:

[0074] A computer readable storage medium storing a computer program, the computer program being executed by a processor to implement a wideband wave impedance modeling method based on multi-information fusion.

[0075] The above only describes the preferred embodiments of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical solution falling within the concept of the present application shall fall within the protection scope of the present application. It should be noted that, for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application shall also be considered as falling within the protection scope of the present application.​​​

Claims

1. A broadband wave impedance modeling method based on multi-information fusion, characterized in that, The method comprises the following steps: S1: establishing a constraint model of broadband reflection coefficient and broadband wave impedance, inputting background wave impedance and broadband reflection coefficient after magnitude calibration into the constraint model; S2: sparsely promoting the broadband reflection coefficient, and inversing broadband wave impedance from large to small by using broadband reflection coefficient from large to small; S3: outputting a broadband wave impedance modeling result.

2. The method of claim 1, wherein, The constraint model of broadband reflection coefficient and broadband wave impedance in step S1 is as follows: where R (k) is the sparse reflection coefficient information extracted from the input broadband reflection coefficient r, R is the ideal broadband reflection coefficient, r is the input broadband reflection coefficient; β, γ, μ, η, λ are all hyperparameters, which are artificially given scalars; L is a low-pass filtering operator, v bg is the input background wave impedance, v is the expected output broadband semi-log wave impedance; v (k) is the broadband wave impedance obtained after iteration, k is the number of iterations, TV(v) represents the total variation algorithm, β0 is the hyperparameter for controlling the sparsity of the reflection coefficient, D z is the difference operator of the broadband wave impedance along the z direction, q k <1 is the sparsity relaxation factor.

3. The method of claim 2, wherein, The input background wave impedance v bg is three-dimensional data, and the input wideband reflection coefficient r is three-dimensional data after fidelity imaging, resolution improvement processing, and magnitude calibration.

4. The method of claim 3, wherein, The calculation formula of TV(v) is: (D z v) m = [v i,j,m -v i,j,m-1 ], m = 1, 2,..., nz (D x v) i = [v i,j,m -v i-1,j,m ], i = 1, 2,..., nx (D y in) j =[in i,j,m -v i,j-1,m ], j=1,2,…,ny where i, j, m are the sampling point subscripts of the desired output wideband semi-logarithmic wave impedance v in the x, y, z directions respectively, nx, ny, nz are the total number of sampling points of the desired output wideband semi-logarithmic wave impedance v in the x, y, z directions respectively, D x ,D y ,D z are the difference operators of the wideband wave impedance along the x, y, z directions respectively.

5. The method of claim 4, wherein, The step S2 is specifically as follows: after the r sparse lifting, the constraint model of the wideband reflection coefficient and the wideband wave impedance established in the step S1 is substituted, the constraint model is solved, r and v are iteratively updated and constrained to each other, and v from large to small is inverted from large to small r (k) .

6. An electronic device, comprising: The electronic device comprises: A memory storing executable instructions; A processor running the executable instructions in the memory to implement the method according to any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method according to any one of claims 1-5.

Images