A multi-channel pre-stack waveform inversion method and device

Abstract

An embodiment of the present invention provides a multi-channel pre-stack waveform inversion method and apparatus, which solves the problems of large inversion error and insufficient inversion resolution. A multi-channel pre-stack waveform inversion method provided by an embodiment of the present invention includes: Step 1: Obtain an angle gather based on the elastic wave equation; Step 2: Obtain a multi-channel inversion objective function based on the angle gather; Step 3: Remove the correlation between elastic parameters from the multi-channel inversion objective function; Step 4: Select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between elastic parameters and the initial model of elastic parameters; Step 5: Repeatedly iterate Step 4 to obtain a multi-channel pre-stack waveform inversion result.

Description

Technical Field

[0001] The present invention relates to the technical field of oil exploration, and particularly relates to a multi-channel pre-stack waveform inversion method and device. Background Art

[0002] In recent years, the degree of oil exploration has been continuously improved. Simple structural oil and gas reservoirs have been almost exhausted, and more and more research directions have begun to shift to complex structural oil and gas reservoirs, subtle oil and gas reservoirs, and lithologic oil and gas reservoirs. However, traditional post-stack inversion no longer meets the needs, and pre-stack inversion has gradually emerged and become mature. Compared with post-stack seismic data, pre-stack seismic data carry rich lithologic information and fluid information of underground media. However, existing pre-stack inversion methods have problems such as large inversion errors and insufficient inversion resolution when facing complex structures and reservoirs with well-developed thin interbeds. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a multi-channel pre-stack waveform inversion method and device, which solve the problems of large inversion errors and insufficient inversion resolution.

[0004] A multi-channel pre-stack waveform inversion method provided by an embodiment of the present invention includes:

[0005] Step 1: Obtain an angle gather based on the elastic wave equation of motion;

[0006] Step 2: Obtain a multi-channel inversion objective function based on the angle gather;

[0007] Step 3: Remove the correlation between elastic parameters from the multi-channel inversion objective function;

[0008] Step 4: Select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between elastic parameters and the initial model of elastic parameters;

[0009] Step 5: Repeatedly iterate Step 4 to obtain a multi-channel pre-stack waveform inversion result.

[0010] In one embodiment, the step of Step 1: Obtain an angle gather based on the elastic wave equation of motion includes:

[0011] Use the vectorized composite matrix method for forward modeling to analytically solve the one-dimensional elastic wave equation of motion. Starting from the bottom-layer reflection response, obtain the comprehensive reflection response layer by layer to obtain the comprehensive response of the reflection coefficient of the layered medium below the surface;

[0012] Add wavelet information to the comprehensive response of the reflection coefficient of the layered medium below the surface to obtain a seismic record; perform flattening processing on the seismic record and perform coordinate domain conversion to obtain an angle gather.

[0013] In one embodiment, the step of obtaining the multi-channel inversion objective function based on the angle gather in step 2 includes:

[0014] Expanding the model vector, data vector, and the corresponding forward matrix based on the angle gather to obtain a multi-channel wavelet-domain reflection coefficient convolution matrix;

[0015] Obtaining the multi-channel inversion objective function based on the two-dimensional prior constraint introduced by the Bayesian framework and the multi-channel wavelet-domain reflection coefficient convolution matrix.

[0016] In one embodiment, the step of removing the correlation between the elastic parameters in step 3 for the multi-channel inversion objective function includes:

[0017] Calculating the covariance matrix of the elastic parameters, performing eigenvector matrix and eigenvalue matrix decomposition on the covariance matrix to obtain the corresponding eigenvector matrix and eigenvalue matrix;

[0018] Decomposing the eigenvector matrix and the eigenvalue matrix, and using a linear matrix to perform variable substitution on the parameters in the objective function to remove the correlation between the elastic parameters.

[0019] In one embodiment, step 4: selecting an initial model of elastic parameters, and obtaining the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between the elastic parameters and the initial model of the elastic parameters includes:

[0020] Selecting the initial model of the elastic parameters, and solving the gradient and Hessian matrix of the objective function under the multi-channel condition;

[0021] Based on the gradient and Hessian matrix of the objective function under the multi-channel condition, using the chain rule to solve the gradient and Hessian matrix under the decorrelation algorithm condition;

[0022] Applying the gradient and Hessian matrix under the decorrelation algorithm condition to the Gauss-Newton method to obtain the perturbation amount of the elastic parameters relative to the initial model

[0023] In one embodiment, after step 5: repeatedly iterating step 4 to obtain the multi-channel pre-stack waveform inversion result, the steps of actual wavelet extraction, well-seismic calibration, horizon picking, initial model establishment, and well-side trace testing are further included to establish a multi-channel pre-stack waveform inversion model.

[0024] A multi-channel pre-stack waveform inversion device includes:

[0025] An angle gather acquisition module for obtaining an angle gather based on the elastic wave equation;

[0026] An objective function acquisition module, configured to obtain a multi - trace inversion objective function based on the angle gather;

[0027] A decorrelation module, configured to decorrelate the elastic parameters in the multi - trace inversion objective function;

[0028] A processing module, configured to select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi - trace inversion objective function after removing the correlation between the elastic parameters and the initial model of the elastic parameters;

[0029] An iteration module, configured to repeatedly iterate step 4 to obtain a multi - trace pre - stack waveform inversion result.

[0030] In one implementation, it further includes: a model establishment module, configured to extract actual wavelets, perform well - seismic calibration, pick horizons, establish an initial model, and test well - side traces to establish a multi - trace pre - stack waveform inversion model.

[0031] An electronic device includes a memory and a processor, where the memory is used to store one or more computer instructions, and when the one or more computer instructions are executed, the multi - trace pre - stack waveform inversion method as described in any one of the above is implemented.

[0032] A computer - readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi - trace pre - stack waveform inversion method as described in any one of the above is implemented.

[0033] Due to the adoption of the above - mentioned multi - trace pre - stack waveform inversion method, the present invention has the following advantages: 1. This technology can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the influence of seismic wave phase changes, formation thickness, and transmission loss, the inversion accuracy is higher than that of traditional pre - stack AVO methods. 2. This technical solution is no longer based on linear approximation and is applicable to any angle - range gather, providing a solid theoretical basis for the inversion of large - angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with traditional regularization methods, this method has higher stability of the solution and accuracy of the result. 4. This technology introduces a decorrelation algorithm into pre - stack waveform inversion, reducing the influence of the statistical correlation of three parameters on inversion, improving the stability of the inversion process, and thus improving the noise resistance of inversion. 5. The inversion process of this technology is carried out through multi - trace inversion, fully considering the cross - correlation relationship between adjacent traces, greatly improving the lateral continuity of the inversion profile, and solving the problem of "hanging noodles" in single - trace inversion. Description of the Drawings

[0034] Figure 1 The figure shows a schematic flow chart of a multi - trace pre - stack waveform inversion method provided by an embodiment of the present invention.

[0035] Figure 2 The following is another schematic flow diagram of a multi-channel pre-stack waveform inversion method provided by an embodiment of the present invention.

[0036] Figure 3 The following is a schematic diagram of the longitudinal wave velocity, transverse wave velocity, and density of a thin interbed model obtained by inverting the original pre-stack waveform at a signal-to-noise ratio of 1 provided by an embodiment of the present invention.

[0037] Figure 4 The following is a schematic diagram of the longitudinal wave velocity, transverse wave velocity, and density obtained by pre-stack waveform inversion after adding a decorrelation algorithm at a signal-to-noise ratio of 1 provided by an embodiment of the present invention.

[0038] Figure 5 The following is a schematic diagram of the longitudinal wave velocity obtained by single-channel pre-stack waveform inversion in a noise-free case provided by an embodiment of the present invention.

[0039] Figure 6 The following is a schematic diagram of the longitudinal wave velocity obtained by multi-channel inversion in a noise-free case provided by an embodiment of the present invention.

[0040] Figure 7 The following is a schematic diagram of the longitudinal wave velocity obtained by inverting the actual work area of single-channel pre-stack waveform provided by an embodiment of the present invention.

[0041] Figure 8 The following is a schematic diagram of the longitudinal wave velocity obtained by inverting the actual work area of multi-channel pre-stack waveform provided by an embodiment of the present invention.

[0042] Figure 9 The following is a schematic flow diagram of a multi-channel pre-stack waveform inversion device provided by an embodiment of the present invention. Detailed implementation manners

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0044] AVO technology is an effective way to utilize prestack information and has become the core technology in oil and gas exploration and development in recent years. Compared with post-stack inversion that yields only a single wave impedance data, AVO inversion can provide multiple elastic parameters such as P-wave velocity, S-wave velocity, density, Poisson's ratio, etc. simultaneously. Furthermore, crossplot interpretation can be carried out to reduce the uncertainty in reservoir prediction. Through the research of the inventors of this application, it is found that traditional AVO inversion technology usually performs inversion based on the exact Zoeppritz equation or some of its approximate equations. However, the Zoeppritz equation only considers the petrophysical properties on both sides of a single interface and does not take into account the influence of formation thickness on it. Therefore, it is difficult to apply to the prediction of thin interbedded hydrocarbon reservoirs. At the same time, the approximate formula of the Zoeppritz equation is also valid under the premise assumptions of small-angle (small offset) incidence and slow change of elastic parameters on both sides of the interface. Conventional inversion based on the exact Zoeppritz equation and its approximate equations assumes that the received seismic data only contains primary reflection waves, ignoring the influence of wave propagation effects, such as multiple waves, various types of converted waves, and transmission losses, etc. At the same time, due to the prestack inversion based on the approximate formula of the Zoeppritz equation, it requires a relatively small incident angle (about 30 degrees) and slow change of elastic parameters on both sides of the interface. When the elastic parameters on both sides of the interface change violently or the seismic data offset is relatively large, the AVO inversion error based on the Zoeppritz approximate formula will be relatively large. In the face of complex structures and reservoirs with well-developed thin interbeds, the resolution of traditional prestack AVO inversion is insufficient.

[0045] Theoretically, the prestack waveform inversion method based on the elastic wave equation is the most suitable method for predicting the elastic parameters of thin interbedded reservoirs. This method can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmission waves. Due to its consideration of the influence of seismic wave phase changes, formation thickness, and transmission losses, the inversion accuracy is higher than that of traditional prestack AVO methods. However, the inventors of this application have found that the prestack waveform inversion method still has deficiencies. Constrained by the ill-posedness of the inversion problem, prior constraints are often introduced in the Bayesian basic framework during the inversion process to improve the stability of the inversion. Since the prior constraints ignore the statistical correlation between elastic parameters, and this correlation will affect the inversion results, causing the results to deviate from the constraints of the seismic records, especially when the noise in the seismic records is relatively strong.

[0046] To solve the above problems, the present invention introduces a decorrelation algorithm into the prestack waveform inversion algorithm to remove the statistical correlation between elastic parameters, making the elastic parameters independent and identically distributed, and better satisfying the assumption conditions of prior constraints. The inversion result will be more stable, and the algorithm itself has stronger anti-noise ability. Secondly, when estimating elastic parameters by prestack waveform inversion, it is carried out channel by channel, that is, the seismic record of the current channel and the initial model of elastic parameters are input, the elastic parameters of this channel are inversely obtained, and then the inversion of the next channel is carried out. However, single-channel inversion often considers the relationship between adjacent channels, and the input data is relatively independent. Therefore, the lateral continuity of the obtained inversion profile is insufficient, and there are phenomena such as "hanging noodles", which is not conducive to subsequent structural interpretation and reservoir prediction. Multi-channel inversion can effectively solve this problem. The present invention expands the parameter matrix, adds two-dimensional prior constraints, fully considers the lateral relationship between channels, and improves the lateral continuity of the inversion result. Due to considering the data connection between channels, the stability and anti-noise performance of the inversion algorithm can also be indirectly improved. The specific implementation manner is as described in the following embodiments.

[0047] Embodiment 1:

[0048] Figure 1 The figure shows a schematic flow chart of a multi-channel prestack waveform inversion method provided by an embodiment of the present invention.

[0049] Figure 2 The figure shows another schematic flow chart of a multi-channel prestack waveform inversion method provided by an embodiment of the present invention.

[0050] Figure 3 The figure shows a schematic diagram of the P-wave velocity, S-wave velocity, and density of a thin interbed model obtained by prestack waveform inversion based on the original data at a signal-to-noise ratio of 1, provided by an embodiment of the present invention.

[0051] Figure 4 The figure shows a schematic diagram of the P-wave velocity, S-wave velocity, and density obtained by prestack waveform inversion after adding a decorrelation algorithm at a signal-to-noise ratio of 1, provided by an embodiment of the present invention.

[0052] Figure 5 The figure shows a schematic diagram of the P-wave velocity obtained by single-channel prestack waveform inversion under noiseless conditions, provided by an embodiment of the present invention.

[0053] Figure 6 The figure shows a schematic diagram of the P-wave velocity obtained by multi-channel inversion under noiseless conditions, provided by an embodiment of the present invention.

[0054] Figure 7 The figure shows a schematic diagram of the P-wave velocity obtained by single-channel prestack waveform inversion in an actual work area, provided by an embodiment of the present invention.

[0055] Figure 8The figure shows a schematic diagram of the P-wave velocity obtained by multi-channel pre-stack waveform inversion in an actual work area provided by an embodiment of the present invention.

[0056] An embodiment provides a multi-channel pre-stack waveform inversion method. Refer to Figures 1 to 2 , the multi-channel pre-stack waveform inversion method includes:

[0057] Step 1: Obtain an angle gather based on the elastic wave equation. Forward simulation is the basis of inversion. The vectorized composite matrix method can be used for the analytical solution of the one-dimensional elastic wave equation by forward simulation. Starting from the bottom-layer reflection response, the comprehensive reflection response is obtained layer by layer to obtain the comprehensive response of the reflection coefficient of the layered medium below the surface; add wavelet information to the comprehensive response of the reflection coefficient of the layered medium below the surface to obtain a seismic record; flatten the seismic record and perform coordinate domain conversion to obtain an angle gather.

[0058] Step 2: Obtain a multi-channel inversion objective function based on the angle gather; expand the model vector, data vector, and the corresponding forward matrix based on the angle gather, and further derive the matrix form of the multi-channel wavelet domain reflection coefficient convolution to obtain the multi-channel wavelet domain reflection coefficient convolution matrix, thereby expanding the wavelet matrix and the reflection coefficient matrix according to the corresponding matrix arrangement. On this basis, introduce two-dimensional prior constraints based on the Bayesian framework to ensure the stability of the solution, convert the inversion problem into a posterior estimation problem, and finally establish a multi-channel inversion objective function.

[0059] Step 3: Remove the correlation between elastic parameters in the multi-channel inversion objective function. Among them, the method for removing the correlation between elastic parameters in the multi-channel inversion objective function includes: calculating the covariance matrix of the elastic parameters, decomposing the covariance matrix into a eigenvector matrix and an eigenvalue matrix to obtain the corresponding eigenvector matrix and eigenvalue matrix; decomposing the eigenvector matrix and the eigenvalue matrix to obtain two linear matrices, and using the linear matrices to perform variable substitution on the parameters in the objective function to remove the correlation between the elastic parameters, ensure that the elastic parameters are independently and identically distributed, and guarantee the accuracy and stability of the inversion result. Use the two-step linear transformation method to remove the correlation between elastic parameters, ensure that the elastic parameters are independently and identically distributed, and further guarantee the accuracy and stability of the inversion result.

[0060] Step 4: Select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between the elastic parameters and the initial model of the elastic parameters. Select an appropriate initial model of elastic parameters and perform a second-order Taylor expansion approximation on the objective function. First, select the initial model of the elastic parameters and solve the gradient and Hessian matrix of the objective function under the multi-channel condition; then, use the chain rule of differentiation to solve the gradient and Hessian matrix of the objective function under the condition of the decorrelation algorithm; finally, apply the finally obtained gradient and Hessian matrix to the Gauss-Newton method to solve the perturbation amount of the elastic parameters relative to the initial model.

[0061] Step 5: Iterate step 4 repeatedly to obtain the multi-channel pre-stack waveform inversion result.

[0062] Among them, after completing the step of repeatedly iterating step 4 to obtain the multi-channel pre-stack waveform inversion result, steps such as actual wavelet extraction, well-seismic calibration, horizon picking, initial model establishment, and well-side trace testing are also included to establish a multi-channel pre-stack waveform inversion model. After completing the model test, starting from actual wavelet extraction, and then performing well-seismic calibration, horizon picking, initial model establishment, well-side trace testing, etc., so as to promote the robust multi-channel pre-stack waveform inversion method to actual work area testing.

[0063] Figure 3 The following shows the longitudinal wave velocity, transverse wave velocity, and density schematic diagrams of a thin interbed model obtained by pre-stack waveform inversion based on the original data at a signal-to-noise ratio of 1 provided by an embodiment of the present invention. Figure 4 The following shows the longitudinal wave velocity, transverse wave velocity, and density schematic diagrams of pre-stack waveform inversion after adding a decorrelation algorithm at a signal-to-noise ratio of 1 provided by an embodiment of the present invention. Figure 1 and Figure 4 The curves in are the actual logging curves, inversion results, and low-frequency models respectively.

[0064] Figure 5 The following shows the longitudinal wave velocity schematic diagram obtained by single-channel pre-stack waveform inversion in a noise-free case provided by an embodiment of the present invention. Figure 6 The following shows the longitudinal wave velocity schematic diagram obtained by multi-channel inversion in a noise-free case provided by an embodiment of the present invention.

[0065] Figure 7 The following shows the longitudinal wave velocity schematic diagram obtained by single-channel pre-stack waveform inversion in an actual work area provided by an embodiment of the present invention. Figure 8 The following shows the longitudinal wave velocity schematic diagram obtained by multi-channel pre-stack waveform inversion in an actual work area provided by an embodiment of the present invention. Figure 7 and Figure 8 The black curve in is the logging data.

[0066] ThroughFigures 3 to 4 , Figures 5 to 6 , Figures 7 to 8 By comparing Figures 3 to 4 , Figures 5 to 6 , and Figures 7 to 8 , it can be found that due to the above-mentioned multi-channel pre-stack waveform inversion method, the present invention has the following advantages: 1. This technology can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the effects of seismic wave phase changes, formation thickness, and transmission losses, the inversion accuracy is higher than that of the traditional pre-stack AVO method. 2. This technical solution is no longer based on linear approximation and is applicable to any angle range gather, providing a solid theoretical basis for the inversion of large-angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with the traditional regularization method, this method has higher stability of the solution and accuracy of the results. 4. This technology introduces a decorrelation algorithm into the pre-stack waveform inversion, reducing the influence of the statistical correlation of the three parameters on the inversion and improving the stability of the inversion process, thereby improving the noise resistance of the inversion. 5. The inversion process of this technology is carried out through multi-channel inversion, fully considering the cross-correlation relationship between adjacent channels, greatly improving the lateral continuity of the inversion profile and solving the problem of "hanging noodles" that appears in single-channel inversion.

[0067] Embodiment 2:

[0068] Since the forward model is the basis for inversion implementation, the present invention first assumes a layered medium and analytically solves the one-dimensional elastic wave equation using the composite matrix method.

[0069] According to the forward theory of the vectorized composite matrix method, the propagation vector is first defined:

[0070] ν n = [Δ -R PS Δ -R SS Δ R PP Δ R SP Δ |RΔ] T (1)

[0071] Its physical meaning represents the propagation response from the nth layer to the bottommost Nth layer. Among them, Δ is the determinant of the system matrix, and its value does not affect the final calculation result. R PS , R SS , R PP , R SP represent the reflection coefficients of PS waves, SS waves, PP waves, and SP waves respectively. The determinant has no physical meaning.

[0072] To calculate the total reflection response, it is necessary to calculate the final total propagation vector v0 from the Nth layer to the surface from bottom to top. For convenient calculation, the algorithm defines the layer propagation matrix Q n :

[0073] ν n= Q n ν n+1

[0074]

[0075] Matrix E n The travel-time phase matrix describes the travel time during the propagation of seismic waves, The amplitude energy distribution matrix describes the energy distribution process of seismic waves at the interface. The above matrices can all be represented by the elastic parameters, P-wave velocity, S-wave velocity, and density.

[0076] Since the overall response of the bottom interface is known, this technique starts from the response ν N = [1 0 0 0 0 0] T and uses formula (2) to solve the reflection coefficient layer by layer from bottom to top. Therefore, the total response ν0 below the top interface is:

[0077] ν0 = Q1Q2…Q N ν N (3)

[0078] The total reflection coefficient of the PP wave in the frequency-wavenumber (τ-p) domain can be calculated from ν0:

[0079]

[0080] Multiply the wavelet by R(ω,p) and perform frequency-domain integration to obtain the seismic record in the τ-p domain:

[0081]

[0082] Furthermore, by performing dynamic correction in the τ-p domain to flatten the seismic record and using formula for the conversion between ray parameter and angle, the final common-angle gather is obtained.

[0083]

[0084] According to the above forward modeling process, it can be mathematized as:

[0085] d = G(m) (7)

[0086] where d is the angle gather; m is the elastic parameter; G is the non-linear operator that maps m to d. In the present invention, G(m) represents the result of the analytical solution forward modeling. According to the convolution relationship, formula (7) is further modified as:

[0087] d = G(m) = W * R(m) (8)

[0088] R(m) represents the reflection coefficient, and W is the seismic wavelet; generalizing it to multiple channels, it can be expressed as:

[0089]

[0090] That is:

[0091] In the formula, is the observed multi-channel seismic data, represents the synthesized corresponding reflection coefficient, represents the multi-channel elastic parameters, and the wavelet W between channels often remains unchanged. l represents the number of multi-channel inversion channels, and d i＝1,2...l represents the synthesized data of the i-th channel, and R i＝1,2…l represents the reflection coefficient of the i-th channel.

[0092] Therefore, the objective function of multi-channel inversion constructed based on prestack waveform inversion can be:

[0093]

[0094] In the formula, represents the two-dimensional regularization term, and λ is the prior information weight.

[0095] Further, a decorrelation operation needs to be performed. The covariance matrix of the three-parameter matrix m can be expressed as:

[0096]

[0097] Among them, x represents the P-wave velocity, y represents the S-wave velocity, and z represents the density. The covariance matrix is decomposed into LU matrices. If U and L represent the eigenvector matrix and eigenvalue matrix of the covariance matrix C respectively, then C can be expressed in the following form:

[0098] C = ULU T (13)

[0099] Let Then the decorrelated three-parameter matrix m' is:

[0100] m' = S -1 U -1 m (14)

[0101] Denote S -1 U -1 = S u , SU = U s After decorrelation, the objective function of inversion can be expressed in the following form:

[0102]

[0103] Among them, U (i) s represents the elastic parameter decorrelation operator of the i-th channel, Similarly expressed.

[0104] The above objective function needs to be solved using the Gauss - Newton method:

[0105]

[0106]

[0107] According to the chain - rule, the gradient matrix:

[0108]

[0109] The Hessian matrix:

[0110]

[0111] Where:

[0112]

[0113]

[0114] According to The partial derivatives in the multi - path form and can be expressed as:

[0115]

[0116]

[0117] and correspond to the reflection coefficient in the \(i\) - th path case and the derivative of the regularization term with respect to the elastic parameters respectively.

[0118] Finally, the entire inversion algorithm can be completed through repeated iteration. After completing the model data test, the algorithm can be further extended to actual data.

[0119] Due to the above multi-channel pre-stack waveform inversion method, the present invention has the following advantages: 1. This technology can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the effects of seismic wave phase changes, formation thickness, and transmission losses, the inversion accuracy is higher than that of traditional pre-stack AVO methods. 2. This technical solution is no longer based on linear approximation and is applicable to any angle range gather, providing a solid theoretical basis for the inversion of large-angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with traditional regularization methods, this method has higher stability of the solution and accuracy of the results. 4. This technology introduces a decorrelation algorithm into pre-stack waveform inversion, reducing the influence of the statistical correlation of three parameters on inversion, improving the stability of the inversion process, and thus enhancing the noise resistance of inversion. 5. The inversion process of this technology is carried out through multi-channel inversion, fully considering the cross-correlation between adjacent channels, greatly improving the lateral continuity of the inversion profile, and solving the problem of "hanging noodles" in single-channel inversion.

[0120] Embodiment 3:

[0121] Figure 9 The following is a schematic flow chart of a multi-channel pre-stack waveform inversion device provided by an embodiment of the present invention.

[0122] This embodiment provides a multi-channel pre-stack waveform inversion device 100. Refer to Figure 9 , the multi-channel pre-stack waveform inversion device 100 includes an angle gather acquisition module 10, a target function acquisition module 20, a decorrelation module 30, a processing module 40, and an iteration module 50. Among them:

[0123] The angle gather acquisition module 10 is used to obtain an angle gather based on the elastic wave equation.

[0124] The target function acquisition module 20 is used to obtain a multi-channel inversion target function based on the angle gather.

[0125] The decorrelation module 30 is used to remove the correlation between elastic parameters from the multi-channel inversion target function.

[0126] The processing module 40 is used to select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion target function after removing the correlation between elastic parameters and the initial model of elastic parameters.

[0127] The iteration module 50 is used to repeatedly iterate step 4 in the embodiment to obtain a multi-channel pre-stack waveform inversion result.

[0128] The angle gather acquisition module 10 obtains the angle gather based on the elastic wave equation, and the objective function acquisition module 20 obtains the multi-channel inversion objective function based on the angle gather; the decorrelation module 30 removes the correlation between elastic parameters from the multi-channel inversion objective function; the processing module 40 selects the initial model of elastic parameters, and obtains the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between elastic parameters and the initial model of elastic parameters; the iteration module 50 iterates step 4 repeatedly to obtain the multi-channel pre-stack waveform inversion result.

[0129] The angle gather acquisition module 10 is further configured to analytically solve the one-dimensional elastic wave equation by using the vectorized composite matrix method for forward modeling, starting from the bottom-layer reflection response, and obtain the comprehensive reflection response layer by layer to obtain the comprehensive response of the reflection coefficient of the layered medium below the surface; add the wavelet information to the comprehensive response of the reflection coefficient of the layered medium below the surface to obtain the seismic record; perform flattening processing on the seismic record and perform coordinate domain conversion to obtain the angle gather.

[0130] The objective function acquisition module 20 is further configured to expand the model vector, the data vector, and the corresponding forward modeling matrix based on the angle gather, and further derive the matrix form of the multi-channel wavelet domain reflection coefficient convolution to obtain the multi-channel wavelet domain reflection coefficient convolution matrix, thereby expanding the wavelet matrix and the reflection coefficient matrix according to the corresponding matrix arrangement. On this basis, introduce two-dimensional prior constraints based on the Bayesian framework to ensure the stability of the solution, convert the inversion problem into a posterior estimation problem, and finally establish the multi-channel inversion objective function.

[0131] The method for the decorrelation module 30 to remove the correlation between elastic parameters from the multi-channel inversion objective function further includes: calculating the covariance matrix of the elastic parameters, performing eigenvector matrix and eigenvalue matrix decomposition on the covariance matrix to obtain the corresponding eigenvector matrix and eigenvalue matrix; decomposing the eigenvector matrix and the eigenvalue matrix to obtain two linear matrices, and using the linear matrices to perform variable substitution on the parameters in the objective function to remove the correlation between the elastic parameters, ensure that the elastic parameters are independently and identically distributed, and guarantee the accuracy and stability of the inversion result. Use the two-step linear transformation method to remove the correlation between elastic parameters, ensure that the elastic parameters are independently and identically distributed, and further guarantee the accuracy and stability of the inversion result.

[0132] The processing module 40 is further configured to select an appropriate initial model of elastic parameters and perform a second-order Taylor expansion approximation on the objective function. First, select the initial model of the elastic parameters, and solve the gradient and Hessian matrix of the objective function under the multi-channel condition; then, use the chain rule of differentiation to solve the gradient and Hessian matrix of the objective function under the condition of the decorrelation algorithm; finally, apply the finally obtained gradient and Hessian matrix to the Gauss-Newton method to solve the perturbation amount of the elastic parameters relative to the initial model.

[0133] In addition, the multi-channel pre-stack waveform inversion device 100 further includes a model establishment module 50, which is configured to extract actual wavelets, perform well-seismic calibration, pick horizons, establish an initial model, and test well-side traces to establish a multi-channel pre-stack waveform inversion model.

[0134] Due to the adoption of the above multi-channel pre-stack waveform inversion device, the present invention has the following advantages: 1. This technology can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the influence of seismic wave phase changes, formation thickness, and transmission losses, the inversion accuracy is higher than that of traditional pre-stack AVO methods. 2. This technical solution is no longer based on linear approximation and is applicable to any angle range gather, providing a solid theoretical basis for the inversion of large-angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with traditional regularization methods, this method has higher stability of the solution and accuracy of the results. 4. This technology introduces the decorrelation algorithm into pre-stack waveform inversion, reduces the influence of the statistical correlation of the three parameters on the inversion, improves the stability of the inversion process, and further improves the noise resistance of the inversion. 5. The inversion process of this technology is carried out through multi-channel inversion, fully considering the cross-correlation relationship between adjacent traces, greatly improving the lateral continuity of the inversion profile, and solving the problem of "hanging noodles" in single-channel inversion.

[0135] Embodiment 4:

[0136] This embodiment provides an electronic device, which can be a mobile phone, a computer, a tablet computer, etc., including a memory and a processor. A computer program is stored on the memory, and when the computer program is executed by the processor, it implements the multi-channel pre-stack waveform inversion method described in Embodiment 1 and Embodiment 2. It can be understood that the electronic device may further include an input / output (I / O) interface and a communication component.

[0137] Among them, the processor is configured to execute all or part of the steps in the multi-channel pre-stack waveform inversion method described in Embodiment 1 and Embodiment 2. The memory is used to store various types of data, which may include, for example, instructions of any application program or method in the electronic device, and data related to the application program.

[0138] The processor may be implemented by an application specific integrated circuit (ASIC), a digital signal processor (DSP), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor, or other electronic components, and is used to execute the multi-channel pre-stack waveform inversion method in the first and second embodiments above.

[0139] The memory may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk, or an optical disk.

[0140] Due to the adoption of the above multi-channel pre-stack waveform inversion method, the present invention has the following advantages: 1. This technology can not only simulate the primary reflection wave, but also simulate the multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the influence of seismic wave phase changes, formation thickness, and transmission loss, the inversion accuracy is higher than that of the traditional pre-stack AVO method. 2. This technical solution is no longer based on linear approximation and is applicable to any angle range gather, providing a solid theoretical basis for the inversion of large-angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with the traditional regularization method, this method has higher stability of the solution and accuracy of the result. 4. This technology introduces the decorrelation algorithm into the pre-stack waveform inversion, reduces the influence of the statistical correlation of the three parameters on the inversion, improves the stability of the inversion process, and further improves the noise resistance of the inversion. 5. The inversion process of this technology is carried out through multi-channel inversion, fully considering the cross-correlation relationship between adjacent channels, greatly improving the lateral continuity of the inversion profile, and solving the problem of "hanging noodles" in single-channel inversion.

[0141] Embodiment 5

[0142] This embodiment also provides a computer-readable storage medium. In each embodiment of the present invention, each functional unit may be integrated into a processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium.

[0143] Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0144] The aforementioned storage medium includes: flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, server, APP application mall, and other various media that can store program check codes. A computer program is stored thereon, and when the computer program is executed by a processor, the following method steps can be implemented:

[0145] Step 1: Obtain an angle gather based on the elastic wave equation. Forward modeling is the basis for inversion. The vectorized composite matrix method can be used for forward modeling to analytically solve the one-dimensional elastic wave equation. Starting from the bottom-layer reflection response, the comprehensive reflection response is obtained layer by layer to obtain the comprehensive response of the reflection coefficient of the layered medium below the surface. Add wavelet information to the comprehensive response of the reflection coefficient of the layered medium below the surface to obtain a seismic record. Flatten the seismic record and perform coordinate domain conversion to obtain an angle gather.

[0146] Step 2: Obtain a multi-channel inversion objective function based on the angle gather; expand the model vector, data vector, and the corresponding forward matrix based on the angle gather, and further derive the matrix form of the multi-channel wavelet domain reflection coefficient convolution to obtain a multi-channel wavelet domain reflection coefficient convolution matrix. Accordingly, the wavelet matrix and the reflection coefficient matrix are expanded according to the corresponding matrix arrangement. On this basis, introduce two-dimensional prior constraints based on the Bayesian framework to ensure the stability of the solution, convert the inversion problem into a posteriori estimation problem, and finally establish a multi-channel inversion objective function.

[0147] Step 3: Remove the correlation between elastic parameters from the multi-channel inversion objective function. The method for removing the correlation between elastic parameters from the multi-channel inversion objective function includes: calculating the covariance matrix of the elastic parameters, performing eigenvector matrix and eigenvalue matrix decomposition on the covariance matrix to obtain the corresponding eigenvector matrix and eigenvalue matrix; decomposing the eigenvector matrix and the eigenvalue matrix to obtain two linear matrices, and using the linear matrices to perform variable substitution on the parameters in the objective function to remove the correlation between the elastic parameters, ensuring that the elastic parameters are independently and identically distributed, and guaranteeing the accuracy and stability of the inversion result. Use the two-step linear transformation method to remove the correlation between elastic parameters, ensure that the elastic parameters are independently and identically distributed, and thus guarantee the accuracy and stability of the inversion result.

[0148] Step 4: Select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between elastic parameters and the initial model of elastic parameters; select an appropriate initial model of elastic parameters and perform a second-order Taylor expansion approximation on the objective function. First, select the initial model of elastic parameters and solve the gradient and Hessian matrix of the objective function under the multi-channel conditions; then, use the chain rule to solve the gradient and Hessian matrix of the objective function under the conditions of the decorrelation algorithm; finally, apply the finally obtained gradient and Hessian matrix to the Gauss-Newton method to solve the perturbation amount of the elastic parameters relative to the initial model.

[0149] Step 5: Iterate step 4 repeatedly to obtain the multi-channel pre-stack waveform inversion result.

[0150] Among them, after completing the step of iterating step 4 repeatedly to obtain the multi-channel pre-stack waveform inversion result, there are also steps of actual wavelet extraction, well-seismic calibration, horizon picking, initial model establishment, and well-side trace testing to establish a multi-channel pre-stack waveform inversion model. After completing the model test, starting from actual wavelet extraction, and then well-seismic calibration, horizon picking, initial model establishment, well-side trace testing, etc., so as to promote the robust multi-channel pre-stack waveform inversion method to actual work area testing.

[0151] When the computer program is executed by a processor, the following method steps can also be implemented:

[0152] Since the forward model is the basis for inversion implementation, the present invention first assumes a layered medium and analytically solves the one-dimensional elastic wave equation using the conforming matrix method.

[0153] According to the forward theory of the vectorized composite matrix method, first define the propagation vector:

[0154] ν n =[Δ -R PS Δ -R SSΔR PP ΔR SP Δ|RΔ] T (1)

[0155] Its physical meaning represents the propagation response from the nth layer to the bottommost Nth layer. Where Δ is the determinant of the system matrix, and its value does not affect the final calculation result. R PS 、R SS 、R PP 、R SP represent the reflection coefficients of PS wave, SS wave, PP wave and SP wave respectively. The determinant has no physical meaning.

[0156] To calculate the total reflection response, it is necessary to calculate the final total propagation vector v0 from the Nth layer to the surface from bottom to top. For the convenience of calculation, the algorithm defines the layer propagation matrix Q n :

[0157] ν n =Q n ν n+1

[0158]

[0159] The matrix E n is the time-shift phase matrix that describes the travel time in the process of seismic wave propagation, and the amplitude energy distribution matrix describes the energy distribution process of seismic waves at the interface. The above matrices can all be expressed by the elastic parameters, P-wave velocity, S-wave velocity and density.

[0160] Since the overall response of the bottom interface is known, this technology starts from the response ν N =[1 0 0 0 0 0] T and uses formula (2) to solve the reflection coefficient layer by layer from bottom to top. Therefore, the total response ν0 below the top interface is:

[0161] ν0 = Q1Q2…Q N ν N (3)

[0162] The total reflection coefficient of the PP wave in the frequency-wavenumber (τ-p) domain can be calculated from ν0:

[0163]

[0164] Multiply the wavelet by R(ω,p) and perform frequency-domain integration to obtain the seismic record in the τ-p domain:

[0165]

[0166] Furthermore, through dynamic correction in the τ-p domain, the seismic record is flattened and formula Convert ray parameters and angles to obtain the final common-angle gather.

[0167]

[0168] According to the above forward modeling process, it can be mathematically expressed as:

[0169] d = G(m) (7)

[0170] Where, d is the angle gather; m is the elastic parameter; G is the non-linear operator that maps m to d. In the present invention, G(m) represents the result of forward modeling of the analytical solution. According to the convolution relationship, formula (7) is further modified to:

[0171] d = G(m) = W * R(m) (8)

[0172] R(m) represents the reflection coefficient, and W is the seismic wavelet; generalizing it to multiple channels, it can be expressed as:

[0173]

[0174] That is:

[0175] In the formula, is the observed multi-channel seismic data, represents the synthetic corresponding reflection coefficient, represents the multi-channel elastic parameter, and the wavelet W is often unchanged between channels. l represents the number of channels for multi-channel inversion, d i＝1,2…l represents the synthetic data of the i-th channel, R i＝1,2…l represents the reflection coefficient of the i-th channel.

[0176] Therefore, the objective function of multi-channel inversion based on prestack waveform inversion can be:

[0177]

[0178] In the formula, represents the two-dimensional regularization term, and λ is the weight of the prior information.

[0179] Further, a decorrelation operation needs to be performed. The covariance matrix of the three-parameter matrix m can be expressed as:

[0180]

[0181] Where, x represents the P-wave velocity, y represents the S-wave velocity, and z represents the density. Perform LU matrix decomposition on the covariance matrix. If U and L represent the eigenvector matrix and eigenvalue matrix of the covariance matrix C respectively, then C can be expressed in the following form:

[0182] C = ULU T (13)

[0183] Let Then the three-parameter matrix m' after decorrelation is:

[0184] m' = S -1 U -1 m(14)

[0185] Denote S -1 U -1 = S u and SU = U s After decorrelation, the objective function for inversion can be expressed in the following form:

[0186]

[0187] where, U (i) s represents the elastic parameter decorrelation operator for the i-th trace, Similarly expressed.

[0188] The above objective function needs to be solved using the Gauss-Newton method:

[0189]

[0190]

[0191] According to the chain rule, the gradient matrix:

[0192]

[0193] The Hessian matrix:

[0194]

[0195] where:

[0196]

[0197]

[0198] According to the partial derivatives in the multi-trace form and can be expressed as:

[0199]

[0200]

[0201] and correspond to the reflection coefficient and the derivative of the regularization term with respect to the elastic parameters in the case of the i-th trace, respectively.

[0202] Finally, the entire inversion algorithm can be completed through repeated iterations. After completing the model data test, the algorithm can be further extended to actual data.

[0203] Due to the above multi-channel pre-stack waveform inversion method, the present invention has the following advantages: 1. This technology can not only simulate primary reflection waves, but also simulate multiple waves, multiple converted waves, and transmitted waves. Since it takes into account the influence of seismic wave phase changes, formation thickness, and transmission loss, the inversion accuracy is higher than that of the traditional pre-stack AVO method. 2. This technical solution is no longer based on linear approximation and is applicable to any angle range gather, providing a solid theoretical basis for the inversion of large-angle seismic data. 3. This technology introduces prior constraints based on the Bayesian inference framework. Compared with the traditional regularization method, this method has higher stability of the solution and accuracy of the result. 4. This technology introduces the decorrelation algorithm into the pre-stack waveform inversion, reducing the influence of the statistical correlation of the three parameters on the inversion, improving the stability of the inversion process, and thus enhancing the noise resistance of the inversion. 5. The inversion process of this technology is carried out through multi-channel inversion, fully considering the cross-correlation relationship between adjacent channels, greatly improving the lateral continuity of the inversion profile, and solving the problem of "hanging noodles" that appears in single-channel inversion.

[0204] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0205] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms. The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0206] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are only examples and not limitations. It cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of the present application. Additionally, the above-disclosed specific details are only for the purposes of illustration and easy understanding, rather than limitations. These details do not limit the present application to necessarily implement using the above specific details.

[0207] The block diagrams of the devices, apparatuses, equipment, and systems related to the present application are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any way.

[0208] It should also be noted that in the devices, equipment, and methods of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present application.

[0209] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

[0210] In the description of this application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically defined. In the embodiments of this application, all directional indications (such as up, down, left, right, front, back, top, bottom...) are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will change accordingly. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include steps or units not listed, or may optionally further include other steps or units inherent to these processes, methods, products or devices.

[0211] In addition, the mention of "embodiment" in this document means that the specific features, structures or characteristics described in connection with the embodiment may be included in at least one embodiment of this application. The occurrence of this phrase at various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0212] As described above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims. The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

[0213] The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-channel pre-stack waveform inversion method, characterized in that, Including: Step 1: Obtain an angle gather based on the elastic wave equation; Step 2: Obtain a multi-channel inversion objective function based on the angle gather; The step of obtaining a multi-channel inversion objective function based on the angle gather in Step 2 includes: expanding the model vector, data vector, and the corresponding forward matrix based on the angle gather to obtain a multi-channel wavelet domain reflection coefficient convolution matrix; Obtain the multi-channel inversion objective function based on the two-dimensional prior constraint introduced by the Bayesian framework and the multi-channel wavelet domain reflection coefficient convolution matrix; Step 3: Remove the correlation between the elastic parameters in the multi-channel inversion objective function so that the elastic parameters are independently and identically distributed; Step 4: Select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between the elastic parameters and the initial model of the elastic parameters; Step 5: Repeatedly iterate Step 4 to obtain a multi-channel prestack waveform inversion result.

2. The multi-channel pre-stack waveform inversion method according to claim 1, wherein The step of obtaining an angle gather based on the elastic wave equation in Step 1 includes: Use the vectorized composite matrix method for forward modeling to analytically solve the one-dimensional elastic wave equation, starting from the bottom layer reflection response, and obtain the comprehensive reflection response layer by layer to obtain the comprehensive response of the reflection coefficient of the layered medium below the surface; Add wavelet information to the comprehensive response of the reflection coefficient of the layered medium below the surface to obtain a seismic record; Flatten the seismic record and perform coordinate domain conversion to obtain an angle gather.

3. The multi-channel pre-stack waveform inversion method according to claim 1, wherein The step of removing the correlation between the elastic parameters in the multi-channel inversion objective function in Step 3 includes: calculating the covariance matrix of the elastic parameters, decomposing the covariance matrix into an eigenvector matrix and an eigenvalue matrix to obtain the corresponding eigenvector matrix and eigenvalue matrix; Decompose the eigenvector matrix and the eigenvalue matrix to obtain a linear matrix, and use the linear matrix to perform variable substitution on the parameters in the objective function to remove the correlation between the elastic parameters.

4. The multi-channel pre-stack waveform inversion method according to claim 1, characterized in that The step of obtaining the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion objective function after removing the correlation between the elastic parameters and the initial model of the elastic parameters in Step 4 includes: Select the initial model of the elastic parameters and solve the gradient and Hessian matrix of the objective function under the multi-channel condition; Based on the gradient and Hessian matrix of the objective function under the multi-channel condition, use the chain rule to solve the gradient and Hessian matrix under the decorrelation algorithm condition; Apply the gradient and Hessian matrix under the decorrelation algorithm condition to the Gauss-Newton method to obtain the perturbation amount of the elastic parameters relative to the initial model.

5. The multi-channel pre-stack waveform inversion method according to claim 1, characterized in that After Step 5: Repeatedly iterate Step 4 to obtain a multi-channel prestack waveform inversion result, there are also steps of actual wavelet extraction, well-seismic calibration, horizon picking, initial model establishment, and well-side trace testing to establish a multi-channel prestack waveform inversion model.

6. A multi-channel pre-stack waveform inversion device, characterized in that, Including: An angle gather acquisition module for obtaining an angle gather based on the elastic wave equation; A target function acquisition module, configured to obtain a multi-channel inversion target function based on the angle gather; the target function acquisition module is further configured to expand the model vector, data vector, and the corresponding forward modeling matrix based on the angle gather, and further derive the matrix form of the multi-channel wavelet domain reflection coefficient convolution to obtain a multi-channel wavelet domain reflection coefficient convolution matrix, thereby expanding the wavelet matrix and the reflection coefficient matrix according to the corresponding matrix arrangement method; on this basis, a two-dimensional prior constraint is introduced based on the Bayesian framework to ensure the stability of the solution, the inversion problem is converted into a posterior estimation problem, and finally a multi-channel inversion target function is established. A decorrelation module, configured to decorrelate the elastic parameters in the multi-channel inversion target function. A processing module, configured to select an initial model of elastic parameters, and obtain the perturbation amount of the elastic parameters relative to the initial model based on the multi-channel inversion target function after removing the correlation between the elastic parameters and the initial model of the elastic parameters. An iteration module, configured to perform repeated iterations to obtain a multi-channel pre-stack waveform inversion result.

7. The multi-channel pre-stack waveform inversion device according to claim 6, characterized in that It further includes: a model establishment module, configured to extract actual wavelets, perform well-seismic calibration, pick horizons, establish an initial model, and test well-side traces to establish a multi-channel pre-stack waveform inversion model.

8. An electronic device, characterized in that, It includes a memory and a processor, where the memory is used to store one or more computer instructions, and when the one or more computer instructions are executed, the multi-channel pre-stack waveform inversion method described in any one of the above claims 1-5 is implemented.

9. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the multi-channel pre-stack waveform inversion method described in any one of the above claims 1-5 is implemented.

Images