Parameter Selection Method for Finite Difference Numerical Simulation of Second-Order Acoustic Equations

By setting dispersion error evaluation criteria and error thresholds, suitable combinations of finite difference parameters are selected, and a computational objective function is established, solving the problem of large computational load in the second-order acoustic wave equation and realizing efficient and low-cost selection of finite difference parameters.

CN115270579BActive Publication Date: 2026-11-17CHANGJIANG GEOTECHNICAL ENG CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210960522.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-08-11
Publication Date
2026-11-17
Estimated Expiration
2042-08-11

AI Technical Summary

Technical Problem

Existing technologies cannot effectively solve the problem of selecting finite difference parameters under the principle of minimizing the computational load of the second-order acoustic wave equation, resulting in problems of large computational load, low efficiency, and long processing time.

Method used

By setting dispersion error evaluation criteria and error thresholds, finite difference parameter combinations with errors less than the error thresholds are selected, and a computational objective function is established to optimize the selection of finite difference parameters with the minimum computational cost.

Benefits of technology

This method enables efficient selection of finite difference parameters for the second-order acoustic wave equation while minimizing computational complexity, thereby improving computational efficiency and reducing computational costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115270579B_ABST
    Figure CN115270579B_ABST
Patent Text Reader

Abstract

This invention discloses a method for selecting parameters in a finite-difference numerical simulation of a second-order acoustic wave equation. It includes the following steps: Step 1: Dispersion of the source wavelet; Step 2: Setting a dispersion error evaluation criterion and an error threshold; Step 3: Screening finite-difference parameter combinations based on the error evaluation criterion and a suitable error threshold; Step 4: Finite-difference discretization of the second-order constant-density acoustic wave equation to establish a computational objective function; Step 5: Calculating the finite-difference parameters corresponding to the minimum value of the objective function based on the computational objective function. This invention has the advantage of achieving the selection of finite-difference parameters under the principle of minimizing the computational load of the second-order acoustic wave equation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of finite difference forward modeling numerical simulation for seismic exploration, specifically to a method for selecting parameters in finite difference numerical simulation, and more specifically to a method for selecting parameters in finite difference numerical simulation of second-order acoustic wave equations. Background Technology

[0002] The finite difference method is widely used in numerical simulation of seismic wavefields. However, due to the discrete numerical calculation of the differential equations, finite difference numerical simulations result in varying propagation velocities for different wave frequencies. Higher frequency wavefields exhibit greater dispersion errors and more severe dispersion phenomena in the numerical simulation. Different difference parameters lead to different simulation results. Fine mesh refinement corresponds to high numerical simulation accuracy but also incurs significant computational costs. Currently, higher-order difference operators and optimized higher-order difference operator coefficients can achieve a compromise in terms of computational cost while maintaining high efficiency and accuracy in wavefield numerical simulation. However, they cannot achieve the principle of minimizing higher-order computational costs when selecting finite difference parameters, resulting in high computational complexity, low efficiency, and long processing times. Furthermore, no theoretical guidance on parameter selection for finite difference numerical simulation of wave equations has been found to date.

[0003] Therefore, it is necessary to develop a method for selecting finite difference parameters that can achieve the principle of minimizing the computational cost of the second-order acoustic wave equation. Summary of the Invention

[0004] The purpose of this invention is to provide a method for selecting parameters in finite difference numerical simulation of the second-order acoustic wave equation, achieving finite difference parameter selection under the principle of minimizing the computational cost of the second-order acoustic wave equation. It provides theoretical guidance for finite difference parameter selection under the principle of minimizing the computational cost of the second-order acoustic wave equation, with low computational cost and high efficiency. It overcomes the shortcomings of existing technologies that cannot achieve finite difference parameter selection under the principle of minimizing the computational cost of the second-order acoustic wave equation, which suffers from high computational cost, low efficiency, and long time consumption.

[0005] To achieve the above objectives, the technical solution of the present invention is: a method for selecting parameters in a finite difference numerical simulation of a second-order acoustic wave equation, characterized by comprising the following steps:

[0006] Step 1: Dispersion is added to the source wavelet for comparison with a reference wave field without dispersion. In Step 2, the waveform with dispersion and the reference wave field without dispersion are used for error assessment. The waveform with dispersion is the waveform after dispersion is added in Step 1.

[0007] Step Two: Set the dispersion error evaluation criteria and error threshold. Dispersion affects the final waveform, and different degrees of dispersion have different impacts on the waveform. The error evaluation criteria are used to guide the determination of a suitable error threshold; the dispersion waveform corresponding to the error threshold is determined to be invalid dispersion and discarded, and the corresponding finite difference parameters are also determined to be invalid dispersion and discarded; the finite difference parameters corresponding to dispersion less than the error threshold are determined to be valid dispersion. This invention determines the optimal finite difference parameters from the valid dispersion.

[0008] Step 3: Based on the error evaluation criteria and a suitable error threshold, screen the combination of finite difference parameters. The finite difference parameters corresponding to the dispersion less than the error threshold are determined as effective finite difference parameters. These effective finite difference parameters include the optimized finite difference parameters.

[0009] Step 4: Perform finite difference discretization on the second-order constant density wave equation of acoustic waves, and establish a computational objective function. Different finite difference parameters will correspond to different computational costs. By establishing the computational objective function, the combination of finite difference parameters that minimizes the computational cost can be selected.

[0010] Step 5: Based on the objective function of computational complexity, calculate the finite difference parameters corresponding to the minimum value of the objective function.

[0011] In the above technical solution, in step one, the theoretical source wavelet or the actual field source wavelet is dispersed;

[0012] In step one, the selected source wavelet is a theoretical wavelet with an analytical expression (such as a Gaussian function, the first derivative wavelet of the Ricker wavelet, etc.) or a wavelet extracted from field seismic data. The following method is used to perform spatial numerical dispersion:

[0013] S11: For the reference wavefield (i.e., the wavefield without dispersion, which is the wavefield before dispersion in step one or the source wavelet). Perform a discrete Fourier transform to obtain the frequency domain wave field. The formula is as follows:

[0014] (1)

[0015] In formula (1): It is a time variable; For spatial coordinate variables; The distance of wave field propagation; The positively mapped wavenumber of the central difference scheme for the second-order acoustic wave equation is expressed as:

[0016] (2)

[0017] In formula (2): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; Wave number; The length of the finite difference operator; The space is far from the walking distance;

[0018] S12: For frequency domain wave fields Perform a Fast Fourier Transform (FFT) algorithm to inverse transform back to the time domain to obtain the wave field with numerical dispersion. The formula is as follows:

[0019] (3)

[0020] In formula (3): It is a time variable; For spatial coordinate variables; The space is far from the walking distance; For wave number.

[0021] In the above technical solution, the dispersion error evaluation criteria and error threshold are set, and the specific method is as follows:

[0022] In step two, based on the reference wave field from step one... And the wave field with added numerical dispersion The normalized L2 error evaluation criterion is defined as follows:

[0023] (4)

[0024] The error between the numerically dispersed wave field and the real wave field is used as the basis for judging the degree of dispersion, and an appropriate error threshold is selected according to the magnitude of the dispersion error.

[0025] In the above technical solution, the threshold is selected as 0.01.

[0026] In the above technical solution, in step three, based on the normalized L2 error evaluation criterion given in step two, all finite difference parameter combinations corresponding to errors less than or equal to the error threshold are selected; the finite difference parameter combinations include the spatial sampling step size. , length of finite difference operator and finite difference coefficients wait.

[0027] In the above technical solution, in step four, the second-order constant density wave equation of acoustic waves is discretized by finite difference to establish the objective function of computation. The specific steps are as follows:

[0028] S4.1: Discretize the second-order constant-density acoustic wave equation using the central difference scheme:

[0029] (5)

[0030] In equation (5): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; The time spent walking is long; The length of the finite difference operator; Indicates a point in time Spatial coordinates Wave field variables; Indicates the spatial coordinate position The magnitude of the speed;

[0031] The sum of the addition and multiplication operations is called the computational complexity. When the equation is discretized, the computational complexity involved is as follows: For the central difference discrete calculation of the three-dimensional wave equation, the multiplication operation at each grid point at each time step is... The addition operation has a computational cost of 100%. The computational complexity is The two-dimensional wave equation has a multiplication complexity of... The addition operation has a computational cost of 100%. The computational complexity is The one-dimensional wave equation has a multiplication complexity of... The addition operation has a computational cost of 100%. The computational complexity is Referring to the computational complexity analysis above, similar computational complexity can be obtained using different wave equations and different difference schemes;

[0032] S4.2: Mesh the velocity model and establish the objective function for computational complexity.

[0033] In the above technical solution, the velocity model body includes a three-dimensional cube velocity model and a two-dimensional rectangular velocity model;

[0034] The number of mesh points in the velocity model body, NUM, is equal to the volume divided by the mesh spacing.

[0035] Computational complexity of velocity model It is related to the number of grid points NUM and the difference order N.

[0036] In the above technical solution, when a three-dimensional cube velocity model is selected for the velocity model body, the length, width, and height are respectively... A 3D cube of meters with a grid spacing of [missing information] Meter grid division, number of grid points and computational complexity They are respectively:

[0037] (6)

[0038] (7)

[0039] When a two-dimensional rectangular velocity model is selected for the velocity model body, the length and width are respectively... A rectangle of meters with a grid spacing of [value missing] Meter grid division, number of grid points and computational complexity They are respectively:

[0040] (8)

[0041] (9).

[0042] In the above technical solution, in step five, based on all finite difference parameter combinations corresponding to the error being less than or equal to the error threshold selected in step three, and based on the objective function of computational complexity in step four, the finite difference parameter combination corresponding to the minimum value of the objective function is calculated, and the finite difference numerical simulation parameters under the principle of minimum computational complexity are selected.

[0043] The present invention has the following advantages due to the adoption of the above technical solutions: Based on the analytical expression of numerical dispersion in the spatial direction of the wave field, the present invention provides theoretical guidance for the selection of finite difference parameters for the second-order acoustic wave equation while minimizing the computational load of finite difference numerical simulation. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the present invention;

[0045] Figure 2 This is a schematic diagram illustrating the recommended values ​​for selecting finite difference parameters when the wavelet is a Ricker wavelet in an embodiment of the present invention.

[0046] Figure 2 In the middle G, it represents the number of wavelength sampling points corresponding to 2.5 times the main frequency of the theoretical Ricker wavelet. Detailed Implementation

[0047] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, these descriptions do not constitute a limitation of the present invention and are merely illustrative. The advantages of the present invention will become clearer and easier to understand through this description.

[0048] like Figure 1 As shown, this invention provides a method for selecting parameters in a finite-difference numerical simulation of a second-order acoustic wave equation with a minimum computational objective function, which includes the following steps:

[0049] 1) Extract the source wavelet or theoretical source wavelet from the real earthquake record and perform spatial dispersion operation;

[0050] Specifically as follows:

[0051] Set up a source wavefield without numerical dispersion. Using the reference wavefield, a forward spatial dispersion transform (FSDT) is performed on the reference wavefield to add dispersion:

[0052]

[0053] In the formula, For positively mapped wavenumber, For time variables, For spatial coordinate variables, This represents the propagation distance of the wave field.

[0054] The specific implementation process of spatial dispersion is as follows:

[0055] 1.1) For the reference wave field Perform a discrete Fourier transform to obtain the frequency domain wave field. The formula is as follows:

[0056] (1)

[0057] In formula (1): It is a time variable; For spatial coordinate variables; The distance of wave field propagation; The positively mapped wavenumber of the central difference scheme for the second-order acoustic wave equation is expressed as:

[0058] (2)

[0059] In formula (2): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; Wave number; The length of the finite difference operator;

[0060] 1.2) For frequency domain wave fields Perform a Fast Fourier Transform (FFT) algorithm to inverse transform back to the time domain to obtain the wave field with numerical dispersion. The formula is as follows:

[0061] (3)

[0062] In formula (3): It is a time variable; For spatial coordinate variables; For wave number.

[0063] 2) Set the dispersion error evaluation criteria and select an error threshold of 0.01;

[0064] Specifically as follows:

[0065] It is suggested that the normalized L2 norm between the numerically dispersed wavefield and the reference wavefield be used to measure the dispersion error.

[0066] (4)

[0067] It is recommended to use the error between the numerically dispersed wave field and the real wave field as the basis for judging the degree of dispersion. A suitable error threshold should be selected according to the magnitude of the dispersion error, and a threshold of 0.01 is recommended.

[0068] 3) Based on an error threshold of 0.01, different combinations of finite difference parameters are filtered to obtain the finite difference parameters under the given error threshold: spatial sampling step size. , length of finite difference operator and finite difference coefficients Combinations, etc.

[0069] 4) Establish the objective function for computational complexity;

[0070] The specific implementation process is as follows:

[0071] 4.1) Discretize the second-order constant-density acoustic wave equation using central grid difference:

[0072] (5)

[0073] In equation (5): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; The time spent walking is long; The length of the finite difference operator; Indicates a point in time Spatial coordinates Wave field variables; Indicates the spatial coordinate position The speed magnitude. The central difference of the three-dimensional wave equation is calculated discretely; the multiplication operation at each grid point in each time step is... The addition operation has a computational cost of 100%. The computational complexity is (The sum of the addition and multiplication operations is called the computational complexity). For a two-dimensional wave equation, the multiplication complexity is... The addition operation has a computational cost of 100%. The computational complexity is (The sum of the addition and multiplication operations is called the computational complexity). For a one-dimensional wave equation, the multiplication complexity is... The addition operation has a computational cost of 100%. The computational complexity is (The sum of the addition and multiplication operations is called the computational complexity). Referring to the computational complexity analysis above, similar computational complexity can be obtained for different wave equations and different difference schemes.

[0074] 4.2) Mesh the velocity model:

[0075] For example, a three-dimensional cube velocity model, with length, width, and height respectively... A 3D cube of meters with a grid spacing of [missing information] Meter grid division, number of grid points for:

[0076] (6)

[0077] For example, in a two-dimensional rectangular velocity model, with length and width respectively... A rectangle of meters with a grid spacing of [value missing] Meter grid division, number of grid points for:

[0078] (8)

[0079] 4.3) Establish the objective function for computational complexity. :

[0080] 3D cube velocity model:

[0081] (7)

[0082] Two-dimensional rectangular velocity model:

[0083] (9)

[0084] 5) Based on the objective function of computational complexity By filtering different sets of finite difference parameters, the finite difference parameters corresponding to the minimum objective function (with minimal computational cost) are calculated (e.g., ...). Figure 2 (As shown).

[0085] In summary, this invention provides theoretical guidance for selecting finite difference parameters of the second-order acoustic wave equation under the principle of minimizing computational complexity.

[0086] Example

[0087] This paper presents a detailed description of the invention using an example of selecting finite difference parameters under the principle of minimizing the computational cost of realizing the second-order acoustic wave equation in a Ricker wavelet reference wavefield. The invention also provides guidance for the application of this invention to spatial numerical dispersion simulation and finite difference parameter optimization of other source wavelets (e.g., theoretical wavelets with analytical expressions (e.g., Gaussian functions, etc.) or wavelets extracted from field seismic data).

[0088] like Figure 1 As shown, this embodiment uses the present invention to optimize the selection of finite difference parameters for the second-order acoustic wave equation with the objective function of minimizing computational complexity, which includes the following steps:

[0089] 1) Perform spatial dispersion operation based on Ricker wavelet;

[0090] Specifically as follows:

[0091] Set up a source wavefield without numerical dispersion. For the reference wavefield, a spatial dispersion positive transform is performed on the Ricker wavelet of the reference wavefield, and a dispersion operation is performed:

[0092]

[0093] In the formula, For positively mapped wavenumber, For time variables, For spatial coordinate variables, This represents the propagation distance of the wave field.

[0094] The specific implementation process of spatial dispersion is as follows:

[0095] 1.1) For the reference wave field Perform a discrete Fourier transform to obtain the frequency domain wave field. The formula is as follows:

[0096] (1)

[0097] In formula (1): It is a time variable; For spatial coordinate variables; The distance of wave field propagation; in equation (1) For positive mapping wavenumber:

[0098] (2)

[0099] In formula (2): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; Wave number; The length of the finite difference operator;

[0100] 1.2) For frequency domain wave fields Perform a Fast Fourier Transform (FFT) algorithm to inverse transform back to the time domain to obtain the wave field with numerical dispersion. The formula is as follows:

[0101] (3)

[0102] In formula (3): It is a time variable; For spatial coordinate variables; For wave number.

[0103] 2) Set the dispersion error evaluation criteria. In this embodiment, the error threshold is selected as 0.01.

[0104] In this embodiment, the Ricker wavelet is a second-order acoustic wave, and the method for setting the dispersion error evaluation criterion is as follows:

[0105] It is suggested that the normalized L2 norm between the numerically dispersed wavefield and the reference wavefield be used to measure the dispersion error.

[0106] (4)

[0107] This embodiment suggests using the error between the numerically dispersed wavefield and the real wavefield as the basis for judging the degree of dispersion. Based on the magnitude of the dispersion error, an appropriate error threshold should be selected. This embodiment suggests selecting a threshold of 0.01.

[0108] 3) Based on an error threshold of 0.01, different combinations of finite difference parameters are filtered to obtain the finite difference parameters under the given error threshold: spatial sampling step size. , length of finite difference operator and finite difference coefficients Combinations, etc.

[0109] 4) Establish the objective function for computational complexity;

[0110] The specific implementation process of this embodiment is as follows:

[0111] 4.1) Discretize the second-order constant-density acoustic wave equation using central grid difference:

[0112] (5)

[0113] In equation (5): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; The time spent walking is long; The length of the finite difference operator; Indicates a point in time Spatial coordinates Wave field variables; Indicates the spatial coordinate position The speed magnitude. The two-dimensional wave equation has a multiplication operation complexity of... The addition operation has a computational cost of 100%. The computational complexity is (The sum of the addition and multiplication operations is called the computational workload).

[0114] 4.2) Mesh the velocity model:

[0115] This embodiment uses a two-dimensional rectangular velocity model, with length and width respectively... A rectangle of meters with a grid spacing of [value missing] Meter grid division, number of grid points for:

[0116] (8)

[0117] 4.3) Establish the objective function for computational complexity. :

[0118] Two-dimensional rectangular velocity model:

[0119] (9)

[0120] 5) Based on the objective function of computational complexity By filtering different sets of finite difference parameters, the finite difference parameters corresponding to the minimum objective function (with minimal computational cost) are calculated (e.g., ...). Figure 2 (As shown).

[0121] Conclusion: The following are suggestions for selecting finite difference parameters obtained using the method of this invention in this embodiment: Figure 2 As shown, Figure 2 To use the theoretical Ricker wavelet as the reference wavefield, based on the Taylor series expansion method and the Remitz exchange method, we recommend the following finite difference parameter values. Figure 2 In the middle G, it represents the number of wavelength sampling points corresponding to 2.5 times the main frequency of the theoretical Ricker wavelet. Figure 2 Figure (a) shows the recommended values ​​for the finite difference parameters under the objective of minimizing the computational cost of the difference coefficients calculated using the Taylor series expansion method; Figure 2 In Figure (a), it is recommended that the order of the difference operator calculated based on the Taylor series expansion method be selected as 16 to 20, and the number of wavelength sampling points be selected as 3 to 3.2. Figure 2 Figure (b) shows the recommended values ​​for the finite difference parameters under the objective of minimizing computational cost corresponding to the optimization class difference coefficients calculated by the Remitz exchange method; Figure 2In Figure (b), it is recommended that the order of the difference operator calculated based on the Remitz exchange method be selected as 14-18, and the number of wavelength sampling points be selected as 2.6-2.7. This embodiment can be referenced for any finite difference method when the source wavelet is a Ricker wavelet. Figure 2 Selection of finite difference parameters.

[0122] Other source wavelets (such as theoretical wavelets with analytical expressions (e.g., Gaussian functions, etc.) or wavelets extracted from field seismic data) can also have their reference wavefields optimized using this method to obtain the optimal finite difference numerical simulation parameters.

[0123] The above embodiments are only used to illustrate the present invention. Each step can be varied. Based on the technical solution of the present invention, any improvements and equivalent transformations made to individual steps according to the principles of the present invention should not be excluded from the protection scope of the present invention.

[0124] All other unspecified parts belong to the prior art.

Claims

1. A method for selecting parameters in finite difference numerical simulation of second-order acoustic wave equations, characterized in that: Includes the following steps, Step 1: Dispersion of the source wavelet; Select a theoretical wavelet with an analytical expression or a wavelet extracted from field seismic data, and perform dispersion on the source wavelet. The following method is used to perform spatial numerical dispersion: S11: For the reference wave field Perform a discrete Fourier transform to obtain the frequency domain wave field. The formula is as follows: (1) In formula (1): Wave number; It is a time variable; For spatial coordinate variables; The distance of wave field propagation; The positively mapped wavenumber of the central difference scheme for the second-order acoustic wave equation is expressed as: (2) In formula (2): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; Wave number; The length of the finite difference operator; S12: For frequency domain wave fields Perform a Fast Fourier Transform (FFT) algorithm to inverse transform back to the time domain to obtain the wave field with numerical dispersion. The formula is as follows: (3) In formula (3): It is a time variable; For spatial coordinate variables; Wave number; Step 2: Set the dispersion error evaluation criteria and error threshold; Step 3: Based on the error evaluation criteria and error threshold, select finite difference parameter combinations; the finite difference parameter combinations include the spatial sampling step size. , length of finite difference operator and finite difference coefficients ; Step 4: Perform finite difference discretization on the second-order constant density wave equation of acoustic waves and establish the objective function for computational complexity; Step 5: Based on the objective function of computational complexity, calculate the finite difference parameters corresponding to the minimum value of the objective function.

2. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 1, characterized in that: The dispersion error evaluation criteria and error threshold are set, and the specific method is as follows: In step two, based on the reference wave field in step one... And the wave field with added numerical dispersion The normalized L2 error evaluation criterion is defined as follows: (4) The error between the numerically dispersed wave field and the real wave field is used as the basis for judging the degree of dispersion, and the error threshold is selected according to the magnitude of the dispersion error.

3. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 2, characterized in that: The threshold was set to 0.

01.

4. The method for selecting parameters for finite difference numerical simulation of the second-order acoustic wave equation according to claim 3, characterized in that: In step three, based on the normalized L2 error evaluation criteria given in step two, all finite difference parameter combinations corresponding to errors less than or equal to the error threshold are selected.

5. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 4, characterized in that: In step four, the second-order constant density wave equation of acoustic waves is discretized by finite difference to establish the objective function of computation. The specific steps are as follows: S4.1: Discretize the second-order constant-density acoustic wave equation using the central difference scheme: (5) In formula (5): These are the coefficients of the higher-order difference operator; The space is far from the walking distance; The time spent walking is long; The length of the finite difference operator; Indicates a point in time Spatial coordinates Wavefield variables; Indicates the spatial coordinate position The magnitude of the speed; S4.2: Mesh the velocity model and establish the objective function for computational complexity.

6. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 5, characterized in that: The velocity model includes a three-dimensional cube velocity model and a two-dimensional rectangular velocity model; The number of mesh points NUM for the velocity model body = volume / mesh spacing.

7. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 6, characterized in that: When a three-dimensional cube velocity model is selected for the velocity model body, the length, width, and height are respectively... A 3D cube of meters with a grid spacing of [missing information] Meter grid division, number of grid points and computational complexity They are respectively: (6) (7) When a two-dimensional rectangular velocity model is selected for the velocity model body, the length and width are respectively... A rectangle of meters with a grid spacing of [value missing] Meter grid division, number of grid points and computational complexity They are respectively: (8) (9)。 8. The method for selecting parameters in the finite difference numerical simulation of the second-order acoustic wave equation according to claim 7, characterized in that: In step five, based on all finite difference parameter combinations that are less than or equal to the error threshold selected in step three, and based on the objective function of computational complexity in step four, the finite difference parameter combination corresponding to the minimum value of the objective function is calculated, and the finite difference numerical simulation parameters under the principle of minimum computational complexity are selected.