A skeleton inversion method and system

Abstract

The present application discloses a skeleton inversion method and system, including: respectively analyzing the observed and simulated CSG seismic reflection wave data CSG to obtain the observed and simulated seismic reflection wave data velocity spectrum, and the observed and simulated data Semblance curve; recording the difference square sum of the observed data Semblance curve and the simulated data Semblance curve as the objective function, and using the velocity gradient calculation formula to obtain the velocity gradient; the velocity gradient calculation formula is obtained according to the link function and hidden variable function theory, and the link function is the cross-correlation of the observed and simulated seismic reflection wave data velocity spectrum; iteratively updating the initial velocity model according to the velocity gradient until the value of the objective function is lower than the set threshold, and obtaining the final velocity model. Through the present application, effective and accurate inversion modeling of underground deep background velocity information based on seismic reflection wave data is achieved.

Description

Technical Field

[0001] The present application relates to the field of exploration geophysical technology, and in particular to a skeleton inversion method and system. Background Art

[0002] Seismic depth domain velocity inversion modeling is a key technology in the field of geophysical exploration. It infers the velocity distribution of underground structures by analyzing the propagation characteristics of seismic waves in underground media. The importance of this technology lies in its ability to provide high-resolution underground velocity models, which are essential for accurate seismic imaging, effective resource exploration, and a deep understanding of geological structures. Depth domain velocity models not only help geologists identify potential oil and gas reservoirs and other mineral resources, but can also be used for geological hazard assessment and monitoring, such as seismicity and fault characteristics analysis. In addition, accurate velocity models play a significant role in improving the accuracy of seismic data interpretation, optimizing exploration costs, and reducing development risks, and are a bridge between seismic data acquisition and geological interpretation.

[0003] In traditional technology, seismic depth domain velocity inversion modeling can be roughly divided into the following methods:

[0004] (1) Velocity analysis: Velocity analysis uses the time-distance curve formula to perform velocity scanning on the common center point data set to obtain a velocity spectrum with the horizontal and vertical coordinates being the stacking velocity and the two-way travel time, respectively. By picking the optimal stacking velocity at each time point on the velocity spectrum, a stacking velocity curve that changes with time is finally obtained. By repeating the above work for each common center point, a time domain velocity model can finally be obtained. The time domain velocity model can then be converted into a depth domain velocity model using the Dix formula. Although this method can use reflection wave data to invert the deep medium model, the problem is that the Dix formula has a horizontal layered model assumption, which is difficult to meet the actual geological conditions. Therefore, the depth domain velocity model converted using the Dix formula has low accuracy and is usually not directly used for geological interpretation and offset imaging.

[0005] (2) Tomographic inversion: Tomographic inversion uses the high-frequency approximate equation of the wave equation, namely the eikonal equation, to invert the travel time information of seismic data. Currently developed methods include using the travel time of the first arrival wave to invert the near-surface model, and using reflection wave data to invert the deep velocity model. Although tomographic inversion has high computational efficiency and can use reflection wave data to invert the deep medium model, its problem is that it has high-frequency approximate assumptions and is difficult to accurately describe the complex wave field in complex media. Therefore, to accurately describe the wave field propagation characteristics in complex media, it is necessary to return to solving the wave equation.

[0006] (3) Full waveform inversion: Full waveform inversion is a representative wave equation inversion method. It can theoretically achieve an inversion resolution of one quarter wavelength and is currently the inversion method with the highest resolution. Since traditional full waveform inversion is insufficient to update the deep background velocity, the corresponding reflection wave full waveform inversion has been derived to effectively construct the deep background velocity information. Full waveform inversion can also use reflection wave data to invert the deep medium model, but no matter which method is used, since the objective function takes the form of the square of the difference between the observed data and the simulated data, the inversion objective function is highly nonlinear and contains many local minima. Therefore, if the initial inversion model (background model) deviates far from the true model, the inversion is very likely to converge to the local minimum, which is the biggest problem of waveform inversion.

[0007] (4) Skeleton inversion: Skeleton inversion is an inversion method based on the wave equation inversion kernel (i.e., full waveform inversion kernel) that inverts the skeleton information of seismic data rather than the waveform information. The advantage of this method is that the objective function has better convexity, and the inversion is easier to converge to the vicinity of the global optimal solution than waveform inversion. However, its limitation is that it can only restore the background model information of the underground medium. Since 1991, a series of skeleton inversion methods have been developed based on the skeleton inversion framework, involving skeleton information such as travel time, frequency shift, and dispersion. However, the problem is that the currently published skeleton inversion work mainly focuses on using shallow seismic data (such as first arrival waves, surface waves, etc.) to invert shallow structures, and there is no method involving deep background velocity modeling using reflection wave data.

[0008] In summary, how to obtain a method that can use seismic reflection wave data to accurately construct a deep underground velocity model has become an urgent problem to be solved. Summary of the invention

[0009] Based on this, in response to the above technical problems, a skeleton inversion method and system are provided to solve the problem that the existing technology cannot use seismic reflection wave data to construct a deep underground velocity model, or the deep underground velocity model constructed using seismic reflection wave data has low accuracy.

[0010] In a first aspect, a skeleton inversion method is provided, the method comprising:

[0011] Performing velocity analysis on the observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data to obtain a velocity spectrum of the observed seismic reflection wave data and a velocity spectrum of the simulated seismic reflection wave data, and obtaining an observed data Semblance curve and a simulated data Semblance curve according to the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data by automatic picking technology;

[0012] The sum of squares of the difference between the observed data Semblance curve and the simulated data Semblance curve is recorded as the objective function;

[0013] The forward and reverse wave fields are calculated by using the observed CSG seismic reflection wave data and the positions of the shot points and the detection points, and then the velocity gradient is obtained by using the velocity gradient calculation formula according to the positions of the shot points and the detection points and the corresponding forward and reverse wave fields; the velocity gradient calculation formula is obtained according to the link function and hidden variable function theory, and the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity;

[0014] The current initial velocity model is iteratively updated using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then the iterative update is stopped to obtain the final velocity model.

[0015] In the above scheme, optionally, the simulated CSG data is generated by calculating according to the initial velocity model using wave equation simulation.

[0016] In the above scheme, optionally, performing velocity analysis on the observed CSG data and the simulated CSG data to obtain the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data respectively comprises:

[0017] The observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data are respectively converted into the observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data by using the extraction channel set;

[0018] The observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data are respectively subjected to velocity scanning using the time-distance curve to generate corresponding observed seismic reflection wave data velocity spectra and simulated seismic reflection wave data velocity spectra.

[0019] In the above scheme, optionally, the iterative update algorithm includes steepest descent or conjugate gradient.

[0020] In the above scheme, optionally, the set threshold value is 30% of the initial error value; the initial error is obtained by comparing the difference between the data predicted by the initial model and the actual observed data.

[0021] In a second aspect, a skeleton inversion system is provided, the system comprising:

[0022] Data Semblance curve extraction module: used for performing velocity analysis on observed CSG seismic reflection wave data and simulated CSG seismic reflection wave data to obtain observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum, and obtaining observed data Semblance curve and simulated data Semblance curve according to the observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum respectively;

[0023] Objective function construction module: used to record the sum of squares of the difference between the observed data Semblance curve and the simulated data Semblance curve as the objective function;

[0024] The velocity gradient calculation module is used to calculate the forward and reverse wave fields using the CSG seismic reflection wave data and the positions of the shot points and the detection points, and then obtain the velocity gradient using the velocity gradient calculation formula according to the positions of the shot points and the detection points and the corresponding forward and reverse wave fields; the velocity gradient calculation formula is obtained based on the hidden variable function theory according to the link function, and the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity;

[0025] Velocity model iterative generation module: used to iteratively update the current initial velocity model using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then stop the iterative update and obtain the final velocity model.

[0026] In the above scheme, optionally, performing velocity analysis on the observed CSG data and the simulated CSG data to obtain the observed seismic reflection wave data velocity spectrum and the simulated seismic reflection wave data velocity spectrum respectively comprises:

[0027] The observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data are respectively converted into the observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data by using the extraction channel set;

[0028] The observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data are respectively subjected to velocity scanning using the time-distance curve to generate corresponding observed seismic reflection wave data velocity spectra and simulated seismic reflection wave data velocity spectra.

[0029] In a third aspect, a computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a skeleton inversion method described in the first aspect when executing the computer program.

[0030] In a fourth aspect, a computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps of a skeleton inversion method described in the first aspect.

[0031] This application has at least the following beneficial effects:

[0032] This application uses the square of the difference between the Semblance curve of observed seismic reflection wave data and the Semblance curve of simulated seismic reflection wave data as a new objective function, and then obtains the velocity gradient based on the wave equation inversion, and iterates the initial velocity model according to the velocity gradient to obtain the final velocity model. In the process of this application, since the square of the difference between the Semblance curve (superposition velocity curve) of the observed and simulated reflection wave data is used as a new objective function, the objective function has better convexity, so this technology can quickly converge to the vicinity of the global minimum. At the same time, the use of wave equation inversion avoids the high-frequency approximate assumptions in the tomography method, and when the underground structure is complex, the inversion accuracy is high; and for seismic reflection wave data, effective inversion modeling of deep underground background velocity information can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of a skeleton inversion method provided in one embodiment of the present application;

[0034] Figure 2 A schematic diagram of a complete process of a skeleton inversion method provided for one embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0036] Wave equation Semblance curve inversion is a velocity inversion method based on the skeleton inversion framework. The skeleton inversion framework originated from the wave equation travel time inversion proposed by Yi Luo and Jerry Schuster in 1991. The core idea of ​​skeleton inversion is to use the kernel of wave equation inversion to invert the skeleton information of seismic data, such as the travel time of seismic waves (Luo and Schuster, 1991), peak frequency (Dutta and Schuster, 2016) and dispersion (Li and Schuster, 2016). The original intention of the development of skeleton inversion is to improve the cycle skipping problem of full waveform inversion. Since the full waveform inversion method attempts to minimize the waveform difference between the model data and the observed data in the objective function, the inversion is highly nonlinear and easy to converge to the local extreme value. The skeleton inversion can effectively improve the cycle skipping problem of the inversion by inverting the skeleton information of the seismic data, and can better converge to the global optimal solution. The key step to realize skeleton inversion is how to define a suitable link function to realize the calculation of the derivative of the skeleton information to the elastic parameter, that is, the gradient. Different skeleton inversion methods use different seismic skeleton information, and the corresponding link functions and inverted elastic parameters are also different.

[0037] At present, skeleton inversion mainly uses seismic first arrival waves or surface waves to invert P-wave velocity and S-wave velocity respectively, and is mainly used for inversion modeling of near-surface models. At present, there is no skeleton inversion method that can use seismic reflection wave data to update deep underground medium information. In addition to skeleton inversion, full waveform inversion and reflection wave waveform inversion can both use reflection wave data to invert deep medium models, but the problem of the former is that the inversion and update capability of low-wavenumber information is poor, and it is easy to converge to local minimum values. Although the latter (reflection wave waveform inversion) can effectively restore the low-wavenumber information of deep underground media, it also cannot get rid of the problem of easy convergence to local extreme values.

[0038] In addition to wave equation inversion, velocity analysis + Dix formula in the standard process of seismic data processing is also a method of using seismic reflection wave data to construct a deep underground velocity model. Velocity analysis is a method of converting the reflection wave data in the CMP gather into a Semblance spectrum using the time-distance curve formula, and then performing velocity picking on the spectrum to obtain a stacking velocity model. Although this velocity model is a time domain velocity model, it can be converted into a depth domain velocity model through the Dix formula. The problem with this method is that the velocity conversion method based on the Dix formula has a layered model assumption, which is usually difficult to meet in practice, so the accuracy of the converted layer velocity model is very low.

[0039] It is not difficult to see from the above that whether it is the velocity analysis method, full waveform inversion method or existing skeleton inversion method in traditional seismic data processing, it is impossible to accurately construct an underground deep velocity model using seismic reflection wave data. Therefore, the method in this application is produced to solve this problem. It should be noted that this application does not limit the wave equation used, which can be an acoustic wave equation, an elastic wave equation, a viscous acoustic wave equation, a viscous elastic wave method, an acoustic anisotropy equation, etc.

[0040] The innovation of this application is to use the Semblance curve (stacked velocity curve) obtained by velocity analysis as the skeleton information, and propose a new skeleton inversion method based on the skeleton inversion framework to use seismic reflection wave data to invert the underground deep background velocity model. This method can replace the Dix formula and convert the time domain velocity model into a depth domain velocity model. The conversion result is greatly improved in accuracy compared to the Dix formula method, and can directly serve geological media, offset imaging, and can be used as the initial model for waveform inversion.

[0041] In one embodiment, Figure 1 As shown, a skeleton inversion method is provided, the method comprising:

[0042] Step S1: performing velocity analysis on the observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data to obtain a velocity spectrum of the observed seismic reflection wave data and a velocity spectrum of the simulated seismic reflection wave data, and obtaining an observed data Semblance curve and a simulated data Semblance curve according to the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data by automatic picking technology;

[0043] Step S2: The sum of squares of the differences between the observed data Semblance curve and the simulated data Semblance curve is recorded as the objective function;

[0044] Step S3: using the observed CSG seismic reflection wave data and the positions of the shot points and the detection points to calculate the forward and reverse propagation wave fields, and then using the velocity gradient calculation formula to obtain the velocity gradient according to the positions of the shot points and the detection points and the corresponding forward and reverse propagation wave fields; the velocity gradient calculation formula is obtained according to the link function and hidden variable function theory, the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity;

[0045] Step S4: Iteratively update the current initial velocity model using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then stop the iterative update and obtain the final velocity model.

[0046] Specifically, in step S2, the objective function of the wave equation Semblance curve inversion is defined as the square of the Semblance curve difference between the observed data and the simulated data:

[0047]

[0048] Where t0 represents the round-trip travel time, and They are the Semblance curves (stacked velocity curves) of the observed data and the simulated data respectively. The Semblance curve can be obtained by first performing velocity analysis on the common center point gather to obtain the velocity spectrum, and then stacking the velocity spectrum to pick the velocity.

[0049] In step S3, the derivative of the objective function with respect to the velocity c is calculated, and the gradient of the velocity can be obtained:

[0050]

[0051] in, because It cannot be directly calculated, so it is necessary to define a link function, that is, the observed data velocity spectrum S obs The speed spectrum S of the simulated data pred The cross-correlation of:

[0052] F=∫S obs (V nm0 (t0)+δV nmo (t0))S pred (V nm0 (t0))dV nmo , (3)

[0053] And when When δV nmo =ΔV nm0 , at this time the cross-correlation result of formula (3) reaches the maximum value:

[0054]

[0055] In order to simplify the derivation, the subscripts of t0 and nmo are dropped in the subsequent derivation. According to implicit function theory, the derivative in equation (2) can be expressed as follows:

[0056]

[0057] in:

[0058]

[0059] The result of formula (5.1) is a constant term E. Substituting formula (5), (5.1) and (5.2) into formula (2), we can get:

[0060]

[0061] because Therefore, the derivative terms on the right side of the equal sign in formula (6) can be expressed as:

[0062]

[0063] If Then formula (6) can be written as:

[0064]

[0065] Assume that when h i ≠h j hour, and If the product of any two terms in is close to 0, then formula (7) can be further simplified as:

[0066]

[0067] The stacking velocity difference ΔV is approximated by the time difference Δt(h) at each offset.

[0068]

[0069] where p(x,t; x s ) and g(x,-t;x r ,0) respectively represent the shot point x r The forward wave field and detection point x g Therefore, equation (8) can be rewritten as:

[0070]

[0071] in,

[0072]

[0073] In step S4, after the gradient is obtained, the current velocity model can be iteratively updated using iterative convergence methods such as steepest descent and conjugate gradient:

[0074]

[0075] When the value of the objective function ∈ is lower than the preset threshold, the iterative update stops.

[0076] In the above-mentioned skeleton inversion method, the square of the difference between the Semblance curve of the observed seismic reflection wave data and the Semblance curve of the simulated seismic reflection wave data is used as a new objective function, and then the velocity gradient is obtained based on the wave equation inversion, and the initial velocity model is iterated according to the velocity gradient to obtain the final velocity model. In the process of this application, since the square of the difference between the Semblance curve (superposition velocity curve) of the observed and simulated reflection wave data is used as a new objective function, the objective function has better convexity, so this technology can quickly converge to the vicinity of the global minimum. At the same time, the use of wave equation inversion avoids the high-frequency approximate assumptions in the tomography method, and when the underground structure is complex, the inversion accuracy is high; and for seismic reflection wave data, effective inversion modeling of deep underground background velocity information can be achieved.

[0077] In one embodiment, Figure 2 As shown, the simulated CSG data is generated by wave equation simulation based on the initial velocity model.

[0078] Specifically, Figure 2 As shown in the figure, in the data preparation stage, an initial velocity model is first established, and then an observation system consistent with that used in field observation data collection is used to simulate and calculate the simulated data common shot gather (CSG) using the wave equation.

[0079] In one embodiment, performing velocity analysis on the observed CSG data and the simulated CSG data to obtain the observed seismic reflection wave data velocity spectrum and the simulated seismic reflection wave data velocity spectrum comprises:

[0080] The observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data are respectively converted into the observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data by using the extraction channel set;

[0081] The observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data are respectively subjected to velocity scanning using the time-distance curve to generate corresponding observed seismic reflection wave data velocity spectra and simulated seismic reflection wave data velocity spectra.

[0082] Specifically, in the data preparation stage, the simulated data and the observed data are converted into a common center point gather (CMP) using the extraction gather. In the skeleton information extraction stage: its steps are basically the same as the velocity analysis steps in traditional seismic data processing, that is, the observed and simulated CMP gather data are velocity scanned using the time-distance curve to generate the corresponding observed and simulated velocity spectra.

[0083] In one embodiment, the observed data Semblance curve and the simulated data Semblance curve are obtained by picking up the observed data velocity spectrum and the simulated data velocity spectrum from the two velocity spectra through automatic picking technology.

[0084] In one embodiment, the iterative update algorithm comprises steepest descent or conjugate gradient.

[0085] Specifically, in the seismic inversion stage: the simulated and observed Semblance curves are subtracted and squared to construct the objective function. Then the connection function is constructed using formula (3). Finally, the velocity gradient is calculated using formulas (10) and (11), and the current velocity model is updated using iterative update algorithms such as the steepest descent and conjugate gradient methods. When the value of the objective function is lower than the preset threshold, the final inversion velocity model is obtained.

[0086] In one embodiment, the set threshold value is 30% of the initial error value; the initial error is obtained by comparing the difference between the data predicted by the initial model and the actual observed data.

[0087] In one embodiment, a skeleton inversion system is provided, comprising:

[0088] Data Semblance curve extraction module: used for performing velocity analysis on observed CSG seismic reflection wave data and simulated CSG seismic reflection wave data to obtain observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum, and obtaining observed data Semblance curve and simulated data Semblance curve according to the observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum respectively;

[0089] Objective function construction module: used to record the sum of squares of the difference between the observed data Semblance curve and the simulated data Semblance curve as the objective function;

[0090] The velocity gradient calculation module is used to calculate the forward and reverse wave fields using the CSG seismic reflection wave data and the positions of the shot points and the detection points, and then obtain the velocity gradient using the velocity gradient calculation formula according to the positions of the shot points and the detection points and the corresponding forward and reverse wave fields; the velocity gradient calculation formula is obtained based on the hidden variable function theory according to the link function, and the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity;

[0091] Velocity model iterative generation module: used to iteratively update the current initial velocity model using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then stop the iterative update and obtain the final velocity model.

[0092] In one embodiment, the simulated CSG data is generated by wave equation simulation based on the initial velocity model.

[0093] The advantages of this application include the following:

[0094] (1) Compared with ray tomography inversion technology, this technology avoids the high-frequency approximate assumptions in tomography methods by adopting wave equation inversion. When the underground structure is complex, the inversion accuracy is higher than that of tomography methods.

[0095] (2) Compared with full waveform inversion, it is more difficult to converge to the local minimum. This technology uses the square of the difference between the Semblance curve (stacked velocity curve) of the observed and simulated reflection wave data as a new objective function. This objective function has better convexity, so this technology can quickly converge to the vicinity of the global minimum.

[0096] (3) Compared with traditional skeleton inversion, this technology mainly targets seismic reflection wave data, so it can realize effective inversion modeling of deep underground background velocity information.

[0097] (4) Compared with the traditional velocity analysis + Dix formula depth domain velocity modeling method, although both use Semblance curve as input data, this technology avoids the horizontal layered medium assumption of the Dix formula through wave equation inversion, and the inversion accuracy is greatly improved compared with the Dix formula. The inversion results can be directly used for geological interpretation, migration imaging (as a migration velocity model) and full waveform inversion (as an initial model). This technology can replace the Dix formula as a part of conventional processing.

[0098] For the specific definition of a skeleton inversion system, please refer to the definition of a skeleton inversion method above, which will not be repeated here. Each module in the above-mentioned skeleton inversion system can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0099] In one embodiment, a computer device is provided, which may be a server, and includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above-mentioned skeleton inversion method is implemented.

[0100] In one embodiment, a computer program product is also provided, including a computer program / instruction, which involves all or part of the process in the above-mentioned embodiment method.

[0101] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0102] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A skeleton inversion method, characterized in that: The method comprises: Performing velocity analysis on the observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data to obtain a velocity spectrum of the observed seismic reflection wave data and a velocity spectrum of the simulated seismic reflection wave data, and obtaining an observed data Semblance curve and a simulated data Semblance curve according to the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data by automatic picking technology; The sum of squares of the difference between the observed data Semblance curve and the simulated data Semblance curve is recorded as the objective function; The forward and reverse wave fields are calculated by using the observed CSG seismic reflection wave data and the positions of the shot points and the detection points, and then the velocity gradient is obtained by using the velocity gradient calculation formula according to the positions of the shot points and the detection points and the corresponding forward and reverse wave fields; the velocity gradient calculation formula is obtained according to the link function and hidden variable function theory, and the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity; The current initial velocity model is iteratively updated using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then the iterative update is stopped to obtain the final velocity model.

2. The method according to claim 1, characterized in that The simulated CSG data is generated by wave equation simulation based on the initial velocity model.

3. The method according to claim 1, characterized in that The performing velocity analysis on the observed CSG data and the simulated CSG data to obtain the observed seismic reflection wave data velocity spectrum and the simulated seismic reflection wave data velocity spectrum comprises: The observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data are respectively converted into the observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data by using the extraction channel set; The observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data are respectively subjected to velocity scanning using the time-distance curve to generate corresponding observed seismic reflection wave data velocity spectra and simulated seismic reflection wave data velocity spectra.

4. The method according to claim 1, characterized in that The iterative update algorithm includes steepest descent or conjugate gradient.

5. The method according to claim 1, characterized in that The set threshold value is 30% of the initial error value; the initial error is obtained by comparing the difference between the data predicted by the initial model and the actual observed data.

6. A skeleton inversion system, characterized in that: The system comprises: Data Semblance curve extraction module: used for performing velocity analysis on observed CSG seismic reflection wave data and simulated CSG seismic reflection wave data to obtain observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum, and obtaining observed data Semblance curve and simulated data Semblance curve according to the observed seismic reflection wave data velocity spectrum and simulated seismic reflection wave data velocity spectrum respectively; Objective function building module: used to record the sum of squares of the difference between the observed data Semblance curve and the simulated data Semblance curve as the objective function; The velocity gradient calculation module is used to calculate the forward and reverse wave fields using the CSG seismic reflection wave data and the positions of the shot points and the detection points, and then obtain the velocity gradient using the velocity gradient calculation formula according to the positions of the shot points and the detection points and the corresponding forward and reverse wave fields; the velocity gradient calculation formula is obtained based on the hidden variable function theory according to the link function, and the link function is the cross-correlation of the velocity spectrum of the observed seismic reflection wave data and the velocity spectrum of the simulated seismic reflection wave data; the velocity gradient is the derivative of the objective function with respect to the velocity; Velocity model iterative generation module: used to iteratively update the current initial velocity model using an iterative update algorithm according to the velocity gradient until the value of the objective function is lower than a preset threshold value, then stop the iterative update and obtain the final velocity model.

7. The system according to claim 6, characterized in that The simulated CSG data is generated by wave equation simulation based on the initial velocity model.

8. The system according to claim 6, characterized in that The performing velocity analysis on the observed CSG data and the simulated CSG data to obtain the observed seismic reflection wave data velocity spectrum and the simulated seismic reflection wave data velocity spectrum comprises: The observed CSG seismic reflection wave data and the simulated CSG seismic reflection wave data are respectively converted into the observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data by using the extraction channel set; The observed CMP seismic reflection wave data and the simulated CMP seismic reflection wave data are respectively subjected to velocity scanning using the time-distance curve to generate corresponding observed seismic reflection wave data velocity spectra and simulated seismic reflection wave data velocity spectra.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images