An elastic impedance tensor-based HTI medium sensitivity equalization seismic inversion method

Abstract

The present application relates to the field of oil and gas exploration and development, and particularly relates to a HTI medium sensitivity balancing seismic inversion method based on elastic impedance tensor. The method comprises the following steps: step S1, constructing a HTI medium accurate reflection coefficient equation based on elastic impedance tensor; step S2, constructing a HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor directly characterized by P-wave velocity, S-wave velocity, density and anisotropy parameters; step S3, constructing a target functional to which the to-be-inverted model parameters are subjected; step S4, introducing a sensitivity balancing parameter adaptive Split-Bregman algorithm to solve the target functional, and obtaining P-wave velocity, S-wave velocity, density and anisotropy parameters. The present application introduces a parameter adaptive strategy guided by sensitivity information, reduces the update difference and coupling influence between different parameters, and improves the update ability of weak sensitivity anisotropy parameters and the stability and precision of multi-parameter inversion.

Description

Technical Field

[0001] This invention relates to the field of oil and gas exploration and development, and specifically to a seismic inversion method based on HTI medium sensitivity equilibrium using elastic impedance tensor. Background Technology

[0002] Fractures, serving as crucial reservoir spaces and migration pathways for oil and gas, are key geological factors inducing rock anisotropy. Widely present in major oil and gas-bearing strata such as shale oil, carbonate rocks, and tight sandstone, they significantly impact oil and gas accumulation, enrichment scale, and development effectiveness, making them a primary focus of exploration and development for unconventional oil and gas reservoirs. Under the combined effects of long-term tectonic stress, compaction, and diagenetic evolution, subsurface strata extensively develop directional vertical fractures. The fracture structure leads to significant differences in rock elastic stiffness, wave velocity, and attenuation characteristics with varying observation azimuth, exhibiting strong azimuth anisotropy. Horizontal Transverse Isotropy (HTI) media can describe the azimuth anisotropy induced by a single set of vertical fractures and is currently the most commonly used theoretical model for seismic characterization of fractured reservoirs. It effectively simplifies the seismic response description of complex fracture systems and provides a reliable theoretical basis for the quantitative inversion of key parameters. Prestack seismic inversion is an effective means of predicting anisotropic reservoir parameters. However, most existing inversion methods are based on approximate reflection coefficient equations. Although these equations are simple in form, they rely on the assumption of weak anisotropy and are not accurate enough under conditions of large incident angles and strong anisotropy, making it difficult to meet the requirements of high-precision exploration.

[0003] In HTI media, the azimuth anisotropy caused by fracture development is significant, making it an important research subject in the exploration of fractured hydrocarbon reservoirs. Currently, most reflection coefficient equations used in seismic inversion of HTI media are derived based on the assumption of weak anisotropy. These approximate equations are simple in form and only have a certain accuracy under small incident angles and weak anisotropy scenarios. Under large incident angles and strong anisotropy conditions, the approximate assumptions no longer hold, the calculation error of the equations increases, and it is difficult to accurately characterize the true seismic wave reflection patterns of HTI media. This makes it impossible to meet the requirements for detailed reservoir description under wide azimuth and large angle conditions, thus limiting the accuracy of detailed description and quantitative evaluation of fractured reservoirs. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention proposes a seismic inversion method for horizontally transversely isotropic media based on the elastic impedance tensor. The elastic impedance tensor can fully characterize the elastic properties of the medium and is not limited by the strength of anisotropy. It can break through the accuracy bottleneck of traditional approximation equations and provide important support for the fine description of reservoirs.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by this invention is: a seismic inversion method based on the elastic impedance tensor and HTI medium sensitivity equilibrium, comprising: Step S1: Construct the exact reflection coefficient equation for HTI medium based on the elastic impedance tensor; Step S2: Construct the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, which is directly characterized by P-wave and S-wave velocities, densities, and anisotropic parameters, according to the HTI medium accurate reflection coefficient equation based on elastic impedance tensor. Step S3: Construct a forward model based on the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, and introduce a parameter equilibrium matrix to construct a sensitivity equilibrium objective functional that the parameters of the model to be inverted obey, according to the sensitivity of different model parameters to seismic response. Step S4: The sensitivity equalization parameter adaptive Split-Bregman algorithm is introduced to solve the sensitivity equalization objective functional to obtain the P-wave and S-wave velocities, densities, and anisotropy parameters.

[0006] In a preferred embodiment, step S1, constructing the exact reflection coefficient equation of the HTI medium based on the elastic impedance tensor, includes: When an isotropic background contains vertical or near-vertical cracks, it is equivalent to an HTI medium. The precise elastic impedance tensor describing this medium is expressed as: (1), In equation (1), , ; Represents the elastic resistance tensor; These correspond to the incident P-wave, reflected P-wave, incident SV-wave, reflected SV-wave, incident SH-wave, and reflected SH-wave, respectively. Indicates the longitudinal wave velocity; Indicates the transverse wave velocity; Indicates density; Indicates anisotropy parameters; Indicates horizontal slowness; Indicates vertical slowness; Considering P-wave incident on the HTI medium layer interface, based on the stress and displacement continuity condition, then... The surface contains: (2), In equation (2), the superscripts 1 and 2 represent the upper and lower media, respectively; the subscripts + and - represent the down-going wave and the up-going wave, respectively. It is the vibration vector; For the propagation matrix; Let be the amplitude of the seismic wave propagation, and we have: (3), (4), In formula (3) This indicates the vertical slowness of the incident P-wave and the reflected P-wave. The exact reflection coefficient equation for HTI medium based on the elastic impedance tensor is obtained from equation (2): (5), have The coefficient matrix The exact reflection coefficient equation for HTI medium based on the elastic impedance tensor then transforms into: (6), Existence matrix and Make: (7), and and It can be represented as: (8), The identity matrix , These correspond to the upper and lower layers of media, respectively. Combining equations (7) and (8), the exact reflection coefficient equation of HTI medium based on the elastic impedance tensor is further transformed into: (9), in (10).

[0007] In a preferred embodiment, step S2, which involves constructing the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor, directly characterized by the P-wave and S-wave velocities, density, and anisotropic parameters, according to the precise reflection coefficient equation of the HTI medium based on the elastic impedance tensor, includes: When the difference between the upper and lower media is small: (11), matrix and The eigenvector matrix of the HTI media system matrix satisfies the following relationship: (12), Symbols in equations (11) and (12) This indicates the difference between the model parameters of the upper and lower media layers, indicated by the superscript. Represents the transpose of a matrix. For scaling factor: (13) By combining equations (9), (10), (11), and (12), equation (9) can be approximately written as: (14); Each matrix Substituting into equation (14), we get: (15) in, (16) In formula (15) It is the P-wave phase velocity, and has , , ,parameter Both are functions of P-wave and S-wave velocities and anisotropic parameters: (17) The vertical slow-motion perturbation term is written in the form of perturbation terms consisting of P-wave and S-wave velocities, densities, and anisotropy parameters: (18) in (19) and (20) Combining equations (15) and (18), we obtain the equation for the weak contrast reflection coefficient of HTI medium based on the elastic impedance tensor: (twenty one).

[0008] In a preferred embodiment, step S3, which involves constructing a forward model based on the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor, includes: Using the convolution model, equation (21) is convolved with the azimuth seismic wavelet to generate the corresponding seismic record. For seismic records containing... Angle of incidence Azimuth angle and The forward problem, which involves a wide-azimuth seismic data volume containing information about a reflection interface, is represented in matrix form as follows: (twenty two) in, , (twenty three), In equation (23), Represents the wavelet matrix, Represents the difference operation matrix. Represents earthquake records. Represents a diagonal matrix, with superscript This represents the transpose of a matrix.

[0009] In a preferred embodiment, step S3, which involves introducing a parameter equilibrium matrix to construct a sensitivity equilibrium objective functional to which the parameters of the model to be inverted conform, based on the sensitivity of different model parameters to the seismic response, includes: Based on the response characteristics of different parameters in the forward operator, the forward operator is divided into blocks: (twenty four), In equation (24), each parameter matrix block represents the forward matrix column block corresponding to the P-wave velocity, S-wave velocity, density, and anisotropic parameter, respectively. Calculate parameter sensitivity based on the energy of the matrix block corresponding to each parameter: ,in, Denotes the Frobenius norm of a matrix. Indicates the first The data dimensions corresponding to each parameter; The calculated sensitivity information is then normalized. (25) Construct a sensitivity balance matrix based on the sensitivity magnitudes of different parameters: (26) By using sensitivity equalization transformation, the original model parameters are converted into new parameter variables: (27) In equation (27), For sensitivity balancing matrix, These are the original model parameters. These are the parameter variables after sensitivity balancing; Convert the original forward model to: (28) In equation (28), Forward operators after sensitivity equalization; Construct the inverted objective functional within the sensitivity equilibrium parameter space: (29) In equation (29), the first term is the data fitting term, the second term is the sparse constraint term, the third term is the smoothing constraint term, and the fourth term is the low-frequency model constraint term.

[0010] In a preferred embodiment, step S4, which involves using an adaptive Split-Bregman algorithm to solve the sensitivity equalization objective functional and obtain the P-wave and S-wave velocities, densities, and anisotropy parameters, includes: The adaptive Split-Bregman algorithm, which introduces sensitivity balancing parameters, first introduces auxiliary variables: (30) The original objective function is transformed into a constrained optimization form: (31), To further address the differences in parameter sensitivity among different models, an adaptive penalty matrix related to parameters is constructed: (32), In equation (32), , , , , and These represent the penalty factors corresponding to different model parameters; Equation (31) can be written in Split-Bregman form: (33), In equation (33), Indicates the current iteration number. For the first The Bregman variable corresponding to the next iteration; In each iteration, the auxiliary variable is first fixed. and Bregman variables The model parameters in the sensitivity equilibrium space are updated by solving the following linear equation: (34), In equation (34): (35); Subsequently, with fixed model parameters In the case of auxiliary variables Update the following for the first... The update format for each model parameter block is as follows: (36) In equation (36) The soft threshold shrinkage operator is defined as follows: (37) After updating the auxiliary variables, correct the Bregman variables: (38) The penalty factors for each parameter are initialized based on the sensitivity information. For the first parameter... The initial penalty factor for each model parameter block is defined as: (39) In equation (39), Basic penalty factor, This is the sensitivity control coefficient; During the iteration process, the residual balance index is calculated based on the prior residuals and model residuals of each parameter block: (40), Based on the residual balance principle, the penalty factor is dynamically adjusted: (41), In equation (41), and To adjust the control parameters; The iteration stops when the model parameter updates meet the preset convergence condition, and an inverse sensitivity equalization transformation is performed: (42), By restoring the original parameter space, the inversion results of the longitudinal wave velocity, transverse wave velocity, density, and anisotropic parameters of the HTI medium are finally obtained.

[0011] Beneficial effects of this invention: The reflection coefficient equation based on the elastic impedance tensor proposed in this invention is not limited by the strength of anisotropy, has higher accuracy, and has a wider range of applications.

[0012] This invention introduces a parameter adaptation strategy guided by sensitivity information. It realizes parameter space transformation by constructing a sensitivity balance matrix and uses sensitivity information to initialize and dynamically adjust the penalty factor, thereby reducing the update differences and coupling effects between different parameters, improving the update capability of weakly sensitive anisotropic parameters and the stability and accuracy of multi-parameter inversion. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the process for an HTI medium sensitivity equilibrium seismic inversion method based on elastic impedance tensor provided in an embodiment of the present invention.

[0014] Figure 2 This is a schematic diagram of a horizontally isotropic medium model.

[0015] Figure 3 For longitudinal wave velocity Schematic diagram of the inversion result profile.

[0016] Figure 4transverse wave velocity Schematic diagram of the inversion result profile.

[0017] Figure 5 density Schematic diagram of the inversion result profile.

[0018] Figure 6 Anisotropic parameters Schematic diagram of the inversion result profile.

[0019] Figure 7 Anisotropic parameters Schematic diagram of the inversion result profile.

[0020] Figure 8 Anisotropic parameters Schematic diagram of the inversion result profile. Detailed Implementation

[0021] The detailed description and technical content of the present invention are explained below with reference to the accompanying drawings. However, the drawings are provided for reference and illustration only and are not intended to limit the present invention.

[0022] This invention derives the weak contrast reflection coefficient equation for HTI medium based on the elastic impedance tensor. Based on the derived equation, a forward model is constructed. A sensitivity balancing strategy is introduced, incorporating a parameter balancing matrix according to the sensitivity of different model parameters to seismic response. A sensitivity balancing objective functional is constructed, including data fitting terms, sparse constraint terms, and model constraint terms. The sensitive information-guided parameter adaptive Split-Bregman algorithm is used to solve the model. By dynamically adjusting the penalty factors corresponding to different parameters, stable inversion of P-wave and S-wave velocities, densities, and anisotropy parameters is achieved.

[0023] like Figure 1 As shown in the figure, the HTI medium sensitivity equalization seismic inversion method based on elastic impedance tensor provided by the present invention includes: Step S1: Construct the exact reflection coefficient equation for HTI medium based on the elastic impedance tensor; Step S2: Construct the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, which is directly characterized by P-wave and S-wave velocities, densities, and anisotropic parameters, according to the HTI medium accurate reflection coefficient equation based on elastic impedance tensor. Step S3: Construct a forward model based on the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, and introduce a parameter equilibrium matrix to construct a sensitivity equilibrium objective functional that the parameters of the model to be inverted obey, according to the sensitivity of different model parameters to seismic response. Step S4: The sensitivity equalization parameter adaptive Split-Bregman algorithm is introduced to solve the sensitivity equalization objective functional to obtain the P-wave and S-wave velocities, densities, and anisotropy parameters.

[0024] In some embodiments, step S1, constructing the exact reflection coefficient equation of the HTI medium based on the elastic impedance tensor, includes: like Figure 2 As shown, when an isotropic background contains vertical or near-vertical cracks, it can be equivalent to an HTI medium. The precise elastic impedance tensor describing this medium can be expressed as: (1), in , ; Represents the elastic resistance tensor; These correspond to the incident P-wave, reflected P-wave, incident SV-wave, reflected SV-wave, incident SH-wave, and reflected SH-wave, respectively. Indicates the longitudinal wave velocity; Indicates the transverse wave velocity; Indicates density; Indicates anisotropy parameters; Indicates horizontal slowness; Indicates vertical slowness.

[0025] Considering P-wave incident on the HTI medium layer interface, based on the stress and displacement continuity condition, then... The surface contains: (2), The superscripts 1 and 2 represent the upper and lower media, respectively; the subscripts "+" and "-" represent the downlink and uplink waves, respectively. It is the vibration vector; For the propagation matrix; This represents the amplitude of the seismic wave propagation.

[0026] And there are: (3), (4), In formula (3) Indicates when middle k The vertical slowness of the incident P-wave and the reflected P-wave at times 1 and 2.

[0027] Therefore, the reflection coefficient equation can be obtained from the system of equations (2): (5), Notice that The coefficient matrix Then the reflection coefficient equation can be transformed into: (6).

[0028] Observation shows that a matrix exists. and Make: (7), and and It can be represented as: (8), The identity matrix , These correspond to the upper and lower layers of media, respectively.

[0029] Combining equations (7) and (8), the reflection coefficient can be further transformed into: (9), in (10).

[0030] In some embodiments, step S2, which involves constructing the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor, directly characterized by the P-wave and S-wave velocities, densities, and anisotropic parameters, according to the precise reflection coefficient equation of the HTI medium based on the elastic impedance tensor, includes: When the difference between the upper and lower media is small, we have: (11), matrix and The eigenvector matrix of the HTI media system matrix therefore satisfies the following relationship: (12), Among the symbols This indicates the difference between the model parameters of the upper and lower media layers, indicated by the superscript. Represents the transpose of a matrix. For scaling factor: (13).

[0031] Combining equations (9), (10), (11), and (12), the reflection coefficient (9) can be approximated as: (14) Each matrix Substituting into equation (14), the HTI medium reflection coefficient equation can be specifically written as: (15) in, (16) In the formula It is the P-wave phase velocity, and has , , .parameter Both are functions of the P-wave and S-wave velocities and anisotropic parameters, specifically: (17) The vertical slow-motion perturbation term can be written in the form of perturbation terms consisting of P-wave and S-wave velocities, densities, and anisotropy parameters: (18) in (19) and (20) Combining equations (15) and (18), and simplifying, we obtain the equation for the weak contrast reflection coefficient of HTI medium based on the elastic impedance tensor: (twenty one).

[0032] In some embodiments, step S3, the step of constructing a forward model based on the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor, includes: By using a convolution model, convolving equation (21) with the azimuth seismic wavelet, the corresponding seismic record can be generated. For those containing Angle of incidence Azimuth angle and The forward problem, which involves a wide-azimuth seismic data volume containing information about a reflection interface, can be represented in matrix form as follows: (twenty two), in, , (twenty three), In the formula, Represents the wavelet matrix, Represents the difference operation matrix. Represents earthquake records. Represents a diagonal matrix, with superscript This represents the transpose of a matrix.

[0033] In some embodiments, step S3, which involves introducing a parameter equilibrium matrix to construct a sensitivity equilibrium objective functional to which the parameters of the model to be inverted conform, based on the sensitivity of different model parameters to the seismic response, includes: Because different model parameters exhibit varying sensitivities to seismic response, uneven parameter updates can easily occur, affecting the accuracy of anisotropic parameter inversion. Therefore, the forward modeling operator is first divided into blocks based on the response characteristics of different parameters in the forward modeling operator: (twenty four), Each parameter matrix block represents the forward modeling matrix column block corresponding to the P-wave velocity, S-wave velocity, density, and anisotropic parameters.

[0034] Calculate parameter sensitivity based on the energy of the matrix block corresponding to each parameter: , in, Denotes the Frobenius norm of a matrix. Indicates the first The data dimensions corresponding to each parameter.

[0035] The calculated sensitivity information is then normalized. (25).

[0036] Based on this, a sensitivity balancing matrix is ​​constructed according to the magnitude of sensitivity to different parameters: (26) By using sensitivity equalization transformation, the original model parameters are converted into new parameter variables: (27) in, For sensitivity balancing matrix, These are the original model parameters. These are the parameter variables after sensitivity equalization. Correspondingly, the original forward model is transformed into: (28) in, This is the forward modeling operator after sensitivity equalization. Through the above transformation, the difference between the response scales of different parameters can be reduced, and the updating ability of weakly sensitive parameters in the inversion process can be improved.

[0037] Therefore, an inversion objective functional is constructed within the sensitivity equilibrium parameter space: (29) The first term is the data fitting term, which is used to ensure the consistency between simulated seismic data and actual observation data; the second term is the sparsity constraint term, which is used to maintain the interface variation characteristics of model parameters; the third term is the smoothing constraint term, which is used to improve the stability of inversion results; and the fourth term is the low-frequency model constraint term, which is used to introduce prior information and reduce inversion uncertainty.

[0038] In some embodiments, step S4, which involves using an adaptive Split-Bregman algorithm to solve the sensitivity equalization objective functional and obtain the P-wave and S-wave velocities, densities, and anisotropy parameters, includes: To solve the above objective functional, the adaptive Split-Bregman algorithm, which introduces sensitivity balancing parameters, first introduces auxiliary variables: (30) The original objective function is transformed into a constrained optimization form: (31).

[0039] To address the differences in parameter sensitivity among different models, this method further constructs a parameter-dependent adaptive penalty matrix: (32), in, , , , , and These represent the penalty factors corresponding to different model parameters.

[0040] Therefore, equation (31) can be written in Split-Bregman form: (33), in, Indicates the current iteration number. For the first The Bregman variable corresponding to the next iteration.

[0041] In each iteration, the auxiliary variable is first fixed. and Bregman variables The model parameters in the sensitivity equilibrium space are updated by solving the following linear equation: (34), in: (35).

[0042] Subsequently, with fixed model parameters In the case of auxiliary variables Update. For the first Each model parameter block is updated in the following manner: (36) in The soft threshold shrinkage operator is defined as follows: (37).

[0043] After updating the auxiliary variables, correct the Bregman variables: (38).

[0044] To further utilize the sensitivity differences between different model parameters, this method initializes the penalty factor for each parameter based on sensitivity information. For the ... The initial penalty factor for each block of model parameters is defined as follows: (39) in, Basic penalty factor, This is the sensitivity control coefficient.

[0045] During the iteration process, the residual balance index is calculated based on the prior residuals and model residuals of each parameter block: (40), The penalty factor is dynamically adjusted based on the residual balance principle. (41), in, and To adjust the control parameters.

[0046] The iteration stops when the model parameter updates meet the preset convergence condition, and an inverse sensitivity equalization transformation is performed: (42), By restoring the original parameter space, the inversion results of the longitudinal wave velocity, transverse wave velocity, density, and anisotropic parameters of the HTI medium are finally obtained.

[0047] The method of this invention was tested using a high-quality source rock reservoir in eastern my country. Because this reservoir has well-developed vertical fractures, it can be considered an HTI medium. The azimuth angles of the seismic data were 15° and 105°. The near, middle, and far stacked average incident angles corresponding to each azimuth angle were 8°, 18°, and 28°, respectively. The model parameters were estimated using the seismic inversion method proposed in this invention. The inversion results are as follows: Figures 3 to 8 As shown, from Figures 3 to 8 It can be seen that the inversion profiles of each parameter have good lateral continuity, the main stratigraphic interfaces and parameter variation characteristics are relatively clear, the target reservoir location can be well characterized, and the well-side inversion results and logging curves have good consistency in overall variation trend and main interface location.

[0048] The embodiments described above are merely preferred embodiments for fully illustrating the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention.

Claims

1. A seismic inversion method based on HTI medium sensitivity equilibrium using elastic impedance tensor, characterized in that, include: Step S1: Construct the exact reflection coefficient equation for HTI medium based on the elastic impedance tensor; Step S2: Construct the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, which is directly characterized by P-wave and S-wave velocities, densities, and anisotropic parameters, according to the HTI medium accurate reflection coefficient equation based on elastic impedance tensor. Step S3: Construct a forward model based on the HTI medium weak contrast reflection coefficient equation based on elastic impedance tensor, and introduce a parameter equilibrium matrix to construct a sensitivity equilibrium objective functional that the parameters of the model to be inverted obey, according to the sensitivity of different model parameters to seismic response. Step S4: The sensitivity equalization parameter adaptive Split-Bregman algorithm is introduced to solve the sensitivity equalization objective functional to obtain the P-wave and S-wave velocities, densities, and anisotropy parameters.

2. The method as described in claim 1, characterized in that, Step S1, which involves constructing the accurate reflection coefficient equation for the HTI medium based on the elastic impedance tensor, includes the following steps: When an isotropic background contains vertical or near-vertical cracks, it is equivalent to an HTI medium. The precise elastic impedance tensor describing this medium is expressed as: （1）， In equation (1), , ; Represents the elastic resistance tensor; These correspond to the incident P-wave, reflected P-wave, incident SV-wave, reflected SV-wave, incident SH-wave, and reflected SH-wave, respectively. Indicates the longitudinal wave velocity; Indicates the transverse wave velocity; Indicates density; Indicates anisotropy parameters; Indicates horizontal slowness; Indicates vertical slowness; Considering P-wave incident on the HTI medium layer interface, based on the stress and displacement continuity condition, then... The surface contains: （2）， In equation (2), the superscripts 1 and 2 represent the upper and lower media, respectively; the subscripts + and - represent the down-going wave and the up-going wave, respectively. It is the vibration vector; For the propagation matrix; Let be the amplitude of the seismic wave propagation, and we have: （3）， （4）， In formula (3) This indicates the vertical slowness of the incident P-wave and the reflected P-wave. The exact reflection coefficient equation for HTI medium based on the elastic impedance tensor is obtained from equation (2): （5）， have The coefficient matrix The exact reflection coefficient equation for HTI medium based on the elastic impedance tensor then transforms into: （6）， Existence matrix and Make: （7）， and and It can be represented as: （8）， The identity matrix , These correspond to the upper and lower layers of media, respectively. Combining equations (7) and (8), the exact reflection coefficient equation of HTI medium based on the elastic impedance tensor is further transformed into: （9）， in （10）。 3. The method as described in claim 2, characterized in that, Step S2, which involves constructing the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor, directly characterized by the P-wave and S-wave velocities, densities, and anisotropic parameters, according to the precise reflection coefficient equation of the HTI medium based on the elastic impedance tensor, includes the following steps: When the difference between the upper and lower media is small: （11）, matrix and The eigenvector matrix of the HTI media system matrix satisfies the following relationship: （12）, Symbols in equations (11) and (12) This indicates the difference between the model parameters of the upper and lower media layers, indicated by the superscript. Represents the transpose of a matrix. For scaling factor: （13）， By combining equations (9), (10), (11), and (12), equation (9) can be approximately written as: （14）； Each matrix Substituting into equation (14), we get: （15）， in, （16）， In formula (15) It is the P-wave phase velocity, and has , , ,parameter Both are functions of P-wave and S-wave velocities and anisotropic parameters: （17）， The vertical slow-motion perturbation term is written in the form of perturbation terms consisting of P-wave and S-wave velocities, densities, and anisotropy parameters: （18）， in （19）， and （20）， Combining equations (15) and (18), we obtain the equation for the weak contrast reflection coefficient of HTI medium based on the elastic impedance tensor: （21）。 4. The method as described in claim 3, characterized in that, In step S3, the step of constructing a forward model based on the HTI medium weak contrast reflection coefficient equation based on the elastic impedance tensor includes: Using the convolution model, equation (21) is convolved with the azimuth seismic wavelet to generate the corresponding seismic record. For seismic records containing... Angle of incidence Azimuth angle and The forward problem, which involves a wide-azimuth seismic data volume containing information about a reflection interface, is represented in matrix form as follows: （22）， in, ， （23）， In equation (23), Represents the wavelet matrix, Represents the difference operation matrix. Represents earthquake records. Represents a diagonal matrix, with superscript This represents the transpose of a matrix.

5. The method as described in claim 4, characterized in that, In step S3, the step of introducing a parameter equilibrium matrix to construct the sensitivity equilibrium objective functional of the model parameters to be inverted, based on the sensitivity of different model parameters to seismic response, includes: Based on the response characteristics of different parameters in the forward operator, the forward operator is divided into blocks: （24）， In equation (24), each parameter matrix block represents the forward matrix column block corresponding to the P-wave velocity, S-wave velocity, density, and anisotropic parameter, respectively. Calculate parameter sensitivity based on the energy of the matrix block corresponding to each parameter: ,in, Denotes the Frobenius norm of a matrix. Indicates the first The data dimensions corresponding to each parameter; The calculated sensitivity information is then normalized. （25）， Construct a sensitivity balance matrix based on the sensitivity magnitudes of different parameters: （26）， By using sensitivity equalization transformation, the original model parameters are converted into new parameter variables: （27）， In equation (27), For sensitivity balancing matrix, These are the original model parameters. These are the parameter variables after sensitivity balancing; Convert the original forward model to: （28）， In equation (28), Forward operators after sensitivity equalization; Construct the inverted objective functional within the sensitivity equilibrium parameter space: （29）， In equation (29), the first term is the data fitting term, the second term is the sparse constraint term, the third term is the smoothing constraint term, and the fourth term is the low-frequency model constraint term.

6. The method as described in claim 5, characterized in that, Step S4, which involves using the sensitivity equalization parameter adaptive Split-Bregman algorithm to solve the sensitivity equalization objective functional and obtain the P-wave and S-wave velocities, densities, and anisotropy parameters, includes the following steps: The adaptive Split-Bregman algorithm, which introduces sensitivity balancing parameters, first introduces auxiliary variables: （30）， The original objective function is transformed into a constrained optimization form: （31）， To further address the differences in parameter sensitivity among different models, an adaptive penalty matrix related to parameters is constructed: （32）， In equation (32), , , , , and These represent the penalty factors corresponding to different model parameters; Equation (31) can be written in Split-Bregman form: （33）， In equation (33), Indicates the current iteration number. For the first The Bregman variable corresponding to the next iteration; In each iteration, the auxiliary variable is first fixed. and Bregman variables The model parameters in the sensitivity equilibrium space are updated by solving the following linear equation: （34）， In equation (34): （35）； Subsequently, with fixed model parameters In the case of auxiliary variables Update the following for the first... The update format for each model parameter block is as follows: （36）， In equation (36) The soft threshold shrinkage operator is defined as follows: （37）， After updating the auxiliary variables, correct the Bregman variables: （38）， The penalty factors for each parameter are initialized based on the sensitivity information. For the first parameter... The initial penalty factor for each model parameter block is defined as: （39）， In equation (39), Basic penalty factor, This is the sensitivity control coefficient; During the iteration process, the residual balance index is calculated based on the prior residuals and model residuals of each parameter block: （40）， Based on the residual balance principle, the penalty factor is dynamically adjusted: （41）， In equation (41), and To adjust the control parameters; The iteration stops when the model parameter updates meet the preset convergence condition, and an inverse sensitivity equalization transformation is performed: （42）， By restoring the original parameter space, the inversion results of the longitudinal wave velocity, transverse wave velocity, density, and anisotropic parameters of the HTI medium are finally obtained.

Images