WO2020143645A1

WO2020143645A1 - Full waveform inversion method and apparatus

Info

Publication number: WO2020143645A1
Application number: PCT/CN2020/070780
Authority: WO
Inventors: 桂生; 刘洪�; 王建; 冯海新
Original assignee: 中国科学院地质与地球物理研究所
Priority date: 2019-01-08
Filing date: 2020-01-08
Publication date: 2020-07-16
Also published as: CN109633742A; CN109633742B

Abstract

A full waveform inversion method and apparatus, the method comprising: using an optimal inversion frequency group to implement forward modelling of a stratum velocity model (100); using original single shot data and the result of the forward modelling of the stratum velocity model to generate a ground surface reception wave field (200); using a frequency spectrum extension method to generate a basis function of the original single shot data (300); using the basis function and the ground surface reception wave field to recover low frequency data of the original single shot data (400); on the basis of the low frequency data, generating a frequency domain gradient field of a preset number of single shot data sets (500); using the frequency domain gradient field and an optimised step size to implement inversion of the stratum velocity model (600); and executing iterative operations, using another optimal inversion frequency group to implement forward modelling of the stratum velocity model in order to obtain a final inversion result of the stratum velocity model (700). The present full waveform inversion method and apparatus can recover low frequency data from seismic material, and the low frequency data can be used for accurately obtaining full waveform inversion results.

Description

Full waveform inversion method and device

Technical field

The invention relates to the field of petroleum exploration, especially seismic data processing, and in particular to a full waveform inversion method and device.

Background technique

With the deepening of seismic exploration, the geological problems faced are becoming more and more complex; short wavelength speed can not only improve the accuracy of depth domain migration, but also can be used to indicate small uneven bodies, which can indicate sand bodies, fractures, caves, and fractures. Geological targets with differences in rock physical properties such as belts and overpressure belts. The full-waveform inversion method can invert short-wavelength velocity changes. It can be applied to short-wavelength modeling of medium-shallow velocities, or it can be used in combination with inversion to reflect wave inversion. China is in-depth research and rapid development methods. Full waveform inversion has obvious effect on constructing high-resolution velocity models, but lack of low-frequency information increases the dependence on high-precision initial models. Inaccurate initial models are likely to cause periodic jumps in the absence of low frequencies, which is not very good The background velocity of the structure makes the full waveform inversion not good. At the same time, when the actual wavelet is missing, the amplitude information is difficult to apply. The full waveform inversion based on the waveform subtraction is difficult to modify the speed correctly. The full waveform inversion based on the travel time residual will lose part of the amplitude information. As mentioned above, there is currently a lack of a A method of full waveform inversion that can recover low-frequency information of seismic data.

Summary of the invention

In view of the problems in the prior art, the present invention can establish a full waveform inversion method that can recover low-frequency information of seismic data, which can avoid the "morbidity" due to full waveform inversion, that is, the initial formation velocity model and seismic data data. Uncoupling, so that the inversion results do not converge and lead to the phenomenon of periodic jumps caused by local extreme values and lack of low frequency data.

To solve the above technical problems, the present invention provides the following technical solutions:

In a first aspect, the present invention provides a full waveform inversion method, including:

Use an optimal inversion frequency group to forward model the formation velocity;

Use the original single shot data and the forward result of the formation velocity model to generate the surface reception wave field;

Use the spectrum extension method to generate the basis function of the original single shot data;

Low frequency data of original single shot data is restored by using the basis function and the ground receiving wave field;

Generate a frequency-domain gradient field of a preset number of single shot data sets based on low-frequency data;

Invert the formation velocity model using the frequency domain gradient field and the optimized step size;

Perform an iterative operation, and use another optimal inversion frequency group to forward model the inversion formation velocity model to obtain the final inversion result of the formation velocity model.

In an embodiment, before performing forward modeling on the formation velocity model using an optimal inversion frequency group, the method further includes: transmitting the velocity model data as a whole to the GPU memory.

In an embodiment, before the overall transmission of the speed model data to the GPU memory, the method further includes:

Use Sirgue frequency optimization strategy to select the best inversion frequency;

Group the optimal inversion frequencies according to the preset frequency interval.

In one embodiment, the forward model of the velocity model using an optimal inversion frequency group includes:

Using the 12th order finite difference time domain propagation operator, the seismic wave numerical simulation is performed according to the optimal inversion frequency group and the formation velocity model, and the forward result of the formation velocity model is generated.

In one embodiment, the use of the original single shot data and the forward result of the formation velocity model to generate the surface reception wavefield includes:

Using discrete Fourier transform method, based on the original single shot data and the forward result of the formation velocity model, the frequency domain surface reception wave field of the optimal inversion frequency group is generated.

In an embodiment, generating a preset number of single shot data set frequency domain gradient fields based on low frequency data includes:

Use low frequency data and original single shot data to generate residual wave field;

The residual wave field is used to generate a gradient field in the frequency domain.

In one embodiment, the gradient field in the frequency domain and the optimized step size are used to invert the formation velocity model, including:

Calculate the sum of gradient fields in each frequency domain;

The sum of the gradient fields in each frequency domain and the optimized step size are used to invert the formation velocity model.

In a second aspect, the present invention provides a full waveform inversion device, which includes:

Forward unit, used to forward the formation velocity model using an optimal inversion frequency group;

The surface receiving wave field generation unit is used to generate the surface receiving wave field using the original single shot data and the forward result of the formation velocity model;

The basis function generating unit is used to generate the basis function of the original single shot data using the spectrum extension method;

Low-frequency data recovery unit, used to recover low-frequency data of original single shot data using basis function and ground reception wave field;

Frequency domain gradient field generation unit, used to generate a frequency domain gradient field of a preset number of single shot data sets based on low frequency data;

The inversion unit is used to invert the formation velocity model using the frequency domain gradient field and the optimized step size;

The iterative unit is used to perform iterative operations, and use another optimal inversion frequency group to forward the inversion formation velocity model to obtain the final inversion result of the formation velocity model.

In an embodiment, the full waveform inversion device further includes: a GPU transmission unit, configured to transmit the velocity model data to the GPU memory.

In an embodiment, the full-waveform inversion device further includes: an optimal inversion frequency selection unit for selecting an optimal inversion frequency using a Sirgue frequency optimization strategy;

The optimal inversion frequency grouping unit is used to group the optimal inversion frequencies according to a preset frequency interval.

In an embodiment, the forward modeling unit is specifically used to: use a 12th-order spatial finite difference time domain propagation operator to perform seismic wave numerical simulation according to the optimal inversion frequency group and the formation velocity model to generate a forward modeling result of the formation velocity model.

In one embodiment, the surface reception wave field generation unit is specifically used to generate the surface reception wave in the frequency domain of the optimal inversion frequency group based on the original single shot data and the forward result of the formation velocity model using the discrete Fourier transform method field.

In an embodiment, the frequency domain gradient field generating unit includes:

Residual wave field generation module, used to generate residual wave field using low frequency data and original single shot data;

The frequency domain gradient field generation module is used to generate a frequency domain gradient field using the residual wave field.

In an embodiment, the inversion unit includes:

Frequency domain gradient field summation module, used to calculate the sum of gradient fields in each frequency domain;

The formation velocity model inversion module is used to invert the formation velocity model by using the sum of gradient fields in each frequency domain and the optimized step size.

In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of the full waveform inversion method are implemented .

In a fourth aspect, the present invention provides a computer-readable storage medium on which a computer program is stored, which when executed by a processor implements the steps of a full waveform inversion method.

As can be seen from the above description, the present invention provides a full waveform inversion method and device, which can select the optimal inversion frequency and then use one or a group of the optimal inversion frequencies to perform forward modeling on the initial formation velocity model , Combined with a single shot data at the current frequency and the forward modeling results to generate the surface reception wave field, and then generate the basis function of the single shot data through the spectrum extension method, the basis function and the above surface reception wave field can restore the single shot The low-frequency data in the data, which is essential for the full-wavelength inversion, and the frequency-domain gradient field of the single shot data is generated based on the low-frequency data, and so on, to generate the data of all single shot data under the frequency or frequency group Gradient field, use the gradient field and optimization step to forward model the formation velocity model, and then use the result of the forward model as the initial formation velocity model of the optimal inversion frequency group for the next optimal inversion frequency, and iterate until the The optimal inversion result, or all the optimal inversion frequency or the optimal inversion frequency group are all iterated. The present invention starts from the actual data, uses the spectrum extension principle (modulation), uses the frequency shift characteristics of the signal, and uses different bases. The function reconstructs the actual data to restore the low-frequency part of the seismic data. The low-frequency data can reduce the nonlinear problem of inversion. Its objective function is relatively smooth and can easily converge to the global minimum point. It depends on the initial model. Relatively much lower, the biggest feature of this method is that the basis function obtained is dynamic, so that each time the band falls, the forward data can better match the actual data. In the initial stage of the inversion, the lower frequency data is used to obtain a more accurate smooth background velocity and a large-scale structure. On this basis, the high frequency data is used to describe the fine structure, which can improve the stability of the inversion and make the objective function gradually. Converging to near the global minimum can obtain good inversion results.

BRIEF DESCRIPTION

In order to more clearly explain the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, without paying any creative work, other drawings can be obtained based on these drawings.

1 is a schematic flowchart of a full waveform inversion method in an embodiment of the present invention;

2 is a schematic flowchart of step 500 in a full waveform inversion method in an embodiment of the present invention;

3 is a schematic flowchart of step 600 in the full waveform inversion method in the embodiment of the present invention;

4 is a schematic diagram of seismic records of original data in a specific embodiment of the present invention;

FIG. 5 is a schematic diagram of a seismic record after original data restores low-frequency information in a specific embodiment of the present invention;

6 is a schematic diagram of the spectrum analysis result of the 190th data of the original data in a specific embodiment of the present invention;

7 is a schematic diagram of the spectrum analysis result of the 190th data after the original data recovers low-frequency information in a specific embodiment of the present invention;

8 is a schematic diagram of a real formation velocity model of block A in a specific embodiment of the present invention;

9 is a schematic diagram of an initial formation velocity model of Block A in a specific embodiment of the present invention;

10 is a schematic diagram of the inversion result of the A block with only low frequency data in the original data in the specific embodiment of the present invention;

FIG. 11 is a schematic diagram of the inversion result of the A block with only high-frequency data in the original data in the specific embodiment of the present invention;

12 is a schematic diagram of the inversion result of the A block containing the full-band data in the original data in the specific embodiment of the present invention;

13 is a schematic structural diagram of a full waveform inversion device in an embodiment of the present invention;

14 is a schematic structural diagram of an electronic device in an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by a person of ordinary skill in the art without making creative efforts fall within the protection scope of the present invention.

An embodiment of the present invention provides a specific implementation of a full waveform inversion method. Referring to FIG. 1, the full waveform inversion method specifically includes the following:

Step 100: Use an optimal inversion frequency group to forward model the formation velocity model.

Step 100 is specifically: establishing a model of the formation velocity model based on the use of existing data (measurement, drilling, etc.), and using this model and the optimal inversion frequency group to obtain the corresponding seismic response.

Step 200: Use the original single shot data and the forward result of the formation velocity model to generate the surface reception wavefield.

It can be understood that the original shot data contains a lot of single shot data, first enter the shot loop, read the single shot data, determine the single shot reception range, and then generate the surface reception wave field.

Step 300: Generate a base function of the original single shot data by using the spectrum extension method.

It can be understood that the concept of spectrum extension in step 300 comes from the communication principle. The essence of spectrum extension is modulation, and the frequency shift in the time domain is the convolution in the frequency domain, which is the product of the time domain. Multiply, broaden the spectrum of information.

Step 400: Restore the low frequency data of the original single shot data by using the basis function and the ground reception wave field.

It can be understood that the basis function in step 400 is dynamic, and the low-frequency data in the original single shot data can be restored by using the basis function and the ground reception wave field generated in step 200, thereby providing data support for full waveform inversion.

Step 500: Generate a frequency domain gradient field of a preset number of single shot data sets according to the low frequency data.

Step 500 is specifically: using low-frequency data to calculate the corresponding frequency domain gradient field of the current shot recorded at the current frequency or frequency group; at the same time, the frequency domain gradient field of the common shot point gather at the current frequency is accumulated until all shot points are cycled complete.

Step 600: Invert the formation velocity model using the frequency domain gradient field and the optimized step size.

It can be understood that step 600 is to use the frequency domain gradient field and the optimized step size to estimate the structural morphology inside the formation and the change in the internal characteristics of the formation.

Step 700: Perform an iterative operation and forward the inversion formation velocity model with another optimal inversion frequency group to obtain the final inversion result of the formation velocity model.

It can be understood that step 700 describes an iterative process from step 100 to step 600. The last update of the formation velocity model, the updated result is the initial formation velocity model of the next optimal frequency group. It should be noted that At the first execution of step 100, the formation velocity model is the initial formation velocity model.

As can be seen from the above description, the present invention provides a full waveform inversion method, which can select the optimal inversion frequency, and then use one of the optimal inversion frequencies or a group of frequencies to forward the initial formation velocity model, and then Combine the single shot data at the current frequency with the forward modeling results to generate the surface reception wave field, and then generate the basis function of the single shot data by the spectrum extension method. The basis function and the above surface reception wave field can restore the single shot data Low-frequency data, which is essential for full-wavelength inversion, and generates a frequency-domain gradient field of the single shot data based on the low-frequency data, and so on, generates a gradient field of all single-shot data under the frequency or frequency group , Use the gradient field and the optimization step to forward model the formation velocity model, and then use the result of the forward model as the initial formation velocity model of the optimal inversion frequency group of the next optimal inversion frequency, and iterate until the optimal inversion is obtained. The results of the evolution, or all the optimal inversion frequencies or the optimal inversion frequency groups are all iterated. The present invention starts from the actual data, uses the spectrum extension principle (modulation), uses the frequency shift characteristics of the signal, and uses different basis functions to The actual data is reconstructed by the frequency spectrum to restore the low-frequency part of the seismic data. The low-frequency data can reduce the nonlinear problem of inversion. Its objective function is relatively smooth, and it can easily converge to the global minimum point. Its dependence on the initial model is relatively low a lot of. In the initial stage of the inversion, the lower frequency data is used to obtain a more accurate smooth background velocity and a large-scale structure. On this basis, the high frequency data is used to describe the fine structure, which can improve the stability of the inversion and make the objective function gradually. Converging to near the global minimum can obtain good inversion results.

In a specific embodiment, the present invention also provides a specific implementation method of the full waveform inversion method, which specifically includes the following steps:

Step 001: Use the Sirgue frequency optimization strategy to select the optimal inversion frequency.

Specifically, the optimal inversion frequency can be selected according to the Sirgue frequency optimization strategy (Sirgue and Pratt proposed in 2004).

Step 002: Group the optimal inversion frequencies according to a preset frequency interval.

Step 002 can be

To compare the 100 optimal inversion frequencies selected in step 001 and group the frequencies, it can be understood that each group may contain one or more optimal inversion frequencies.

Step 003: Transfer the speed model data to the GPU memory.

Step 100 is specifically: for a preset formation velocity model (built using logging data, drilling data and other geological data), using a 12th order spatial finite difference time domain propagation operator, according to the optimal inversion frequency group and formation The velocity model obtains the corresponding seismic response and generates the forward result of the formation velocity model. This step uses two-dimensional forward modeling to calculate synthetic seismic profiles.

Step 200 is specifically: using a discrete Fourier transform method, based on the original single shot data and the forward result of the formation velocity model, to generate a frequency domain surface reception wavefield of the optimal inversion frequency group.

Preferably, the surface received wave field at each moment of forward evolution can be recorded, and the forward wave field in the frequency domain of the corresponding frequency is extracted using the discrete Fourier transform method, that is, the wave field at each moment is calculated at the moment The contributions to the current wavefield in the frequency domain are accumulated.

It can be understood that the method used in step 300 is the principle of spectrum extension in communication signal processing. The essence of spectrum extension is modulation, and the frequency shift in the time domain is the convolution in the frequency domain, that is, the product in the time domain. , The two signals are multiplied point by point, which broadens the spectrum of information, specifically: two time-domain signals f ₁ (t) f ₂ (t), the expression of the frequency domain is F ₁ (ω) F ₂ (ω ), the convolution in the frequency domain can be expressed as:

F ₁ (u)*F ₂ (ω-u) realizes the frequency shift, and does a product in the time domain, and realizes a frequency shift. Choosing a suitable time domain function can widen the spectrum of the data.

Step 500: Generate a frequency domain gradient field of a preset number of single shot data sets according to the low frequency data. Referring to FIG. 2, step 500 may include:

Step 501: Use the low frequency data and the original single shot data to generate a residual wave field;

Step 502: Generate a gradient field in the frequency domain using the residual wave field.

The formula that can be used in step 502 is:

Where Grad is the gradient field in the frequency domain, u _f is the source forward wave field, u _b is the residual back wave field, v is the velocity of the underground medium, x _s and x _r are the horizontal positions of the source and the detector, respectively. ω is the frequency of current inversion.

In an embodiment, referring to FIG. 3, step 600 may include:

Step 601: Calculate the sum of gradient fields in each frequency domain.

Step 601 specifically uses formula (2) to calculate the current optimal inversion frequency or the sum of the gradient fields of all single shot data under the current optimal inversion frequency group.

Step 602: Invert the formation velocity model using the sum of gradient fields in each frequency domain and the optimized step size.

It can be understood that step 602 is to use the sum of the frequency domain gradient fields calculated in step 601 and the optimized step size to calculate the structural morphology inside the stratum and the change in stratum internal characteristics.

It is understandable that the formation velocity model is continuously updated through the parameters calculated in forward modeling, and then the updated formation velocity model is used as the initial formation velocity model of the next optimal frequency group, and so it is repeated to obtain the final formation velocity model. Inversion results.

To further illustrate this solution, the following will illustrate the effects of the present invention with a specific application example:

The method of the embodiment of the present invention is used to carry out full waveform inversion on block A of an oil field in China (the real formation velocity model of block A is known). The initial formation velocity model size is 3840mx1220m, and the spatial interval Δx=Δy=10m. Adopt frequency domain full waveform inversion, step size selection adopts step size attenuation method, inversion algorithm adopts quasi-Hessian matrix method, a total of 10 shots, time sampling interval is 1ms, sampling length is 1s, boundary strip adopts 15 layers of pml absorption boundary .

For comparison, two sets of data are used for inversion, one is the original data of the unrecovered low frequency, the other is the original data after the low frequency is restored, and then the inversion results of the original data using the unrecovered low frequency and the use of recovery The inversion results of the original data after low frequency are compared. This specific application example includes the following steps:

S1: Use the Sirgue frequency optimization strategy to select the optimal inversion frequency.

S2: to

Group the optimal inversion frequencies for equal ratio.

S3: Transfer the preset speed model data to GPU memory.

It is understandable that the initial formation velocity model established in advance using well logging data, drilling data and other geological data is transferred to the GPU memory.

S4: Use the 12th order finite difference time domain propagation operator to obtain the corresponding seismic response according to the optimal inversion frequency group (or optimal inversion frequency) and the formation velocity model to generate the two-dimensional forward model of the formation velocity model result.

S5: Using the discrete Fourier transform method, based on the original single shot data and the two-dimensional forward results of the formation velocity model, the frequency domain surface reception wave field of the optimal inversion frequency group is generated.

S6: Generate a basis function of the original single shot data by using the spectrum extension method.

S6: Recover low frequency data of original single shot data using basis function and ground reception wavefield, see Figure 4 to Figure 7,

Comparing Figure 4 and Figure 5, it can be clearly seen that the low-frequency information in the original data can be recovered better by the spectrum extension method (modulation).

Comparing Fig. 6 and Fig. 7, it can be clearly seen that the low-frequency information in the spectrum analysis of 190 data in the original data is recovered better by the spectrum extension method (modulation), especially the frequency below 20 Hz is obviously recovered.

S7: Use the low frequency data and the original single shot data to generate a residual wave field.

S8: Use formula (2) to generate a gradient field in the frequency domain.

S9: Calculate the current optimal inversion frequency or the sum of the gradient fields of all single shot data under the current optimal inversion frequency group.

S10: Use the step size attenuation method to obtain the optimized step size

S11: Invert the formation velocity model using the sum of gradient fields in each frequency domain and the optimized step size.

S12: Perform an iterative operation, use another optimal inversion frequency group to forward the inversion formation velocity model to obtain the final inversion result of the formation velocity model, and use the method provided in this example to the original data separately Only low-frequency data, only high-frequency data and full-band data are inverted, and compared with the formation velocity model (real formation velocity model) of Block A and the initial model. It can be seen that when there is only low-frequency data in the original data , Can restore the background speed to a certain extent, the shallow part, when there is only high-frequency data, is not much different from the initial model, this is because the first iteration falls into a local minimum, iterative update is not possible, and it is easy to produce cycles The jump falls into a local minimum; and the full-band data after recovering the low-frequency data can obtain the most accurate inversion results, see FIGS. 8-12.

Based on the same inventive concept, the embodiments of the present application also provide a full waveform inversion device, which can be used to implement the method described in the foregoing embodiments, as described in the following embodiments. Since the principle of the full waveform inversion device to solve the problem is similar to that of the full waveform inversion method, the implementation of the full waveform inversion device can be referred to the implementation of the full waveform inversion method, and the repetition will not be repeated. As used below, the term "unit" or "module" may implement a combination of software and/or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, implementation of hardware or a combination of software and hardware is also possible and conceived.

Embodiments of the present invention provide a specific implementation of a full waveform inversion device. Referring to FIG. 13, the full waveform inversion device specifically includes the following:

The forward unit 10 is used to forward the formation velocity model using an optimal inversion frequency group;

The surface receiving wave field generating unit 20 is used to generate the surface receiving wave field using the original single shot data and the forward result of the formation velocity model;

The basis function generating unit 30 is used to generate the basis function of the original single shot data by using the spectrum extension method;

A low-frequency data restoration unit 40, configured to restore the low-frequency data of the original single shot data by using the basis function and the ground reception wave field;

The frequency domain gradient field generating unit 50 is configured to generate a frequency domain gradient field of a preset number of single shot data sets according to the low frequency data;

The inversion unit 60 is used to invert the formation velocity model using the frequency domain gradient field and the optimized step size;

The iterative unit 70 is used to perform an iterative operation, and forward the inversion formation velocity model with another optimal inversion frequency group to obtain the final inversion result of the formation velocity model.

In an embodiment, the frequency domain gradient field generating unit includes:

In an embodiment, the inversion unit includes:

As can be seen from the above description, the present invention provides a full-waveform inversion device, which can select the optimal inversion frequency, and then use one or a group of the optimal inversion frequencies to forward model the initial formation velocity model, and then Combine the single shot data at the current frequency with the forward modeling results to generate the surface reception wave field, and then generate the basis function of the single shot data by the spectrum extension method. The basis function and the above surface reception wave field can restore the single shot data Low-frequency data, which is essential for full-wavelength inversion, and generates a frequency-domain gradient field of the single shot data based on the low-frequency data, and so on, generates a gradient field of all single-shot data under the frequency or frequency group , Use the gradient field and the optimization step to forward model the formation velocity model, and then use the result of the forward model as the initial formation velocity model of the optimal inversion frequency group of the next optimal inversion frequency, and iterate until the optimal inversion is obtained. The results of the evolution, or all the optimal inversion frequencies or the optimal inversion frequency groups are all iterated. The present invention starts from the actual data, uses the spectrum extension principle (modulation), uses the frequency shift characteristics of the signal, and uses different basis functions to The actual data is reconstructed by the frequency spectrum to restore the low-frequency part of the seismic data. The low-frequency data can reduce the nonlinear problem of inversion. Its objective function is relatively smooth, and it can easily converge to the global minimum point. Its dependence on the initial model is relatively low a lot of. In the initial stage of the inversion, the lower frequency data is used to obtain a more accurate smooth background velocity and a large-scale structure. On this basis, the high frequency data is used to describe the fine structure, which can improve the stability of the inversion and make the objective function gradually. Converging to near the global minimum can obtain good inversion results.

The embodiments of the present application also provide a specific implementation manner of an electronic device capable of implementing all the steps in the full waveform inversion method in the above embodiments. Referring to FIG. 14, the electronic device specifically includes the following content:

Processor (processor) 1201, memory (memory) 1202, communication interface (Communications) Interface 1203 and bus 1204;

The processor 1201, the memory 1202, and the communication interface 1203 communicate with each other through the bus 1204; the communication interface 1203 is used to implement information between related devices such as server-side devices, detection devices, and user-end devices. transmission;

The processor 1201 is configured to call a computer program in the memory 1202. When the processor executes the computer program, all steps in the full waveform inversion method in the foregoing embodiment are implemented, for example, the processor executes The computer program implements the following steps:

As can be seen from the above description, the electronic device in the embodiment of the present application can select the optimal inversion frequency, and then use one or a group of the optimal inversion frequencies to forward model the initial formation velocity model, and then combine the current A single shot data in the frequency and the forward modeling result generate the surface reception wave field, and then generate the basis function of the single shot data by the spectrum extension method, the basis function and the above surface reception wave field can restore the low frequency in the single shot data Data, this low-frequency data is essential for the full-wavelength inversion, and the frequency-domain gradient field of the single shot data is generated based on the low-frequency data, and so on, to generate the gradient field of all single shot data under this frequency or frequency group, using Gradient field and optimization step are used to forward the formation velocity model, and then use the forward result as the initial formation velocity model of the optimal inversion frequency group of the next optimal inversion frequency, and iterate until the optimal inversion result is obtained , Or all the optimal inversion frequencies or the optimal inversion frequency groups are all iterated. The present invention starts from the actual data, uses the spectrum extension principle (modulation), uses the frequency shift characteristics of the signal, and uses different basis functions to convert the actual data Perform spectrum reconstruction to restore the low-frequency part of the seismic data. The low-frequency data can reduce the nonlinear problem of inversion. Its objective function is relatively smooth and can easily converge to the global minimum point. Its dependence on the initial model is relatively low. In the initial stage of the inversion, the lower frequency data is used to obtain a more accurate smooth background velocity and a large-scale structure. On this basis, the high frequency data is used to describe the fine structure, which can improve the stability of the inversion and make the objective function gradually. Converging to near the global minimum can obtain good inversion results.

An embodiment of the present application also provides a computer-readable storage medium capable of implementing all the steps in the full waveform inversion method in the above embodiments, the computer-readable storage medium stores a computer program, which is processed by a processor During execution, all steps of the full waveform inversion method in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are realized:

As can be seen from the above description, the computer-readable storage medium in the embodiments of the present application can select the optimal inversion frequency, and then use one or a group of the optimal inversion frequencies to forward model the initial formation velocity model. Combined with the single shot data at the current frequency and the forward result, a ground reception wave field is generated, and then the basis function of the single shot data is generated by the spectrum extension method. The basis function and the above-mentioned surface reception wave field can restore the single shot data The low-frequency data in the medium, this low-frequency data is essential for the full-wavelength inversion, and the frequency-domain gradient field of the single shot data is generated based on the low-frequency data, and so on, and the gradient of all single shot data under the frequency or frequency group Field, use the gradient field and the optimization step to forward model the formation velocity model, and then use the result of the forward model as the initial formation velocity model of the optimal inversion frequency group of the next optimal inversion frequency, and iterate until the optimal The inversion result, or all the optimal inversion frequencies or the optimal inversion frequency groups are all iterated. The present invention starts from the actual data, uses the spectrum extension principle (modulation), uses the frequency shift characteristics of the signal, and uses different basis functions Reconstruct the spectrum of the actual data and restore the low-frequency part of the seismic data. The low-frequency data can reduce the nonlinear problem of inversion. Its objective function is relatively smooth, and it can easily converge to the global minimum point. Its dependence on the initial model is relatively Much lower. In the initial stage of the inversion, the lower frequency data is used to obtain a more accurate smooth background velocity and a large-scale structure. On this basis, the high frequency data is used to describe the fine structure, which can improve the stability of the inversion and make the objective function gradually. Converging to near the global minimum can obtain good inversion results.

The embodiments in this specification are described in a progressive manner. The same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method embodiment.

The foregoing describes specific embodiments of the present specification. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the particular order shown or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Although the present application provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative labor. The order of the steps listed in the embodiment is only one way among the execution order of many steps, and does not represent a unique execution order. When the actual device or client product is executed, it may be executed sequentially or in parallel according to the method shown in the embodiments or the drawings (for example, a parallel processor or a multi-threaded processing environment).

Although the embodiments of the present specification provide method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive means. The order of the steps listed in the embodiment is only one way among the execution order of many steps, and does not represent a unique execution order. When the actual device or terminal product is executed, it may be executed sequentially or in parallel according to the method shown in the embodiments or the drawings (such as a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "include", "include", or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, product, or device that includes a series of elements includes not only those elements, but also others that are not explicitly listed Elements, or also include elements inherent to such processes, methods, products, or equipment. Without more restrictions, it does not exclude that there are other identical or equivalent elements in the process, method, product or equipment including the elements.

For the convenience of description, when describing the above device, the functions are divided into various modules and described separately. Of course, when implementing the embodiments of the present specification, the functions of each module may be implemented in one or more software and/or hardware, or the modules that implement the same function may be implemented by a combination of multiple submodules or subunits. The device embodiments described above are only schematic. For example, the division of the unit is only a division of logical functions. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or integrated To another system, or some features can be ignored, or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical, or other forms.

Those skilled in the art also know that, in addition to implementing the controller in the form of pure computer-readable program code, the method can be logically programmed to enable the controller to use logic gates, switches, special integrated circuits, programmable logic controllers and embedded The same function is realized in the form of a microcontroller or the like. Therefore, such a controller can be regarded as a hardware component, and the devices included therein for realizing various functions can also be regarded as structures within the hardware component. Or even, the means for realizing various functions can be regarded as both a software module of the implementation method and a structure within a hardware component.

The present invention is described with reference to flowcharts and/or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and/or block in the flowchart and/or block diagram and a combination of the flow and/or block in the flowchart and/or block diagram may be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processing machine, or other programmable data processing device to produce a machine that enables the generation of instructions executed by the processor of the computer or other programmable data processing device A device for realizing the functions specified in one block or multiple blocks of one flow or multiple blocks of a flowchart.

These computer program instructions may also be stored in a computer readable memory that can guide a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, the instructions The device implements the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operating steps are performed on the computer or other programmable device to generate computer-implemented processing, which is executed on the computer or other programmable device The instructions provide steps for implementing the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.

In a typical configuration, the computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include non-permanent memory, random access memory (RAM) and/or non-volatile memory in computer-readable media, such as read only memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

Computer-readable media, including permanent and non-permanent, removable and non-removable media, can store information by any method or technology. The information may be computer readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape storage or other magnetic storage devices or any other non-transmission media can be used to store information that can be accessed by computing devices. As defined in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

Those skilled in the art should understand that the embodiments of the present specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present specification may take the form of computer program products implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

Embodiments of this specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The embodiments of the present specification can also be practiced in distributed computing environments in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.

The embodiments in this specification are described in a progressive manner. The same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method embodiment. In the description of this specification, the description referring to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means specific features described in conjunction with the embodiments or examples , Structure, material, or characteristics are included in at least one embodiment or example of the embodiments of this specification. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. In addition, without contradicting each other, those skilled in the art may combine and combine different embodiments or examples and features of the different embodiments or examples described in this specification.

The above are only examples of the embodiments of the present specification, and are not intended to limit the embodiments of the present specification. For those skilled in the art, various modifications and changes can be made to the embodiments of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the embodiments of this specification should be included in the scope of the claims of the embodiments of this specification.

Claims

A full waveform inversion method, which is characterized by including:

Use an optimal inversion frequency group to forward model the formation velocity;

Use the original single shot data and the forward result of the formation velocity model to generate the surface reception wave field;

Generating the basis function of the original single shot data by using the spectrum extension method;

Recovering the low frequency data of the original single shot data by using the basis function and the ground receiving wave field;

Generating a frequency domain gradient field of a preset number of single shot data sets according to the low frequency data;

Invert the formation velocity model using the frequency domain gradient field and the optimized step size;

Perform an iterative operation, and use another optimal inversion frequency group to forward model the inversion formation velocity model to obtain the final inversion result of the formation velocity model.
The full waveform inversion method according to claim 1, wherein before the forward modeling of the formation velocity model using an optimal inversion frequency group, the method further comprises:

Transfer the speed model data to the GPU memory.
The full waveform inversion method according to claim 2, characterized in that before the overall transmission of the velocity model data to the GPU memory, the method further comprises:

Use Sirgue frequency optimization strategy to select the best inversion frequency;

Group the optimal inversion frequencies according to preset frequency intervals.
The full waveform inversion method according to claim 1, wherein the forward modeling of the velocity model using an optimal inversion frequency group includes:

A 12-order spatial finite difference time domain propagation operator is used to perform seismic wave numerical simulation according to the optimal inversion frequency group and formation velocity model to generate a forward result of the formation velocity model.
The full waveform inversion method according to claim 1, wherein the generating of the surface reception wave field using the original single shot data and the forward result of the formation velocity model includes:

Using the discrete Fourier transform method, the frequency domain surface reception wave field of the optimal inversion frequency group is generated according to the original single shot data and the forward result of the formation velocity model.
The full waveform inversion method according to claim 1, wherein the generating a frequency domain gradient field of a preset number of single shot data sets based on the low frequency data includes:

Generating a residual wave field using the low frequency data and the original single shot data;

The residual wave field is used to generate a gradient field in the frequency domain.
The full waveform inversion method according to claim 1, wherein the inversion of the formation velocity model using the frequency domain gradient field and the optimized step size includes:

Calculate the sum of gradient fields in each frequency domain;

The sum of the gradient fields in each frequency domain and the optimized step size are used to invert the formation velocity model.
A full waveform inversion device, which is characterized by including:

Forward unit, used to forward the formation velocity model using an optimal inversion frequency group;

The surface receiving wave field generation unit is used to generate the surface receiving wave field using the original single shot data and the forward result of the formation velocity model;

A basis function generating unit, which is used to generate a basis function of the original single shot data by using a spectrum extension method;

A low-frequency data recovery unit for recovering the low-frequency data of the original single shot data using the basis function and the ground reception wave field;

A frequency domain gradient field generating unit, configured to generate a frequency domain gradient field of a preset number of single shot data sets according to the low frequency data;

An inversion unit for inverting the formation velocity model using the frequency domain gradient field and the optimized step size;

The iterative unit is used to perform iterative operations, and use another optimal inversion frequency group to forward the inversion formation velocity model to obtain the final inversion result of the formation velocity model.
The full waveform inversion device according to claim 8, further comprising:

The GPU transmission unit is used to transmit the speed model data to the GPU memory.
The full waveform inversion device according to claim 8, further comprising:

The optimal inversion frequency selection unit is used to select the optimal inversion frequency using the Sirgue frequency optimization strategy;

The optimal inversion frequency grouping unit is used to group the optimal inversion frequencies according to a preset frequency interval.
The full waveform inversion device according to claim 8, characterized in that the forward inversion unit is specifically used for: using a 12th order spatial finite difference time domain propagation operator, according to the optimal inversion frequency group and formation velocity The model performs numerical simulation of seismic waves to generate forward results of the formation velocity model.
The full waveform inversion device according to claim 8, characterized in that the ground surface reception wave field generating unit is specifically used to: use a discrete Fourier transform method, based on the original single shot data and the formation velocity model The result of the forward modeling is to generate the received wave field in the frequency domain of the optimal inversion frequency group.
The full waveform inversion device according to claim 8, wherein the frequency domain gradient field generating unit comprises:

A residual wave field generation module, configured to generate a residual wave field using the low frequency data and the original single shot data;

The frequency domain gradient field generation module is used to generate a frequency domain gradient field using the residual wave field.
The full waveform inversion device according to claim 8, wherein the inversion unit comprises:

Frequency domain gradient field summation module, used to calculate the sum of gradient fields in each frequency domain;

The formation velocity model inversion module is used to invert the formation velocity model using the sum of the gradient fields in each frequency domain and the optimized step size.
An electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, when the processor executes the program, any one of claims 1 to 7 is realized The steps of the full waveform inversion method.
A computer-readable storage medium on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the full waveform inversion method according to any one of claims 1 to 7 are realized.