Velocity model inversion method, device, equipment and storage medium

Abstract

This application discloses a method, apparatus, device, and storage medium for inverting a velocity model, relating to the field of Earth sciences. The method includes: generating multi-order picking dispersion curves for P-guided waves based on acquired seismic data; performing forward modeling on the multi-order picking dispersion curves based on an initial velocity model to obtain at least one depth-sensitive core for the P-guided waves; performing inversion using inversion equations based on the depth-sensitive core to obtain an update to the initial velocity model; adjusting the initial velocity model based on the update to obtain an updated initial velocity model; if the updated initial velocity model satisfies a first convergence condition, then re-establishing the initial velocity model based on the updated initial velocity model; if the updated initial velocity model satisfies a second convergence condition, then determining the updated initial velocity model as the inverted velocity model; and executing the steps corresponding to the forward modeling. This method improves the agreement between the final obtained velocity model and the theoretical model.

Description

Technical Field

[0001] This application relates to the field of earth sciences, and in particular to a method, apparatus, device, and storage medium for inverting a velocity model. Background Technology

[0002] Seismic waves, as an important tool in geophysical exploration, can provide valuable information about the structure and properties of subsurface media. P-guided waves are a propagation mode primarily dominated by longitudinal waves (P-waves), characterized by propagation along a waveguide in environments with large Poisson distributions, and observationally approximate an ideal guided wave mode. Therefore, studying P-guided waves can invert the shallow P-wave velocity structure, thereby revealing variations in the physical properties of the subsurface media.

[0003] In related technologies, the P-guided wave inversion method is similar to the surface wave inversion method, mainly relying on damped least squares inversion techniques. This method utilizes the P-guided wave phase velocity dispersion curve to perform one-dimensional velocity-structure inversion on seismic data at a single point. Through the inversion process, a velocity model of the P-wave is obtained, thus providing a basis for shallow seismic structure research.

[0004] Because the velocity model obtained through related technologies is a continuously changing model, it is difficult to accurately reflect the location and properties of the velocity abrupt change interface. Summary of the Invention

[0005] This application provides a method, apparatus, device, and storage medium for inverting a velocity model. The technical solution provided by this application is as follows:

[0006] According to one aspect of the embodiments of this application, a method for inverting a velocity model is provided, the method comprising:

[0007] Acquire earthquake data, which is data collected after seismic waves propagate through the subsurface medium;

[0008] Based on the earthquake data, a multi-order picking dispersion curve of the P-guided wave is generated, which is used to describe the relationship between the wave velocity of the P-guided wave and the frequency.

[0009] Forward modeling is performed on the multi-order picking dispersion curve to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on wave velocity in the first underground medium, which is the underground medium simulated by the initial velocity model.

[0010] If the multi-order predicted dispersion curve satisfies the first convergence condition, the initial velocity model is re-established based on the initial velocity model to obtain the reconstructed initial velocity model.

[0011] If the multi-order predicted dispersion curve satisfies the second convergence condition, the initial velocity model is determined as the inverted velocity model.

[0012] Based on the at least one depth-sensitive kernel, an inversion equation is used to perform inversion and obtain the update amount of the initial velocity model. Based on the update amount of the initial velocity model, the initial velocity model is adjusted to obtain the updated initial velocity model. The inversion equation is a relationship between the at least one depth-sensitive kernel, the update amount of the initial velocity model, the multi-order pickup dispersion curve of the P-guided wave and the multi-order prediction dispersion curve of the P-guided wave.

[0013] The process begins again from the step of performing forward modeling on the multi-order pickup dispersion curve based on the initial velocity model.

[0014] According to one aspect of the embodiments of this application, an inversion apparatus for a velocity model is provided, the apparatus comprising:

[0015] The acquisition module is used to acquire earthquake data, which is data collected after seismic waves propagate in the underground medium;

[0016] The generation module is used to generate a multi-order picking dispersion curve of the P-guided wave based on the seismic data. The multi-order picking dispersion curve is used to describe the relationship between the wave velocity of the P-guided wave and the frequency.

[0017] The forward modeling module is used to perform forward modeling on the multi-order picking dispersion curve based on the initial velocity model to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on the wave velocity in the first underground medium, which is the underground medium simulated by the initial velocity model.

[0018] The reconstruction module is used to re-establish the initial velocity model based on the initial velocity model, and obtain the reconstructed initial velocity model, provided that the multi-order predicted dispersion curve satisfies the first convergence condition.

[0019] The determination module is used to determine the initial velocity model as the inverted velocity model when the multi-order predicted dispersion curve satisfies the second convergence condition.

[0020] The inversion module is used to perform inversion based on the at least one depth-sensitive kernel using an inversion equation to obtain the update amount of the initial velocity model, and to adjust the initial velocity model based on the update amount of the initial velocity model to obtain an updated initial velocity model; wherein, the inversion equation is a relationship between the at least one depth-sensitive kernel, the update amount of the initial velocity model, the multi-order pickup dispersion curve of the P-guided wave and the multi-order prediction dispersion curve of the P-guided wave;

[0021] The execution module is used to start execution again from the step of forward modeling the multi-order picking dispersion curve based on the initial velocity model.

[0022] According to one aspect of the present application, a computer device is provided, the computer device including a processor and a memory, the memory storing a computer program, the computer program being loaded and executed by the processor to implement the above-described speed model inversion method.

[0023] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, wherein a computer program is stored in the storage medium, the computer program being loaded and executed by a processor to implement the above-described speed model inversion method.

[0024] According to one aspect of the embodiments of this application, a computer program product is provided, the computer program product including a computer program, the computer program being loaded and executed by a processor to implement the above-described speed model inversion method.

[0025] The technical solutions provided in this application have at least the following beneficial effects:

[0026] The method provided in this application obtains the initial velocity model after the first round of inversion by picking up the dispersion curves of the P-guided wave through multiple forward modeling in the first round. Then, it analyzes the initial velocity model to re-layer it. Based on the re-layered initial velocity model, a new initial velocity model is re-established. The above steps are then repeated using this reconstructed initial velocity model to obtain the final inverted velocity model. This method reduces the inversion parameters required for inversion by re-classifying the initial velocity model, thus improving the agreement between the final velocity model and the theoretical model. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the implementation environment of a solution provided in one embodiment of this application;

[0028] Figure 2 This is a flowchart of a forward modeling method for a velocity model provided in one embodiment of this application;

[0029] Figure 3This is a schematic diagram of the multi-order pickup dispersion curve provided in the embodiments of this application;

[0030] Figure 4 This is a schematic diagram of a velocity change rate curve provided in one embodiment of this application;

[0031] Figure 5 This is a flowchart of a velocity model inversion method provided in another embodiment of this application;

[0032] Figure 6 This is a schematic diagram of a multi-order prediction dispersion curve provided in one embodiment of this application;

[0033] Figure 7 This is a schematic diagram of an initial velocity model provided in one embodiment of this application;

[0034] Figure 8 This is a schematic diagram of a multi-order prediction dispersion curve provided in another embodiment of this application;

[0035] Figure 9 This is a schematic diagram of an initial velocity model provided in another embodiment of this application;

[0036] Figure 10 This is a schematic diagram of the conformity change curve provided in one embodiment of this application;

[0037] Figure 11 This is a block diagram of a velocity model inversion device provided in one embodiment of this application;

[0038] Figure 12 This is a structural block diagram of a computer device provided in one embodiment of this application. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0040] Please refer to Figure 1 The diagram illustrates an implementation environment for a solution provided in one embodiment of this application. This implementation environment may include: computer device 10.

[0041] Please refer to Figure 1 The diagram illustrates an implementation environment for a solution provided in one embodiment of this application. This implementation environment may include: a terminal device 10 and a server 20.

[0042] Terminal devices 10 include, but are not limited to, PCs (Personal Computers), host computers of cloud computing technology platforms, mobile phones, tablet computers, smart voice interaction devices, game consoles, wearable devices, multimedia playback devices, vehicle terminals, smart home appliances, AR (Augmented Reality) devices, VR (Virtual Reality) devices, and other electronic devices.

[0043] Server 20 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms, but it is not limited to these.

[0044] Terminal device 10 and server 20 can communicate with each other via a network. This network can be a wired network or a wireless network.

[0045] Please refer to Figure 2 The diagram illustrates a flowchart of a velocity model inversion method provided in one embodiment of this application. The execution entity for each step of this method can be a computer device; for example, the computer device could be... Figure 1 Terminal device 10 in the middle can also be Figure 1 Server 20. The method may include at least one of the following steps (210-270):

[0046] Step 210: Obtain earthquake data. Earthquake data is data collected after seismic waves propagate through the underground medium.

[0047] Seismic data is generated during geophysical exploration by artificially generating seismic waves at the Earth's surface. As these waves propagate downwards, reflection and refraction occur due to inconsistencies in the geological medium and lithology. The reflected seismic waves are then received at the surface by seismic exploration geophones, and the resulting data includes records of the reflected seismic waves, their arrival time, and intensity.

[0048] In some embodiments, seismic data is shot gather data. Shot gather data refers to the collection of seismic wave signals recorded by multiple geophones deployed on or beneath the surface after a single seismic source excites the earthquake. Due to the complex environment of seismic exploration, the raw data typically contains various types of noise, such as random noise, multipath effects, and instrument noise. In some embodiments, the seismic data is denoised to obtain denoised seismic data.

[0049] Step 220: Based on the seismic data, generate multi-order picking dispersion curves for the P-guided wave. The multi-order picking dispersion curves are used to describe the relationship between the wave velocity of the P-guided wave and the frequency.

[0050] Because real wave field propagation involves the coupling of P-waves (longitudinal waves) and S-waves (transverse waves), when the longitudinal wave reaches the lower interface, some energy always escapes into the half-space as S-waves. Therefore, the observed "guided wave" dispersion curve in this case is not a strictly guided wave mode, but a leakage mode. However, in some environments with high Poisson's ratios (such as shallow seas or shallow surfaces), the coupling degree between the longitudinal and transverse waves is very low, which may cause the observed dispersion of this leakage mode to be almost identical to the guided wave dispersion under pure acoustic wave theory. In addition, since this part of the wave field energy is dominated by longitudinal waves and no energy leaks into the half-space in the form of longitudinal waves, these waves are also called P-guided waves. This mode is dominated by longitudinal waves and does not radiate longitudinal wave energy into the half-space, so the attenuation rate is very slow when propagating along the waveguide. P-guided waves have a strong dispersion effect and a high phase velocity, approaching the velocity of reflected P-waves at large offsets. Shallow sea environments are prone to generating P-guided waves. P-guided waveguides have the following characteristics: their dispersion modes have a resonant frequency tuned appearance; they have a relatively high cutoff frequency; and the phase velocity of the P-guided wave exceeds the propagation speed of the P-wave in water.

[0051] The multi-order pickup dispersion profile of a P-guided wave refers to the set of pickup dispersion profiles extracted from seismic data that reflect the wave velocity variations of the P-guided wave at different high frequencies. Dispersion refers to the phenomenon that the phase velocity or group velocity of the P-guided wave changes with frequency. That is, the wave velocity of the P-guided wave corresponds to the phase velocity in the multi-order pickup dispersion profile. In some embodiments, the multi-order pickup dispersion profile includes a fundamental pickup dispersion profile and at least one higher-order pickup dispersion profile. Different pickup dispersion profiles correspond to different propagation modes. The propagation mode is used to indicate the frequency, wave number, and wave velocity of the P-guided wave in the subsurface medium. The fundamental pickup dispersion profile is used to indicate the characteristics of the shallow layers of the subsurface medium. Higher-order pickup dispersion profiles are used to indicate structural information of deeper layers in the subsurface medium.

[0052] In some embodiments, the phase-shifting method or the FJ method (Frequency-Bessel transform) can be used to extract multi-order pickup dispersion curves of P-guided waves from seismic data. Other methods can also be used to extract multi-order pickup dispersion curves of P-guided waves from seismic data, and this application does not limit the specific methods used. The phase-shifting method is a frequency-domain based pickup technique. The FJ method is a pickup technique that combines time-domain and frequency-domain methods.

[0053] In some embodiments, the multi-order picking dispersion curve can be obtained based on seismic data of the Z component or seismic data of the Press direction component. The Z component refers to seismic data in the vertical direction. The Press direction component refers to seismic data in the ocean direction.

[0054] For example, when the seismic data is terrestrial seismic data, the phase shift method or the FJ method can be used to extract the multi-order pickup dispersion curve of the P-guided wave of the Z component from the seismic data.

[0055] For example, when the seismic data is marine data, the phase shift method or the FJ method can be used to extract the multi-order pickup dispersion curve of the P-guided wave of the Press direction component from the seismic data.

[0056] In some embodiments, the multi-stage picking dispersion curve includes at least one picking dispersion curve, each picking dispersion curve having a different starting frequency; based on the starting frequency of at least one picking dispersion curve, a model depth is determined, the model depth being used to indicate the vertical depth of the subsurface medium simulated by the initial velocity model; the maximum wave velocity in at least one picking dispersion curve is determined as the maximum wave velocity of the multi-stage picking dispersion curve; the minimum wave velocity in at least one picking dispersion curve is determined as the minimum wave velocity of the multi-stage picking dispersion curve; based on the model depth, maximum wave velocity, minimum wave velocity, and preset layer thickness, an initial velocity model is established, the preset layer thickness being used to indicate the thickness of each layer in the subsurface medium simulated by the initial velocity model.

[0057] The starting frequency of a pickup dispersion curve is the lowest frequency required to excite the transmission mode corresponding to that pickup dispersion curve. For at least one pickup dispersion curve, higher-order pickup dispersion curves require higher frequencies and energy to excite and propagate; therefore, different pickup dispersion curves will have different starting frequencies.

[0058] For example, please refer to Figure 3 It shows a schematic diagram of the multi-order pickup dispersion curve provided in the embodiments of this application. Figure 3 It includes three pickup dispersion curves, corresponding to the first-order pickup dispersion curve 31, the second-order pickup dispersion curve 32, and the third-order pickup dispersion curve 33, respectively. For example... Figure 3 It can be seen that the starting frequency of the first-order pickup dispersion curve 31 is 6, the starting frequency of the second-order pickup dispersion curve 32 is 11, and the starting frequency of the third-order pickup dispersion curve 33 is 18.

[0059] Model depth refers to the maximum vertical range of the velocity model established in seismic exploration or inversion, that is, the depth from the surface to the deepest point that the velocity model can describe.

[0060] In some embodiments, based on the starting frequency of at least one picked dispersion curve, the corresponding model depth is matched in the theoretical depth table, which indicates the starting frequency of the P-guided wave at different model depths obtained through theoretical calculation. In some embodiments, based on a preset layer thickness and the starting frequency of at least one picked dispersion curve, the corresponding model depth is matched in the model depth table. The preset layer thickness refers to the pre-defined thickness of each layer in the subsurface medium, assuming that each layer has the same thickness.

[0061] For example, Table 1 is a model depth table with a preset layer thickness of 5 meters, according to Figure 3 The starting frequencies of the three pickup dispersion curves are matched in Table 1. The first-order starting frequency represents the starting frequency of the first-order pickup dispersion curve, the second-order starting frequency represents the starting frequency of the second-order pickup dispersion curve, the third-order starting frequency represents the starting frequency of the third-order pickup dispersion curve, the fourth-order starting frequency represents the starting frequency of the fourth-order pickup dispersion curve, and the fifth-order starting frequency represents the starting frequency of the fifth-order pickup dispersion curve.

[0062] Based on the fact that the initial frequency of the first-order pickup dispersion curve 31 is 6, the initial frequency of the second-order pickup dispersion curve 32 is 11, and the initial frequency of the third-order pickup dispersion curve 33 is 18, it can be matched with a model depth of 200 meters, 250 meters, or 300 meters. Therefore, the model depth of the initial velocity model can be set to 200 meters, 250 meters, or 300 meters, and the preset layer thickness can be set to 5 meters.

[0063] Table 1 Model Depth Table

[0064]

[0065]

[0066] In some embodiments, the largest wave velocity among all the picked-up dispersion curves is determined as the maximum wave velocity of the multi-order picked-up dispersion curve. In some embodiments, the smallest wave velocity among all the picked-up dispersion curves is determined as the minimum wave velocity of the multi-order picked-up dispersion curve.

[0067] For example, such as Figure 3 As shown, the maximum wave velocity of the multi-order picking dispersion curve is 3000 m / s, and the minimum wave velocity is 1500 m / s. In order for the final velocity model to cover the maximum and minimum wave velocities of the multi-order picking dispersion curve, the velocity range of the velocity model can be set to 1350–3300 m / s.

[0068] A velocity model is used to describe the wave velocity of P-waves at different depths or locations in the subsurface medium. In some embodiments, an initial velocity model with uniformly varying wave velocity is established based on the model depth, maximum wave velocity, minimum wave velocity, and a preset layer thickness.

[0069] Based on the above example, a uniformly varying initial velocity model is established with a model depth of 300 meters and a preset layer thickness of 5 meters. That is, the initial velocity model can describe a subsurface medium of 300 meters, comprising a total of 60 layers, each with a layer thickness of 5 meters.

[0070] Based on the previous example, the initial velocity model can be established using the following formula:

[0071]

[0072] Among them, V p [i] represents the wave velocity of the P-wave in the i-th layer of the subsurface medium, V max V represents the maximum dispersion velocity, which is also the maximum wave velocity of the P-guided wave. min This represents the minimum dispersion velocity, which is also the minimum wave velocity of a P-guided wave, n. layer This represents the total number of layers in the underground medium. Where i and n... layer It is a positive integer.

[0073] Step 230: Based on the initial velocity model, perform forward modeling on the multi-order picking dispersion curve to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on the wave velocity in the first underground medium. The first underground medium is the underground medium simulated by the initial velocity model.

[0074] The purpose of forward modeling is to calculate the multi-order predicted dispersion curve corresponding to the multi-order picked dispersion curve based on the initial velocity model. During the forward modeling process, at least one depth-sensitive kernel is obtained by solving the forward modeling equations, which is used to quantify the sensitivity of the P-guided wave velocity to the model parameters of the initial velocity model.

[0075] In some embodiments, a root-search method based on dispersion functions, or a discretization solution algorithm combining finite element method with eigenvalue separation, is used to perform forward modeling on multi-order picking dispersion curves based on an initial velocity model, thereby obtaining at least one depth-sensitive core of the P-guided wave. In some embodiments, forward modeling on multi-order picking dispersion curves based on an initial velocity model also yields multi-order predicted dispersion curves.

[0076] In some embodiments, the Spectral Element Method (SEM) is used to construct the forward modeling equations. For example, the forward modeling equations are as follows:

[0077] ω 2 MV-EV=k 2 EV+kDV

[0078] Where ω is the angular frequency, V is the wave field vector of the initial velocity model, M is the global mass matrix of the initial velocity model, K is the global stiffness matrix of the initial velocity model, E is the local stiffness matrix of the initial velocity model, D is the characteristic matrix, and k is the wave number of the P-guided wave.

[0079] In some embodiments, the forward modeling equation is solved to obtain the quadratic eigenvalues ​​of the forward modeling equation; based on the quadratic eigenvalues ​​of the forward modeling equation, the phase velocity of the multi-order prediction dispersion curve is calculated.

[0080] For example, solving the forward equation of the previous embodiment yields the complex root k of the forward equation, which includes the real part Re(k) and the imaginary part Im(k). The complex root k of the forward equation is also the wavenumber, which in turn yields the dispersion curve points of the multi-order prediction dispersion curve, i.e., the phase velocity.

[0081] In some embodiments, at least one deep-sensitive kernel is obtained using the following formula.

[0082]

[0083] in, Represents the partial differential equation, λ is the Lamé coefficient, Re represents the real part, and V T It is the transpose of V, a deep-sensitive kernel: Vp is the wave velocity of a P-wave.

[0084] Step 240: If the multi-order predicted dispersion curve satisfies the first convergence condition, the initial velocity model is re-established based on the initial velocity model to obtain the reconstructed initial velocity model.

[0085] The first convergence condition indicates the convergence condition of the initial velocity model before reconstruction. That is, if the initial velocity model before reconstruction converges, the initial velocity model is rebuilt. In some embodiments, the reconstructed initial velocity model is updated in the same manner to finally obtain the inverted velocity model.

[0086] In some embodiments, the first convergence condition includes: if the initial velocity model is not the reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a first threshold; or, if the initial velocity model is not the reconstructed initial velocity model, the number of updates to the initial velocity model is less than or equal to a second threshold.

[0087] The difference between the multi-order picked dispersion curve and the multi-order predicted dispersion curve is used to indicate the fit between them. In some embodiments, the difference between the multi-order picked dispersion curve and the multi-order predicted dispersion curve can be determined by calculating the second norm of the dispersion residuals of the multi-order picked dispersion curve and the multi-order predicted dispersion curve.

[0088] The first threshold refers to the maximum number of updates to the initial velocity model before reconstruction. In some embodiments, the first threshold is preset by those skilled in the art, and this application does not limit this. The second threshold is used to determine whether the initial velocity model before reconstruction is sufficiently similar to the theoretical model. In some embodiments, the second threshold is preset by those skilled in the art.

[0089] In some embodiments, based on the initial velocity model, a velocity change rate curve is generated, which is used to describe the wave velocity variation among the first layers in the first underground medium; based on the velocity change rate curve, the first underground medium is re-layered to obtain the re-layered first underground medium, which includes N second layers, where N is a positive integer; based on the initial velocity model, the average wave velocity of each of the N second layers is calculated; based on the re-layered first underground medium and the average wave velocity of each of the N second layers, the initial velocity model is re-established to obtain the reconstructed initial velocity model.

[0090] The first stratification refers to the stratification within the first subsurface medium before the initial velocity model is reconstructed. The second stratification refers to the stratification within the first subsurface medium after the initial velocity model is reconstructed.

[0091] In some embodiments, the initial velocity model before the last updated reconstruction is analyzed to calculate the rate of change of wave velocity between any two adjacent first layers in the first underground medium; based on the rate of change of wave velocity, a velocity change rate curve is plotted.

[0092] For example, for any two adjacent first layers in the first underground medium, assuming these two first layers are Vp[j] and Vp[j+1] respectively, the wave velocity change rate of these two first layers is:

[0093]

[0094] Where j is a positive integer.

[0095] In some embodiments, the velocity change rate curve includes N velocity change rate peaks; based on the velocity change rate curve, the position of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium is determined; based on the position of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium, the first underground medium is re-layered to obtain the re-layered first underground medium; wherein, in the first underground medium, the position of each second layer corresponds to the position of a first layer corresponding to a velocity change rate peak.

[0096] The peak velocity change rate refers to the rate of change of wave velocity corresponding to the peak on the velocity change rate curve. The first layer corresponding to the peak velocity change rate refers to the lower of the two first layers used to calculate the peak velocity change rate.

[0097] In some embodiments, for any of the wave velocity change rates mentioned above, two first layers corresponding to the peak velocity change rate are determined; the first layer located in the lower layer of the two first layers is determined as the second layer.

[0098] In other words, on the velocity change rate curve, each velocity change rate peak corresponds to a second stratum. The position of the first stratum within the first underground medium is defined as the position of the second stratum within the first underground medium. The position of the first stratum within the first underground medium refers to the position of its lower boundary line. The position of the second stratum within the first underground medium refers to the position of its lower boundary line.

[0099] In some embodiments, the thickness of each of the N second layers is determined based on the position of the lower boundary line of the N second layers. In some embodiments, the thickness of the first second layer is determined based on the number of first layers included between the ground and the position of the lower boundary line of the first second layer, and a preset layer thickness; for any second layer after the first second layer, the thickness of the second layer is determined based on the number of first layers between the lower boundary line of the previous second layer and the lower boundary line of the second layer, and a preset layer thickness.

[0100] In some embodiments, for any one of the N second layers, the average wave velocity of the first layer included in the second layer is calculated; the average wave velocity of the first layer included in the second layer is determined as the wave velocity of the second layer.

[0101] For example, please refer to Figure 4 It shows a schematic diagram of a velocity change rate curve provided in one embodiment of this application. Figure 4The velocity change rate curve shown includes four peaks, namely peak 41, peak 42, peak 43 and peak 44. Since peak 41 is the location where the simulated seismic wave was recently excited in the first subsurface medium, and the seismic wave has unstable characteristics, peak 41 is not used to re-layer the first subsurface medium. Instead, the subsequent peaks 42, 43 and 44 are used to re-layer the first subsurface medium.

[0102] Step 250: If the multi-order predicted dispersion curve satisfies the second convergence condition, the initial velocity model is determined as the inverted velocity model.

[0103] The second convergence condition indicates the convergence condition of the reconstructed initial velocity model. That is, if the reconstructed initial velocity model converges, it is determined as the final velocity model. In some embodiments, analyzing the inverted velocity model yields the stratification of the subsurface medium.

[0104] In some embodiments, the second convergence condition includes: when the initial velocity model is a reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a third threshold; or, when the initial velocity model is a reconstructed initial velocity model, the number of updates to the initial velocity model is less than or equal to a fourth threshold.

[0105] The third threshold refers to the maximum number of times the reconstructed initial velocity model can be updated. In some embodiments, the third threshold is preset by those skilled in the art, and this application does not limit this. The fourth threshold is used to determine whether the reconstructed initial velocity model is sufficiently similar to the theoretical model. In some embodiments, the fourth threshold is preset by those skilled in the art.

[0106] Step 260: Based on at least one deep sensitive kernel, inversion equations are used to perform inversion to obtain the update amount of the initial velocity model. Based on the update amount of the initial velocity model, the initial velocity model is adjusted to obtain the initial velocity model. The inversion equations are the relationships between at least one deep sensitive kernel, the update amount of the initial velocity model, the multi-order picking dispersion curve of the P-guided wave, and the multi-order predicted dispersion curve of the P-guided wave. The multi-order predicted dispersion curve is generated based on the initial velocity model.

[0107] In some embodiments, the inversion equation is as follows:

[0108] (G T G+L)δVp=G T δd

[0109] Where G is the depth-sensitive kernel matrix, G THere, δVp is the transpose of G, L is the damping coefficient, δD is the update amount of the initial velocity model, and δd is the difference between the multi-order predicted dispersion curve and the multi-order picked dispersion curve. The deep sensitive kernel matrix includes at least one of the above-mentioned deep sensitive kernels. In some embodiments, the above-mentioned inversion model is solved to obtain the update amount δVp of the initial velocity model; the update amount δVp of the initial velocity model is added to the initial velocity model to obtain the initial velocity model. For example, the initial velocity model V p [i] is as follows:

[0110] V p ′[i]=V p [i]+δVp

[0111] Step 270: The process begins again with the forward modeling of the multi-order pickup dispersion curve based on the initial velocity model.

[0112] In summary, the method provided in this application obtains the initial velocity model after the first round of inversion by picking up the dispersion curves of the P-guided wave through multiple forward modeling in the first round. Then, it analyzes the initial velocity model to re-layer it. Based on the re-layered initial velocity model, a new initial velocity model is established. Next, the above steps are performed again using this reconstructed initial velocity model to obtain the final inverted velocity model. This method reduces the inversion parameters required for inversion by re-classifying the initial velocity model, thus improving the agreement between the final velocity model and the theoretical model.

[0113] The complete process of the velocity model inversion method is described below.

[0114] Please refer to Figure 5 The diagram illustrates a flowchart of an inversion method for a velocity model provided in another embodiment of this application.

[0115] The velocity model inversion method can be divided into two parts: the initial inversion part and the reconstruction inversion part. It is assumed that the initial velocity model is based on a model depth of 300 meters and a preset layer thickness of 5 meters.

[0116] (I) First Inversion Section

[0117] (1) The multi-order pickup dispersion curves of the P-guided wave were obtained from the shot gather record as follows: Figure 3 As shown.

[0118] (2) Based on the initial velocity model, the multi-order picking dispersion curves are forward modeled to obtain the multi-order predicted dispersion curves of the P-guided wave.

[0119] (3) Calculate the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve.

[0120] (4) Determine whether the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than the first threshold. If not, proceed to step (5); otherwise, proceed to step (8).

[0121] (5) Inversion is performed based on at least one deep sensitive kernel to obtain the update amount of the initial velocity model.

[0122] (6) Based on the update amount of the initial velocity model, the initial velocity model is adjusted to obtain the updated initial velocity model.

[0123] (7) Start from step (2) again.

[0124] (8) Obtain the initial velocity model for the first time, and perform step (1) of (II) reconstructing the inversion part.

[0125] Please refer to Figure 6 It shows a schematic diagram of a multi-order prediction dispersion curve provided in one embodiment of this application. Figure 6 The curves corresponding to the asterisk symbol are three prediction dispersion curves in the multi-order prediction dispersion curves, while the curves corresponding to the hollow circle symbol are multiple dispersion curves in the multi-order picking dispersion curves. Figure 6 The multi-order prediction dispersion curve shown is the multi-order prediction dispersion curve obtained after 20 inversion iterations. Figure 6 It can be seen that after 20 inversion iterations, the multi-order predicted dispersion curve obtained from the forward modeling based on the initial velocity model fits well.

[0126] Please refer to Figure 7 It shows a schematic diagram of an initial velocity model provided in one embodiment of this application. Figure 7 It includes three velocity models: theoretical velocity model 71, inverted initial velocity model 72, and initial velocity model 73. Figure 7 It can be seen that the initial velocity model obtained after 20 inversion iterations differs from the theoretical velocity model.

[0127] Therefore, re-establishing the initial velocity model and continuing the inversion iteration can improve the agreement between the final velocity model and the theoretical velocity model.

[0128] (II) Reconstruction and Inversion Section

[0129] (1) Based on the initial initial velocity model, a new initial velocity model is established.

[0130] according to Figure 4The velocity change rate curves shown illustrate the re-stratification of the initial velocity model, leading to a new initial velocity model. The first second layer of the first underground medium in the new initial velocity model is 40 meters deep, with the average velocity of the first 8 layers being 1476 m / s; the second second layer is 35 meters deep, with the average velocity of the first 9-15 layers being 1728 m / s; the third second layer is 85 meters deep, with the average velocity of the first 16-32 layers being 2270 m / s; and the fourth second layer is the wireless half-space, with the average velocity of the first 33-61 layers being 2804 m / s.

[0131] (2) Based on the initial velocity model, the multi-order picking dispersion curves are forward modeled to obtain the multi-order predicted dispersion curves of the P-guided wave.

[0132] (3) Calculate the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve.

[0133] (4) Determine whether the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than the first threshold. If not, proceed to step (5); otherwise, proceed to step (8).

[0134] (5) Inversion is performed based on at least one deep sensitive kernel to obtain the update amount of the initial velocity model.

[0135] (6) Based on the update amount of the initial velocity model, the initial velocity model is adjusted to obtain the updated initial velocity model.

[0136] (7) Start from step (2) again.

[0137] (8) Obtain the final initial velocity model.

[0138] Please refer to Figure 8 It shows a schematic diagram of a multi-order prediction dispersion curve provided in another embodiment of this application. Figure 8 The curves corresponding to the asterisk symbol are three prediction dispersion curves in the multi-order prediction dispersion curves, while the curves corresponding to the hollow circle symbol are multiple dispersion curves in the multi-order picking dispersion curves. Figure 8 The multi-order prediction dispersion curve shown is the multi-order prediction dispersion curve obtained after 20 inversion iterations following reconstruction. Figure 8 It can be seen that after 20 inversion iterations, the multi-order predicted dispersion curve obtained from the forward modeling based on the initial velocity model fits well.

[0139] Please refer to Figure 9 This illustrates a schematic diagram of an initial velocity model provided in another embodiment of this application. Figure 9It includes three velocity models: theoretical velocity model 91, reconstructed initial velocity model 92, and reconstructed and inverted initial velocity model 93. Figure 9 It can be seen that after 20 inversion iterations, the reconstructed initial velocity model is quite close to the theoretical velocity model.

[0140] Please refer to Figure 10 It shows a schematic diagram of the conformity change curve provided in one embodiment of this application, by Figure 10 It can be seen that after 20 inversion iterations, the reconstructed initial velocity model differs from the theoretical velocity model by less than 3%, indicating that the reconstructed initial velocity model and the theoretical velocity model match very well after 20 inversion iterations.

[0141] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0142] Please refer to Figure 11 This diagram illustrates a block diagram of a velocity model inversion apparatus according to an embodiment of this application. The apparatus has the functions described above, which can be implemented in hardware or by hardware executing corresponding software. The apparatus can be the computer device 10 described above, or it can be located within the computer device 10. Figure 11 As shown, the device 1100 may include: an acquisition module 1110, a generation module 1120, a forward modeling module 1130, a reconstruction module 1140, a determination module 1150, an inversion module 1160, and an execution module 1170.

[0143] The acquisition module 1110 is used to acquire earthquake data, which is data collected after seismic waves propagate in the underground medium.

[0144] The generation module 1120 is used to generate a multi-order picking dispersion curve of the P-guided wave based on the seismic data. The multi-order picking dispersion curve is used to describe the relationship between the wave velocity of the P-guided wave and the frequency.

[0145] The forward modeling module 1130 is used to perform forward modeling on the multi-order picking dispersion curve based on the initial velocity model to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on the wave velocity in the first underground medium, which is the underground medium simulated by the initial velocity model.

[0146] The reconstruction module 1140 is used to re-establish the initial velocity model based on the initial velocity model, and obtain the reconstructed initial velocity model, provided that the multi-order predicted dispersion curve satisfies the first convergence condition.

[0147] The determination module 1150 is used to determine the initial velocity model as the inverted velocity model when the multi-order predicted dispersion curve satisfies the second convergence condition.

[0148] The inversion module 1160 is used to perform inversion based on the at least one depth-sensitive kernel using an inversion equation to obtain the update amount of the initial velocity model, and to adjust the initial velocity model based on the update amount of the initial velocity model to obtain an updated initial velocity model; wherein, the inversion equation is a relationship between the at least one depth-sensitive kernel, the update amount of the initial velocity model, the multi-order pickup dispersion curve of the P-guided wave and the multi-order prediction dispersion curve of the P-guided wave.

[0149] The execution module 1170 is used to start execution again from the step of performing forward modeling of the multi-order picking dispersion curve based on the initial velocity model.

[0150] In some embodiments, the reconstruction module 1140 is configured to generate a velocity change rate curve based on the initial velocity model, the velocity change rate curve describing the wave velocity variation among the first layers in the first underground medium; based on the velocity change rate curve, re-layer the first underground medium to obtain a re-layered first underground medium, the re-layered first underground medium including N second layers, where N is a positive integer; calculate the average wave velocity of each of the N second layers based on the initial velocity model; and re-establish the initial velocity model based on the re-layered first underground medium and the average wave velocity of each of the N second layers to obtain the reconstructed initial velocity model.

[0151] In some embodiments, the velocity change rate curve includes N velocity change rate peaks; the reconstruction module 1140 is further configured to determine, based on the velocity change rate curve, the position of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium; and based on the position of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium, re-layer the first underground medium to obtain the re-layered first underground medium; wherein, in the first underground medium, the position of each second layer corresponds to the position of a first layer corresponding to a velocity change rate peak.

[0152] In some embodiments, the multi-order pickup dispersion curves include at least one pickup dispersion curve, and the starting frequencies of each pickup dispersion curve are different; the device 1100 further includes a modeling module (in Figure 11 (Not shown in the image) is used to determine the model depth based on the starting frequency of the at least one picked dispersion curve, the model depth indicating the vertical depth of the subsurface medium simulated by the initial velocity model; the maximum wave velocity in the at least one picked dispersion curve is determined as the maximum wave velocity of the multi-order picked dispersion curve; the minimum wave velocity in the at least one picked dispersion curve is determined as the minimum wave velocity of the multi-order picked dispersion curve; the initial velocity model is established based on the model depth, the maximum wave velocity, the minimum wave velocity, and a preset layer thickness, the preset layer thickness indicating the thickness of each layer in the subsurface medium simulated by the initial velocity model.

[0153] In some embodiments, the first convergence condition includes: when the initial velocity model is not the reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a first threshold; or, when the initial velocity model is not the reconstructed initial velocity model, the number of updates to the initial velocity model is less than or equal to a second threshold.

[0154] In some embodiments, the second convergence condition includes: when the initial velocity model is a reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a third threshold; or, when the initial velocity model is a reconstructed initial velocity model, the number of updates to the initial velocity model is less than or equal to a fourth threshold.

[0155] In summary, the method provided in this application obtains the initial velocity model after the first round of inversion by picking up the dispersion curves of the P-guided wave through multiple forward modeling in the first round. Then, it analyzes the initial velocity model to re-layer it. Based on the re-layered initial velocity model, a new initial velocity model is established. Next, the above steps are performed again using this reconstructed initial velocity model to obtain the final inverted velocity model. This method reduces the inversion parameters required for inversion by re-classifying the initial velocity model, thus improving the agreement between the final velocity model and the theoretical model.

[0156] It should be noted that the apparatus provided in the above embodiments is only illustrated by the division of the above functional modules when implementing its functions. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0157] Please refer to Figure 12This diagram illustrates a structural block diagram of a computer device 1200 provided in one embodiment of this application. The computer device 1200 may be... Figure 1 The terminal device 10 in the implementation environment shown can also be Figure 1 Server 20 in the implementation environment shown is used to implement the speed model inversion method provided in the above embodiments. Specifically:

[0158] Typically, computer device 1200 includes a processor 1210 and a memory 1220.

[0159] Processor 1210 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 1210 may be implemented using at least one hardware form selected from Digital Signal Processing (DSP), Field Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). Processor 1210 may also include a main processor and a coprocessor. The main processor, also known as the Central Processing Unit (CPU), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 1210 may integrate a Graphics Processing Unit (GPU), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 1210 may also include an AI processor for handling computational operations related to machine learning.

[0160] The memory 1220 may include one or more computer-readable storage media, which may be non-transitory. The memory 1220 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 1220 are used to store a computer program configured to be executed by one or more processors to implement the speed model inversion method described above.

[0161] Those skilled in the art will understand that Figure 12 The structure shown does not constitute a limitation on the computer device 1200, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0162] In an exemplary embodiment, a computer-readable storage medium is also provided, wherein a computer program is stored in the storage medium, and the computer program, when executed by a processor, implements the speed model inversion method described above. Optionally, the computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), solid-state drives (SSDs), or optical discs, etc. The random access memory may include resistive random access memory (ReRAM) and dynamic random access memory (DRAM).

[0163] In an exemplary embodiment, a computer program product is also provided, the computer program product including a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium, and the processor executes the computer program, causing the computer device to perform the speed model inversion method described above.

[0164] It should be understood that "multiple" as used herein refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, the step numbers described herein are merely illustrative of one possible execution order. In some other embodiments, the steps may not be executed in numerical order, such as two steps with different numbers being executed simultaneously, or two steps with different numbers being executed in the reverse order of the illustration. This application does not limit this.

[0165] The above description is merely an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for inverting a velocity model, characterized in that, The method includes: Acquire earthquake data, which is data collected after seismic waves propagate through the subsurface medium; Based on the earthquake data, a multi-order picking dispersion curve of the P-guided wave is generated, which is used to describe the relationship between the wave velocity of the P-guided wave and the frequency. Based on the initial velocity model, the multi-order picking dispersion curve is forward modeled to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on wave velocity in the first underground medium, which is the underground medium simulated by the initial velocity model. If the multi-order predicted dispersion curve satisfies the first convergence condition, the initial velocity model is re-established based on the initial velocity model to obtain the reconstructed initial velocity model. If the multi-order predicted dispersion curve satisfies the second convergence condition, the initial velocity model is determined as the inverted velocity model. Based on the at least one depth-sensitive kernel, an inversion equation is used to perform inversion and obtain the update amount of the initial velocity model. Based on the update amount of the initial velocity model, the initial velocity model is adjusted to obtain the updated initial velocity model. The inversion equation is a relationship between the at least one depth-sensitive kernel, the update amount of the initial velocity model, the multi-order pickup dispersion curve of the P-guided wave and the multi-order prediction dispersion curve of the P-guided wave. The process begins again from the step of performing forward modeling on the multi-order pickup dispersion curve based on the initial velocity model.

2. The method according to claim 1, characterized in that, The process of reconstructing the initial velocity model based on the initial velocity model to obtain the reconstructed initial velocity model includes: Based on the initial velocity model, a velocity change rate curve is generated, which is used to describe the wave velocity change among the first layers in the first underground medium. Based on the velocity change rate curve, the first underground medium is re-layered to obtain a re-layered first underground medium, which includes N second layers, where N is a positive integer. Based on the initial velocity model, the average wave velocity of each of the N second layers is calculated respectively; Based on the average wave velocity of the first subsurface medium after re-stratification and the N second layers, the initial velocity model is re-established to obtain the reconstructed initial velocity model.

3. The method according to claim 2, characterized in that, The velocity change rate curve includes N velocity change rate peaks; The process of re-stratifying the first underground medium based on the velocity change rate curve to obtain the re-stratified first underground medium includes: Based on the velocity change rate curve, determine the location of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium; Based on the position of the first layer corresponding to each of the N velocity change rate peaks in the first underground medium, the first underground medium is re-layered to obtain the re-layered first underground medium. In the first underground medium, the location of each second layer corresponds to the location of a first layer corresponding to a peak value of the rate of change of velocity.

4. The method according to claim 1, characterized in that, The multi-order pickup dispersion curve includes at least one pickup dispersion curve, and the starting frequencies of each pickup dispersion curve are different; the method further includes: Based on the starting frequency of the at least one picked dispersion curve, the model depth is determined, which indicates the vertical depth of the subsurface medium simulated by the initial velocity model. The maximum wave velocity in the at least one pickup dispersion curve is determined as the maximum wave velocity of the multi-order pickup dispersion curve. The minimum wave velocity in the at least one pickup dispersion curve is determined as the minimum wave velocity of the multi-order pickup dispersion curve. Based on the model depth, the maximum wave velocity, the minimum wave velocity, and the preset layer thickness, the initial velocity model is established, whereby the preset layer thickness is used to indicate the thickness of each layer in the underground medium simulated by the initial velocity model.

5. The method according to claim 1, characterized in that, The first convergence condition includes: if the initial velocity model is not the reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a first threshold; or, if the initial velocity model is not the reconstructed initial velocity model, the number of updates of the initial velocity model is less than or equal to a second threshold.

6. The method according to claim 1, characterized in that, The second convergence condition includes: if the initial velocity model is a reconstructed initial velocity model, the difference between the multi-order picking dispersion curve and the multi-order prediction dispersion curve is less than or equal to a third threshold; or, if the initial velocity model is a reconstructed initial velocity model, the number of updates to the initial velocity model is less than or equal to a fourth threshold.

7. An inversion device for a velocity model, characterized in that, The device includes: The acquisition module is used to acquire earthquake data, which is data collected after seismic waves propagate in the underground medium; The generation module is used to generate a multi-order picking dispersion curve of the P-guided wave based on the seismic data. The multi-order picking dispersion curve is used to describe the relationship between the wave velocity of the P-guided wave and the frequency. The forward modeling module is used to perform forward modeling on the multi-order picking dispersion curve based on the initial velocity model to obtain the multi-order predicted dispersion curve and at least one depth-sensitive kernel of the P-guided wave. The initial velocity model is established based on the maximum and minimum wave velocities of the multi-order picking dispersion curve. The depth-sensitive kernel is used to indicate the degree of influence of different depths on the wave velocity in the first underground medium, which is the underground medium simulated by the initial velocity model. The reconstruction module is used to re-establish the initial velocity model based on the initial velocity model, and obtain the reconstructed initial velocity model, provided that the multi-order predicted dispersion curve satisfies the first convergence condition. The determination module is used to determine the initial velocity model as the inverted velocity model when the multi-order predicted dispersion curve satisfies the second convergence condition. The inversion module is used to perform inversion based on the at least one depth-sensitive kernel using an inversion equation to obtain the update amount of the initial velocity model, and to adjust the initial velocity model based on the update amount of the initial velocity model to obtain an updated initial velocity model; wherein, the inversion equation is a relationship between the at least one depth-sensitive kernel, the update amount of the initial velocity model, the multi-order pickup dispersion curve of the P-guided wave and the multi-order prediction dispersion curve of the P-guided wave; The execution module is used to start execution again from the step of forward modeling the multi-order picking dispersion curve based on the initial velocity model.

8. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing a computer program that is loaded and executed by the processor to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program that is executed by a processor to implement the method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that is loaded and executed by a processor to implement the method as described in any one of claims 1 to 6.

Images