CN111694052A

CN111694052A - Blind inversion method and device

Info

Publication number: CN111694052A
Application number: CN201910192024.4A
Authority: CN
Inventors: 郭欣; 高建虎; 王洪求; 雍学善; 李胜军; 刘炳杨
Original assignee: Petrochina Co Ltd
Current assignee: Petrochina Co Ltd
Priority date: 2019-03-14
Filing date: 2019-03-14
Publication date: 2020-09-22
Anticipated expiration: 2039-03-14
Also published as: CN111694052B

Abstract

The invention discloses a blind inversion method and a blind inversion device, wherein the method comprises the following steps: determining a reflection coefficient matrix of the seismic record by utilizing sparse pulse inversion, converting the reflection coefficient matrix into a reflection coefficient vector, respectively converting a first-order difference matrix and a second-order difference matrix on a time dimension, a line dimension and a channel dimension into a total variation regularization matrix and a Gihonov regularization matrix, and constructing a target function of wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix, wherein the wave impedance comprises transverse abrupt wave impedance and transverse gradual change wave impedance; and determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the minimum value is obtained according to the objective function. The invention utilizes total variation regularization and Gihonov regularization to carry out regularization constraint on the transverse abrupt wave impedance and the transverse gradual change wave impedance respectively, utilizes the constructed objective function to solve the wave impedance, and can improve the continuity of the transverse energy of the wave impedance.

Description

Blind inversion method and device

Technical Field

The invention relates to the technical field of oil and gas field exploration, in particular to a blind inversion method and device.

Background

This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims. The description herein is not admitted to be prior art by inclusion in this section.

The wave impedance inversion actually eliminates the influence of wavelets from the seismic profile, leaves a reflection coefficient, and calculates the physical parameter wave impedance capable of reflecting the change of the formation physical property according to the reflection coefficient. The wave impedance inversion is very important in the field of seismic exploration, is widely used for seismic interpretation and reservoir prediction, and is also well applied to the aspect of rock physics.

The wave impedance inversion problem typically involves deconvolution, where the reflection coefficient is solved by extracting seismic wavelets from the seismic record, and then the wave impedance is obtained according to the recursive principle. However, due to uncertainty of the wavelet, the continuity of the reflection coefficient in the transverse direction is poor, and meanwhile, the recursion process is very sensitive to noise interference, so that the inverted wave impedance has the problems of large energy change and poor continuity in the transverse direction, and a 'fine line' phenomenon appears on a seismic wave impedance profile, which is not beneficial to reservoir prediction and analysis.

Therefore, the conventional wave impedance inversion has a problem of poor continuity of wave impedance lateral energy.

Disclosure of Invention

The embodiment of the invention provides a blind inversion method for improving the continuity of wave impedance transverse energy, which comprises the following steps:

determining a reflection coefficient matrix of the seismic record by utilizing sparse pulse inversion, and converting the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension and channel dimension;

respectively converting the constructed first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension;

respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension;

constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance;

and determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value.

The embodiment of the invention also provides a blind inversion device for improving the continuity of the wave impedance transverse energy, which comprises:

the reflection coefficient determining module is used for determining a reflection coefficient matrix of the seismic record by utilizing sparse pulse inversion and converting the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension and channel dimension;

the total variation regularization conversion module is used for converting the constructed first-order difference matrixes in the time dimension, the line dimension and the channel dimension into total variation regularization matrixes in the time dimension, the line dimension and the channel dimension respectively;

the Gihono regularization conversion module is used for converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihono regularization matrixes in the time dimension, the line dimension and the channel dimension respectively;

the target function construction module is used for constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance;

and the wave impedance determination module is used for determining the wave impedance of the seismic record according to the transverse abrupt change wave impedance and the transverse gradual change wave impedance when the objective function of the wave impedance obtains the minimum value.

The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the blind inversion method when executing the computer program.

Embodiments of the present invention further provide a computer-readable storage medium, which stores a computer program for executing the blind inversion method.

In the embodiment of the invention, a reflection coefficient matrix of a seismic record is determined by utilizing sparse pulse inversion, and the reflection coefficient matrix is converted into a reflection coefficient vector according to the sequence of time dimension, line dimension and channel dimension; respectively converting the constructed first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension; respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension; constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance; and determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value. According to the embodiment of the invention, the first-order difference matrix and the second-order difference matrix are respectively converted into the total variation regularization matrix and the Gihonov regularization matrix, the total variation regularization and the Gihonov regularization are respectively utilized to carry out regularization constraint on the transverse abrupt wave impedance and the transverse gradual wave impedance, and the wave impedance is solved according to the constructed objective function of the wave impedance, so that the continuity of the transverse energy of the wave impedance can be improved, and the reliability of the wave impedance inversion result is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:

fig. 1 is a flowchart illustrating an implementation of a blind inversion method according to an embodiment of the present invention;

fig. 2 is a flowchart illustrating an implementation of step 102 in a blind inversion method according to an embodiment of the present invention;

fig. 3 is a flowchart illustrating an implementation of step 202 in a blind inversion method according to an embodiment of the present invention;

fig. 4 is a flowchart illustrating an implementation of step 103 in the blind inversion method according to an embodiment of the present invention;

FIG. 5 is a flowchart illustrating an implementation of step 402 in a blind inversion method according to an embodiment of the present invention;

fig. 6 is a flowchart illustrating an implementation of step 104 in the blind inversion method according to an embodiment of the present invention;

FIG. 7 is a flowchart illustrating the implementation of step 105 in the blind inversion method according to an embodiment of the present invention;

FIG. 8 is a functional block diagram of a blind inversion apparatus according to an embodiment of the present invention;

fig. 9 is a block diagram of a total variation regularization conversion module 802 in the blind inversion apparatus according to an embodiment of the present invention;

fig. 10 is a block diagram of a total variation regularization conversion unit 902 in the blind inversion apparatus according to an embodiment of the present invention;

fig. 11 is a structural block diagram of the gihonov regularization conversion module 803 in the blind inversion apparatus according to the embodiment of the present invention;

fig. 12 is a block diagram of a structure of a gihonov regularization conversion unit 1102 in the blind inversion apparatus according to the embodiment of the present invention;

fig. 13 is a block diagram of an objective function constructing module 804 in the blind inversion apparatus according to the embodiment of the present invention;

fig. 14 is a block diagram of a wave impedance determination module 805 in a blind inversion apparatus according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.

Although the present invention provides the method operation steps or apparatus structures as shown in the following embodiments or figures, more or less operation steps or module units may be included in the method or apparatus based on conventional or non-inventive labor. In the case of steps or structures which do not logically have the necessary cause and effect relationship, the execution order of the steps or the block structure of the apparatus is not limited to the execution order or the block structure shown in the embodiment or the drawings of the present invention. The described methods or modular structures, when applied in an actual device or end product, may be executed sequentially or in parallel according to embodiments or the methods or modular structures shown in the figures.

Aiming at the defect of poor continuity of wave impedance transverse energy in blind inversion in the prior art, the applicant of the invention provides a blind inversion method and a blind inversion device, wherein a constructed first-order difference matrix on a time dimension, a line dimension and a channel dimension is respectively converted into a total variation regularization matrix on the time dimension, the line dimension and the channel dimension; respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension; and then carrying out regularization constraint on the transverse abrupt wave impedance and the transverse gradual wave impedance by using total variation regularization and Gihonov regularization, and solving the wave impedance according to a constructed target function of the wave impedance, so that the aims of improving the continuity of the transverse energy of the wave impedance and improving the reliability of a blind inversion result are fulfilled.

Fig. 1 shows a flow of implementing the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

as shown in fig. 1, a blind inversion method includes:

101, determining a reflection coefficient matrix of a seismic record by utilizing sparse pulse inversion, and converting the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension and channel dimension;

102, respectively converting the constructed first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension;

103, respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension;

104, constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance;

and 105, determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the minimum value is obtained according to the objective function of the wave impedance.

The seismic recording means that seismic waves are excited by a seismic source and are propagated in an underground medium and finally received by detectors, and seismic wave signals received by each detector form a seismic recording.

The sparse pulse inversion is a recursive inversion method based on sparse pulse deconvolution, mainly comprises maximum likelihood deconvolution, L1 norm deconvolution, minimum entropy deconvolution and the like, can be used for solving and determining the reflection coefficient of a seismic record, and in the embodiment of the invention, the reflection coefficient of the seismic record is embodied in a matrix form.

Specifically, when the reflection coefficient matrix of the seismic record is determined by using sparse pulse inversion, the reflection coefficient matrix of the seismic record can be solved and determined according to the following relation between the discrete wave impedance and the reflection coefficient:

where r represents the reflection coefficient matrix, z represents the discrete wave impedance, and j represents a layer of the subsurface layered medium (assuming that the subsurface is a layered medium). In one embodiment of the invention, the seismic records are three-dimensional seismic records, which mainly comprise three dimensions, namely a time dimension, a line dimension and a trace dimension. In order to facilitate the subsequent construction of an objective function and the determination of wave impedance, the reflection coefficients in the form of a matrix are converted into the reflection coefficients in the form of vectors according to the sequence of time dimension, line dimension and track dimension, that is, the reflection coefficient matrix is converted into the reflection coefficient vectors according to the sequence of time dimension, line dimension and track dimension.

In addition, in order to construct the objective function, the wave impedance in the relation between the discrete wave impedance and the reflection coefficient matrix is replaced, so that

Then there are:

namely:

r[j]＝＝x[j+1]-x[j]；

in addition, the linear relationship between the reflection coefficient matrix and the logarithmic impedance can be expressed as:

wherein r represents a reflection coefficient matrix, and D represents a first order difference matrix (n-1 rows and n columns) including a first order difference matrix D in the time dimension₁₁First order difference matrix D in line dimension₁₂And a first order difference matrix D in the channel dimension₁₃Another order difference matrix D₁₁、D₁₂And D₁₃The number of rows and/or columns may be the same or different.

Regularization refers to when the solution of the linear equation Am ═ b does not exist or is not unique (so-called ill-conditioned problem), for the case that the solution does not exist, some conditions are added to find an approximate solution, for the case that the solution is not unique, some limitations are added to narrow the solution range, and the method of solving the ill-conditioned problem by adding the conditions or the limitations is the regularization method.

Total Variation (TV) regularization in earthquakeThe demonstration aspect is widely applied. In order to constrain the wave impedance by using total variation regularization, first-order difference matrixes D in the time dimension are respectively constructed₁₁First order difference matrix D in line dimension₁₂And a first order difference matrix D in the channel dimension₁₃And respectively combining the first order difference matrix D in the time dimension₁₁Converting into a time-dimension total variation regularization matrix, and converting into a line-dimension first-order difference matrix D₁₂Converting into a total variation regularization matrix in line dimension, and converting into a first-order difference matrix D in channel dimension₁₃Conversion to a total variation regularization matrix in the track dimension.

For the linear equation of morbidity, Am ═ b, giHonowov proposed the use of | | | Am-b | | luminance²+||m||²Is called the Tikhonov matrix.

Similarly, in order to constrain the wave impedance by utilizing the Gihonov regularization, first, a second-order difference matrix D in the time dimension is respectively constructed₂₁Second order difference matrix D in line dimension₂₂And a second order difference matrix D in the channel dimension₂₃And respectively converting the second order difference matrix D in the time dimension₂₁Converting into Gihono regularization matrix in time dimension, and converting second order difference matrix D in line dimension₂₂Converting into Gihono regularization matrix in line dimension, and converting into second order difference matrix D in track dimension₂₃And converting into a Gihono regularization matrix on the track dimension.

After the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix are determined, an objective function of the wave impedance is constructed according to the reflection coefficient vector, the total variation regularization matrix in the time dimension, the total variation regularization matrix in the line dimension, the total variation regularization matrix in the channel dimension, the Gihonov regularization matrix in the time dimension, the Gihonov regularization matrix in the line dimension and the Gihonov regularization matrix in the channel dimension.

In the embodiment of the invention, the wave impedance comprises a transverse abrupt wave impedance part and a transverse gradual change wave impedance part, and the transverse abrupt wave impedance and the transverse gradual change wave impedance are respectively subjected to regularization constraint by using a total variation regularization matrix and a Gihonov regularization matrix, so that the continuity of the transverse energy of the wave impedance is improved, and the reliability of a wave impedance inversion result is improved.

The method comprises the steps of utilizing a total variation regularization matrix and a Gihonov regularization matrix to carry out regularization constraint on transverse abrupt wave impedance and transverse gradient wave impedance respectively, constructing an objective function of wave impedance, processing the objective function of the wave impedance, determining the transverse abrupt wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains a minimum value, further obtaining the transverse abrupt wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value, and determining the wave impedance of a seismic record, so that the wave impedance of the seismic record can be reliably obtained, and the continuity of transverse energy of the wave impedance is improved.

In the embodiment of the invention, a reflection coefficient matrix of a seismic record is determined by utilizing sparse pulse inversion, and the reflection coefficient matrix is converted into a reflection coefficient vector according to the sequence of time dimension, line dimension and trace dimension; respectively converting the constructed first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension; respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension; constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance; and determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value. According to the embodiment of the invention, the first-order difference matrix and the second-order difference matrix are respectively converted into the total variation regularization matrix and the Gihonov regularization matrix, the total variation regularization and the Gihonov regularization are respectively utilized to carry out regularization constraint on the transverse abrupt wave impedance and the transverse gradual wave impedance, and the wave impedance is solved according to the constructed objective function of the wave impedance, so that the continuity of the transverse energy of the wave impedance can be improved, and the reliability of the wave impedance inversion result is improved.

In one embodiment of the invention, the seismic records are real seismic records. The wave impedance is determined by using the real seismic records, and the reliability of the wave impedance inversion result can be improved.

In an embodiment of the invention, based on the above method steps, the blind inversion method further includes:

the method comprises the following steps: determining a theoretical reflection coefficient according to the established wave impedance model;

the method comprises the following steps: and (4) performing convolution on the theoretical reflection coefficient and the Rake wavelets to obtain the seismic record.

In addition, when the seismic record is obtained, a wave impedance model can be established firstly, then a theoretical reflection coefficient is determined according to the established wave impedance model, then the theoretical reflection coefficient and the Rake wavelets are convoluted to obtain the seismic record, and the seismic record obtained at the moment is the theoretical seismic record. In an embodiment of the present invention, the Rake wavelet is a Rake wavelet with a dominant frequency of 30 Hz.

and adding noise with a preset signal-to-noise ratio into the seismic record to obtain the seismic record after the noise is added.

Correspondingly, step 101, determining a reflection coefficient matrix of the seismic record by utilizing sparse pulse inversion, and converting the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension and trace dimension, including:

and determining a reflection coefficient matrix of the seismic record after the noise is added by utilizing sparse pulse inversion, and converting the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension and channel dimension.

In order to simulate real seismic record, after theoretical seismic record is obtained, noise with a certain signal-to-noise ratio is added into the theoretical seismic record to improve the reliability of the seismic record, further improve the continuity of wave impedance transverse energy and improve the reliability of blind inversion results.

In an embodiment of the present invention, the preset signal-to-noise ratio is a preset signal-to-noise ratio, and a person skilled in the art may preset the preset signal-to-noise ratio according to actual conditions and specific requirements. For example, the preset signal-to-noise ratio is preset to 0.2, or 0.22, or 0.18, etc. In addition, it will be understood by those skilled in the art that the preset snr can also be preset to other snrs besides the above snr. For example, the preset snr is preset to 0.25 or 0.17, etc., which is not particularly limited in the embodiment of the present invention.

Fig. 2 shows an implementation flow of step 102 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, and detailed descriptions are as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result, as shown in fig. 2, step 102 is to convert the constructed first-order difference matrices in the time dimension, the line dimension and the track dimension into total variation regularization matrices in the time dimension, the line dimension and the track dimension, respectively, and includes:

step 201, constructing a first-order difference matrix in a time dimension, a line dimension and a channel dimension;

step 202, respectively converting the first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension according to the first-order difference matrixes in the time dimension, the line dimension and the track dimension and the unit matrixes in the time dimension, the line dimension and the track dimension.

When determining the total variation regularization matrix in different dimensions such as time dimension, line dimension and track dimension, firstly, constructing a first-order difference matrix D of the time dimension₁₁First order difference matrix D in line dimension₁₂And a first order difference matrix D in the channel dimension₁₃And further respectively converting the first order difference matrix D of the time dimension₁₁Converting into a time-dimension total variation regularization matrix, and converting into a line-dimension first-order difference matrix D₁₂Converting into a total variation regularization matrix in line dimension, and converting into a first-order difference matrix D in channel dimension₁₃Conversion to a total variation regularization matrix in the track dimension.

In the embodiment of the invention, the first-order difference matrixes in the time dimension, the line dimension and the channel dimension are constructed, and the first-order difference matrixes in the time dimension, the line dimension and the channel dimension are respectively converted into the total variation regularization matrixes in the time dimension, the line dimension and the channel dimension according to the first-order difference matrixes in the time dimension, the line dimension and the channel dimension and the unit matrixes in the time dimension, the line dimension and the channel dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 3 shows an implementation flow of step 202 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result, as shown in fig. 3, step 202 is to convert the first-order difference matrices in the time dimension, the line dimension and the track dimension into total variation regularization matrices in the time dimension, the line dimension and the track dimension, respectively, according to the first-order difference matrices in the time dimension, the line dimension and the track dimension and the unit matrices in the time dimension, the line dimension and the track dimension, and includes:

step 301, determining a total variation regularization matrix in a time dimension according to an identity matrix in a channel dimension and an identity matrix in a line dimension and a first-order difference matrix in the time dimension;

step 302, determining a total variation regularization matrix in a line dimension according to an identity matrix in a channel dimension and a first-order difference matrix in the line dimension and an identity matrix in a time dimension;

step 303, determining a total variation regularization matrix in the channel dimension according to the first-order difference matrix in the channel dimension and the unit matrix in the line dimension, and the unit matrix in the time dimension.

When the total variation regularization matrix in the time dimension, the line dimension and the track dimension is determined, the identity matrix in the time dimension, the line dimension and the track dimension is needed. Specifically, in determining a total variation regularization matrix in a time dimension, an identity matrix in a channel dimension and an identity matrix in a line dimension are required to be utilized; when determining a total variation regularization matrix on a line dimension, utilizing an identity matrix on a track dimension and an identity matrix on a time dimension; when determining the total variation regularization matrix in the track dimension, an identity matrix in the line dimension and an identity matrix in the time dimension are needed.

Wherein the first order difference matrix D in the time dimension can be expressed by the following formula₁₁Conversion to total variation regularization matrix in time dimension:

L₁₁representing a global variation regularization matrix in the time dimension, I₃Representing an identity matrix in the track dimension, I₂Representing identity matrices in line dimension, D₁₁A first order difference matrix in the time dimension is represented,

represents the outer product operator which, in one embodiment of the invention,

is the Kronecker outer product operator.

Wherein the first order difference matrix D in the line dimension can be expressed by the following formula₁₂Conversion to total variation regularization matrix in line dimension:

L₁₂representing the total variation regularization matrix in the line dimension, I₃Representing an identity matrix in the track dimension, D₁₂Representing a first order difference matrix in the line dimension, I₁A unit matrix in the time dimension is represented,

is the Kronecker outer product operator.

Wherein the first order difference matrix D in the track dimension can be expressed by the following formula₁₃Conversion to total variation regularization matrix in the track dimension:

L₁₃representing the total variation regularization matrix in the track dimension, D₁₃Representing a first order difference matrix in the track dimension, I₂Representing identity matrices in line dimensions, I₁A unit matrix in the time dimension is represented,

is the Kronecker outer product operator.

In the embodiment of the invention, the total variation regularization matrix in the time dimension is determined according to the unit matrix in the channel dimension and the unit matrix in the line dimension and the first-order difference matrix in the time dimension, the total variation regularization matrix in the line dimension is determined according to the unit matrix in the channel dimension and the first-order difference matrix in the line dimension and the unit matrix in the time dimension, and the total variation regularization matrix in the channel dimension is determined according to the first-order difference matrix in the channel dimension and the unit matrix in the line dimension and the unit matrix in the time dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 4 shows an implementation flow of step 103 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, and detailed descriptions are as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result, as shown in fig. 4, step 103 is to convert the constructed second-order difference matrices in the time dimension, the line dimension, and the track dimension into gihonov regularization matrices in the time dimension, the line dimension, and the track dimension, respectively, and includes:

step 401, constructing a second-order difference matrix in a time dimension, a line dimension and a channel dimension;

step 402, according to the second-order difference matrix in the time dimension, the line dimension and the lane dimension and the unit matrix in the time dimension, the line dimension and the lane dimension, the second-order difference matrix in the time dimension, the line dimension and the lane dimension is converted into a Gihonov regularization matrix in the time dimension, the line dimension and the lane dimension.

When determining the Gihonov regularization matrix in different dimensions such as time dimension, line dimension and channel dimension, firstly, constructing a second-order difference matrix D of the time dimension₂₁Second order difference matrix D in line dimension₂₂And a second order difference matrix D in the channel dimension₂₃And further respectively converting the second order difference matrix D of the time dimension₂₁Converting into Gihono regularization matrix in time dimension, and converting second order difference matrix D in line dimension₂₂Converting into Gihono regularization matrix in line dimension, and converting into second order difference matrix D in track dimension₂₃And converting into a Gihono regularization matrix on the track dimension.

In the embodiment of the invention, the second-order difference matrixes in the time dimension, the line dimension and the channel dimension are constructed, and the second-order difference matrixes in the time dimension, the line dimension and the channel dimension are respectively converted into the Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension according to the second-order difference matrixes in the time dimension, the line dimension and the channel dimension and the unit matrixes in the time dimension, the line dimension and the channel dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 5 shows an implementation flow of step 402 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, and detailed descriptions are as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result, as shown in fig. 5, step 402, according to the second-order difference matrix in the time dimension, the line dimension, and the track dimension and the unit matrix in the time dimension, the line dimension, and the track dimension, respectively converting the second-order difference matrix in the time dimension, the line dimension, and the track dimension into the genoh regularization matrix in the time dimension, the line dimension, and the track dimension, includes:

step 501, determining a Gihonov regularization matrix on a time dimension according to an identity matrix on a channel dimension, an identity matrix on a line dimension and a second-order difference matrix on the time dimension;

step 502, determining a Gihonov regularization matrix on a line dimension according to an identity matrix on a channel dimension and a second-order difference matrix on the line dimension and an identity matrix on a time dimension;

step 503, determining a gihonov regularization matrix in the track dimension according to the second order difference matrix in the track dimension and the identity matrix in the line dimension, and the identity matrix in the time dimension.

When determining the gihonov regularization matrix in the time dimension, the line dimension and the track dimension, the identity matrix in the time dimension, the line dimension and the track dimension is needed. Specifically, in determining the gihonov regularization matrix in the time dimension, an identity matrix in the lane dimension and an identity matrix in the line dimension need to be utilized; when determining the Gihonov regularization matrix on the line dimension, the unit matrix on the track dimension and the unit matrix on the time dimension are required to be utilized; when determining the Gihono regularization matrix in the track dimension, an identity matrix in the line dimension and an identity matrix in the time dimension are needed.

Wherein the second order difference matrix D in the time dimension can be expressed by the following formula₂₁Converting to a Gihono regularization matrix in the time dimension:

L₂₁representing a Gihonov regularization matrix in the time dimension, I₃Representing an identity matrix in the track dimension, I₂Representing identity matrices in line dimension, D₂₁Representing a second order difference matrix in the time dimension,

is the Kronecker outer product operator.

Wherein the second order difference matrix D in the line dimension can be expressed by the following formula₂₂Converting into a Gihono regularization matrix on a line dimension:

L₂₂representing a Gihonov regularization matrix in a line dimension, I₃Representing an identity matrix in the track dimension, D₂₂Representing a second order difference matrix in the line dimension, I₁A unit matrix in the time dimension is represented,

is the Kronecker outer product operator.

Wherein the second order difference matrix D in the channel dimension can be expressed by the following formula₂₃Converting into a Gihono regularization matrix on a track dimension:

L₂₃gihonov regularization matrix, D, in the representation lane dimension₂₃Representing two in the track dimensionOrder difference matrix, I₂Representing identity matrices in line dimensions, I₁A unit matrix in the time dimension is represented,

is the Kronecker outer product operator.

In the embodiment of the invention, according to the unit matrix on the channel dimension and the unit matrix on the line dimension and the second-order difference matrix on the time dimension, the Gihonov regularization matrix on the time dimension is determined, according to the unit matrix on the channel dimension and the second-order difference matrix on the line dimension and the unit matrix on the time dimension, the Gihonov regularization matrix on the line dimension is determined, and according to the second-order difference matrix on the channel dimension and the unit matrix on the line dimension and the unit matrix on the time dimension, the Gihonov regularization matrix on the channel dimension is determined, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 6 shows an implementation flow of step 104 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, and detailed descriptions are as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result, as shown in fig. 6, step 104 is to construct an objective function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix, and the gihonov regularization matrix, where the method includes:

step 601, constructing a common constraint term aiming at the transverse abrupt wave impedance and the transverse gradual wave impedance according to the reflection coefficient vector;

step 602, constructing a total variation regularization constraint term aiming at the transverse abrupt change wave impedance according to the total variation regularization matrix;

step 603, constructing a Gihonov regularization constraint term aiming at the transverse gradient wave impedance according to the Gihonov regularization matrix;

and step 604, constructing a wave impedance target function according to the common constraint term, the total variation regularization constraint term and the Gihonov regularization constraint term.

When constructing the common constraint term for the transverse abrupt wave impedance and the transverse gradual wave impedance according to the reflection coefficient vector, the common constraint term may be specifically constructed in the following manner:

wherein D is₁₁A first order difference matrix representing the time dimension, so D₁₁x₁、D₁₁x₂Is the difference of wave impedance, and is a reflection coefficient sequence; x is the number of₁Representing the transverse abrupt wave impedance, x₂Representing the transverse gradient wave impedance, r representing a reflection coefficient matrix, and vec (r) being a reflection coefficient vector for converting the reflection coefficient matrix into a single channel according to the sequence of time dimension, line dimension and channel dimension.

In addition, when constructing the total variation regularization constraint term for the transverse abrupt change wave impedance according to the total variation regularization matrix, the total variation regularization constraint term for the transverse abrupt change wave impedance may be specifically constructed in the following manner:

wherein, α_iConstraint coefficient, L, representing total variation regularization constraint term_1iThe matrix is regularized for the total variation in different dimensions (time dimension, line dimension, and trace dimension).

In addition, when constructing the giHonowov regularization constraint term for the transverse gradient wave impedance according to the giHonowov regularization matrix, the geHonowov regularization constraint term for the transverse gradient wave impedance may be specifically constructed according to the following method:

wherein, β_iConstraint coefficient, L, representing a Gihonov regularization constraint term_2iMatrices are regularized for Gihonov in different dimensions (time, line and lane).

So far, after the common constraint term of the transverse abrupt wave impedance and the transverse gradient wave impedance, the gihonov regularization constraint term of the transverse abrupt wave impedance and the gihonov regularization constraint term of the transverse gradient wave impedance are respectively constructed, the objective function of the wave impedance can be constructed according to the common constraint term of the transverse abrupt wave impedance and the transverse gradient wave impedance, the gihonov regularization constraint term of the transverse abrupt wave impedance and the gihonov regularization constraint term of the transverse gradient wave impedance, and specifically, the objective function of the wave impedance can be constructed in the following way:

wherein,

an objective function representing the wave impedance is shown,

a total variation regularization constraint term representing the transverse abrupt wave impedance,

a gehonov regularization constraint term representing a laterally graded wave impedance.

In the embodiment of the invention, a common constraint term aiming at the transverse abrupt change wave impedance and the transverse gradient wave impedance is constructed according to the reflection coefficient vector, a total variation regularization constraint term aiming at the transverse abrupt change wave impedance is constructed according to the total variation regularization matrix, a Gihonov regularization constraint term aiming at the transverse gradient wave impedance is constructed according to the Gihonov regularization matrix, and a target function of the wave impedance is constructed according to the common constraint term, the total variation regularization constraint term and the Gihonov regularization constraint term, so that the continuity of the transverse energy of the wave impedance can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 7 shows an implementation flow of step 105 in the blind inversion method provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, as shown in fig. 7, step 105, determining the wave impedance of the seismic record according to the lateral abrupt change wave impedance and the lateral gradual change wave impedance when the objective function of the wave impedance takes the minimum value, includes:

step 701, determining a transverse abrupt wave impedance and a transverse gradual change wave impedance when an objective function of wave impedance obtains a minimum value by using an alternative direction multiplier;

and step 702, adding the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value to obtain the wave impedance of the seismic record.

Alternative Direction Multipliers (the English full name: alternative Direction methods of Multipliers, ADMM for short) were first proposed by Gloinski & Marrocco and Gabay & Mercier in 1975 and 1976, respectively, and reviewed and proven to be suitable for the large-scale distributed optimization problem by Boyd et al in 2011. The method is a constraint problem optimization method widely used in machine learning, is a simple method for solving a decomposable convex optimization problem, is particularly effective in solving large-scale problems, can equivalently decompose an objective function of an original problem into a plurality of sub-problems which can be solved by using an ADMM algorithm, then solves each sub-problem in parallel, and finally coordinates the sub-problems to obtain a global solution of the original problem.

In the embodiment of the invention, the transverse abrupt change wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value are solved and determined by using the alternative direction multiplier algorithm, after the transverse abrupt change wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value are obtained, the transverse abrupt change wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value are added, and the wave impedance of the seismic record can be obtained, so that the obtained wave impedance has good transverse energy continuity, and the reliability of the wave impedance seismic inversion result is improved.

In the embodiment of the invention, the transverse abrupt change wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value are determined by using the alternative direction multipliers, and the transverse abrupt change wave impedance and the transverse gradient wave impedance when the objective function of the wave impedance obtains the minimum value are added to obtain the wave impedance of the seismic record, so that the continuity of the transverse energy of the wave impedance can be further improved, and the reliability of the wave impedance inversion result is improved.

The embodiment of the invention also provides a blind inversion device, which is described in the following embodiment. Because the principle of solving the problems of the devices is similar to that of a blind inversion method, the implementation of the devices can be referred to the implementation of the method, and repeated details are not repeated.

Fig. 8 shows functional modules of a blind inversion apparatus provided in an embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

referring to fig. 8, each module included in the blind inversion apparatus is used to perform each step in the embodiment corresponding to fig. 1, and specific reference is made to fig. 1 and the related description in the embodiment corresponding to fig. 1, which are not repeated herein. In the embodiment of the present invention, the blind inversion apparatus includes a reflection coefficient determining module 801, a total variation regularization conversion module 802, a gikhonov regularization conversion module 803, an objective function constructing module 804, and a wave impedance determining module 805.

A reflection coefficient determining module 801, configured to determine a reflection coefficient matrix of the seismic record by using sparse pulse inversion, and convert the reflection coefficient matrix into a reflection coefficient vector according to the order of time dimension, line dimension, and trace dimension;

a total variation regularization conversion module 802, configured to convert the constructed first-order difference matrices in the time dimension, the line dimension, and the lane dimension into total variation regularization matrices in the time dimension, the line dimension, and the lane dimension, respectively;

a gihonov regularization conversion module 803, configured to convert the constructed second-order difference matrices in the time dimension, the line dimension, and the lane dimension into gihonov regularization matrices in the time dimension, the line dimension, and the lane dimension, respectively;

a target function constructing module 804, configured to construct a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix, and the gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance;

the wave impedance determination module 805 is configured to determine the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance takes the minimum value.

In the embodiment of the present invention, the reflection coefficient determining module 801 determines a reflection coefficient matrix of the seismic record by using sparse pulse inversion, and converts the reflection coefficient matrix into a reflection coefficient vector according to the order of time dimension, line dimension, and trace dimension; the total variation regularization conversion module 802 converts the constructed first-order difference matrices in the time dimension, the line dimension and the track dimension into total variation regularization matrices in the time dimension, the line dimension and the track dimension respectively; the gihonov regularization conversion module 803 converts the constructed second-order difference matrices in the time dimension, the line dimension and the lane dimension into gihonov regularization matrices in the time dimension, the line dimension and the lane dimension, respectively; the target function construction module 804 constructs a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance; the wave impedance determination module 805 determines the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance takes a minimum value. In the embodiment of the present invention, the total variation regularization conversion module 802 and the gihonov regularization conversion module 803 respectively convert the first-order difference matrix and the second-order difference matrix into the total variation regularization matrix and the gihonov regularization matrix, the objective function construction module 804 respectively performs regularization constraint on the transverse abrupt-change wave impedance and the transverse gradual-change wave impedance by using the total variation regularization and the gihonov regularization, and solves the wave impedance according to the constructed objective function of the wave impedance, so that the continuity of the transverse energy of the wave impedance can be improved, and the reliability of the wave impedance inversion result can be improved.

In an embodiment of the invention, based on the above module structure, the blind inversion apparatus further includes: the device comprises a theoretical reflection coefficient determining module and a seismic record acquiring module.

And the theoretical reflection coefficient determining module is used for determining the theoretical reflection coefficient according to the established wave impedance model.

And the seismic record acquisition module is used for performing convolution on the theoretical reflection coefficient and the Rake wavelets to acquire the seismic record.

In an embodiment of the invention, based on the above module structure, the blind inversion apparatus further includes a noise adding module.

And the noise adding module is used for adding noise with a preset signal-to-noise ratio into the seismic record to obtain the seismic record after the noise is added.

Correspondingly, the reflection coefficient determining module 801 is specifically configured to determine a reflection coefficient matrix of the seismic record to which the noise is added by using sparse pulse inversion, and convert the reflection coefficient matrix into a reflection coefficient vector according to the sequence of time dimension, line dimension, and trace dimension.

Fig. 9 shows functional blocks of a total variation regularization conversion module 802 in a blind inversion apparatus provided in an embodiment of the present invention, and only shows portions related to the embodiment of the present invention for convenience of description, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 9, each unit included in the total variation regularization conversion module 802 is configured to execute each step in the embodiment corresponding to fig. 2, and specifically refer to fig. 2 and the related description in the embodiment corresponding to fig. 2, which is not repeated herein. In the embodiment of the present invention, the total variation regularization conversion module 802 includes a first-order difference matrix construction unit 901 and a total variation regularization conversion unit 902.

A first-order difference matrix constructing unit 901 configured to construct a first-order difference matrix in a time dimension, a line dimension, and a channel dimension;

the total variation regularization conversion unit 902 is configured to convert the first-order difference matrices in the time dimension, the line dimension, and the track dimension into total variation regularization matrices in the time dimension, the line dimension, and the track dimension, respectively, according to the first-order difference matrices in the time dimension, the line dimension, and the track dimension, and the unit matrices in the time dimension, the line dimension, and the track dimension.

In the embodiment of the present invention, the first-order difference matrix constructing unit 901 constructs a first-order difference matrix in a time dimension, a line dimension, and a track dimension; the total variation regularization conversion unit 902 converts the first-order difference matrices in the time dimension, the line dimension and the track dimension into total variation regularization matrices in the time dimension, the line dimension and the track dimension respectively according to the first-order difference matrices in the time dimension, the line dimension and the track dimension and the unit matrices in the time dimension, the line dimension and the track dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result can be improved.

Fig. 10 shows functional modules of a total variation regularization conversion unit 902 in a blind inversion apparatus provided in an embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 10, each sub-unit included in the total variation regularization conversion unit 902 is configured to execute each step in the embodiment corresponding to fig. 3, specifically refer to fig. 3 and the related description in the embodiment corresponding to fig. 3, and details are not repeated here. In the embodiment of the present invention, the total variation regularization conversion unit 902 includes a time dimension total variation regularization conversion subunit 1001, a line dimension total variation regularization conversion subunit 1002, and a lane dimension total variation regularization conversion subunit 1003.

A time dimension total variation regularization conversion subunit 1001, configured to determine a total variation regularization matrix in a time dimension according to a unit matrix in a channel dimension and a unit matrix in a line dimension, and a first-order difference matrix in the time dimension;

a line dimension total variation regularization conversion subunit 1002, configured to determine a total variation regularization matrix in a line dimension according to an identity matrix in a lane dimension and a first-order difference matrix in the line dimension, and an identity matrix in a time dimension;

the track dimension total variation regularization conversion subunit 1003 is configured to determine a total variation regularization matrix in the track dimension according to the first-order difference matrix in the track dimension and the unit matrix in the line dimension, and the unit matrix in the time dimension.

In the embodiment of the present invention, the time dimension total variation regularization conversion subunit 1001 determines the time dimension total variation regularization matrix according to the channel dimension unit matrix and the line dimension unit matrix and the time dimension first-order difference matrix, the line dimension total variation regularization conversion subunit 1002 determines the line dimension total variation regularization matrix according to the channel dimension unit matrix and the line dimension first-order difference matrix and the time dimension unit matrix, and the channel dimension total variation regularization conversion subunit 1003 determines the channel dimension total variation regularization matrix according to the channel dimension first-order difference matrix and the line dimension unit matrix and the time dimension unit matrix, so as to further improve the continuity of the wave impedance transverse energy and improve the reliability of the wave impedance inversion result.

Fig. 11 illustrates functional modules of the gihonov regularization conversion module 803 in the blind inversion apparatus provided in the embodiment of the present invention, and for convenience of description, only portions related to the embodiment of the present invention are illustrated, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 11, each unit included in the givenov regularization conversion module 803 is configured to execute each step in the embodiment corresponding to fig. 4, specifically refer to fig. 4 and the related description in the embodiment corresponding to fig. 4, and details are not repeated here. In the embodiment of the present invention, the gihonov regularization conversion module 803 includes a second-order difference matrix construction unit 1101 and a gihonov regularization conversion unit 1102.

A second order difference matrix constructing unit 1101 configured to construct a second order difference matrix in a time dimension, a line dimension, and a lane dimension;

the gihonov regularization conversion unit 1102 is configured to convert the second-order difference matrices in the time dimension, the line dimension, and the lane dimension into gihonov regularization matrices in the time dimension, the line dimension, and the lane dimension, respectively, according to the second-order difference matrices in the time dimension, the line dimension, and the lane dimension, and the unit matrices in the time dimension, the line dimension, and the lane dimension.

In the embodiment of the present invention, the second order difference matrix constructing unit 1101 constructs a second order difference matrix in a time dimension, a line dimension, and a lane dimension; the gihonov regularization conversion unit 1102 converts the second-order difference matrices in the time dimension, the line dimension and the channel dimension into the gihonov regularization matrices in the time dimension, the line dimension and the channel dimension respectively according to the second-order difference matrices in the time dimension, the line dimension and the channel dimension and the unit matrices in the time dimension, the line dimension and the channel dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result can be improved.

Fig. 12 shows functional modules of a gihonov regularization conversion unit 1102 in a blind inversion apparatus provided in an embodiment of the present invention, and for convenience of description, only parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 12, each sub-unit included in the givenov regularization conversion unit 1102 is configured to perform each step in the embodiment corresponding to fig. 5, specifically please refer to fig. 5 and the related description in the embodiment corresponding to fig. 5, which is not repeated herein. In the embodiment of the present invention, the gihonov regularization conversion unit 1102 includes a time dimension gihonov regularization conversion subunit 1201, a line dimension gihonov regularization conversion subunit 1202, and a lane dimension gihonov regularization conversion subunit 1203.

A time dimension gihonov regularization conversion subunit 1201, configured to determine a gihonov regularization matrix in a time dimension according to an identity matrix in a lane dimension and an identity matrix in a line dimension, and a second-order difference matrix in the time dimension;

a line dimension gihonov regularization conversion subunit 1202, configured to determine a gihonov regularization matrix in a line dimension according to an identity matrix in a lane dimension and a second-order difference matrix in the line dimension, and an identity matrix in a time dimension;

a track dimension gihonov regularization conversion subunit 1203, configured to determine a gihonov regularization matrix in the track dimension according to the second-order difference matrix in the track dimension and the identity matrix in the line dimension, and the identity matrix in the time dimension.

In the embodiment of the present invention, the time dimension gihonov regularization conversion subunit 1201 determines a gihonov regularization matrix in a time dimension according to an identity matrix in a lane dimension and an identity matrix in a line dimension and a second order difference matrix in the time dimension; the line dimension gihonov regularization conversion subunit 1202 determines a gihonov regularization matrix in the line dimension according to the unit matrix in the lane dimension and the second order difference matrix in the line dimension and the unit matrix in the time dimension; the track dimension gihonov regularization conversion subunit 1203 determines a gihonov regularization matrix in the track dimension according to the second-order difference matrix in the track dimension and the unit matrix in the line dimension and the unit matrix in the time dimension, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result can be improved.

Fig. 13 shows functional modules of the objective function constructing module 804 in the blind inversion apparatus provided by the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 13, each unit included in the objective function building module 804 is configured to execute each step in the embodiment corresponding to fig. 6, specifically refer to fig. 6 and the related description in the embodiment corresponding to fig. 6, and are not described herein again. In the embodiment of the present invention, the objective function constructing module 804 includes a common constraint item constructing unit 1301, a total variation regularization constraint item constructing unit 1302, a gikhonov regularization constraint item constructing unit 1303, and an objective function constructing unit 1304.

A common constraint term construction unit 1301, configured to construct a common constraint term for the transverse abrupt wave impedance and the transverse gradual wave impedance according to the reflection coefficient vector;

a total variation regularization constraint term construction unit 1302, configured to determine a total variation regularization constraint term for the transverse abrupt change wave impedance according to the total variation regularization matrix;

the Gihonov regularization constraint term construction unit 1303 is used for determining a Gihonov regularization constraint term aiming at the transverse gradient wave impedance according to the Gihonov regularization matrix;

and an objective function constructing unit 1304, configured to construct an objective function of the wave impedance according to the common constraint term, the total variation regularization constraint term, and the gihonov regularization constraint term.

In the embodiment of the present invention, the common constraint term construction unit 1301 constructs a common constraint term for the transverse abrupt-change wave impedance and the transverse gradual-change wave impedance according to the reflection coefficient vector; the total variation regularization constraint term construction unit 1302 determines a total variation regularization constraint term for the transverse abrupt change wave impedance according to the total variation regularization matrix; the gihonov regularization constraint term construction unit 1303 determines a gihonov regularization constraint term for the transverse gradient wave impedance according to the gihonov regularization matrix; the target function construction unit 1304 constructs a wave impedance target function according to the common constraint term, the total variation regularization constraint term and the Gihonov regularization constraint term, so that the continuity of the wave impedance transverse energy can be further improved, and the reliability of the wave impedance inversion result is improved.

Fig. 14 shows functional blocks of the wave impedance determining module 805 in the blind inversion apparatus provided in the embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which are detailed as follows:

in an embodiment of the present invention, in order to further improve the continuity of the wave impedance lateral energy and improve the reliability of the wave impedance inversion result, referring to fig. 14, each unit included in the wave impedance determining module 805 is configured to execute each step in the embodiment corresponding to fig. 7, specifically refer to fig. 7 and the related description in the embodiment corresponding to fig. 7, and are not described herein again. In the embodiment of the present invention, the wave impedance determination module 805 includes a determination unit 1401 and a summation unit 1402.

A determining unit 1401 for determining the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value by using the alternating direction multiplier;

and an adding unit 1402, configured to add the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance takes a minimum value, so as to obtain the wave impedance of the seismic record.

In the embodiment of the present invention, the determining unit 1401 determines the transverse abrupt wave impedance and the transverse gradual change wave impedance when the objective function of the wave impedance obtains the minimum value by using the alternative direction multiplier; the addition unit 1402 adds the transverse abrupt wave impedance and the transverse gradual change wave impedance when the objective function of the wave impedance takes the minimum value to obtain the wave impedance of the seismic record, so that the continuity of the transverse energy of the wave impedance can be further improved, and the reliability of the inversion result of the wave impedance can be improved.

In summary, in the embodiments of the present invention, a reflection coefficient matrix of a seismic record is determined by sparse pulse inversion, and the reflection coefficient matrix is converted into a reflection coefficient vector according to the order of time dimension, line dimension, and trace dimension; respectively converting the constructed first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension; respectively converting the constructed second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension; constructing a target function of the wave impedance according to the reflection coefficient vector, the total variation regularization matrix and the Gihonov regularization matrix; the wave impedance comprises a transverse abrupt wave impedance and a transverse gradual wave impedance; and determining the wave impedance of the seismic record according to the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value. According to the embodiment of the invention, the first-order difference matrix and the second-order difference matrix are respectively converted into the total variation regularization matrix and the Gihonov regularization matrix, the total variation regularization and the Gihonov regularization are respectively utilized to carry out regularization constraint on the transverse abrupt wave impedance and the transverse gradual wave impedance, and the wave impedance is solved according to the constructed objective function of the wave impedance, so that the continuity of the transverse energy of the wave impedance can be improved, and the reliability of the wave impedance inversion result is improved.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A blind inversion method, comprising:

2. The method of claim 1, wherein converting the constructed first order difference matrices in the time dimension, line dimension, and track dimension to total variation regularization matrices in the time dimension, line dimension, and track dimension, respectively, comprises:

constructing a first-order difference matrix on a time dimension, a line dimension and a track dimension;

and respectively converting the first-order difference matrixes in the time dimension, the line dimension and the track dimension into total variation regularization matrixes in the time dimension, the line dimension and the track dimension according to the first-order difference matrixes in the time dimension, the line dimension and the track dimension and the unit matrixes in the time dimension, the line dimension and the track dimension.

3. The method of claim 1, wherein converting the first-order difference matrices in the time dimension, the line dimension, and the track dimension into total variation regularization matrices in the time dimension, the line dimension, and the track dimension, respectively, based on the first-order difference matrices in the time dimension, the line dimension, and the track dimension, and the identity matrices in the time dimension, the line dimension, and the track dimension, comprises:

determining a total variation regularization matrix in a time dimension according to an identity matrix in a channel dimension and an identity matrix in a line dimension and a first-order difference matrix in the time dimension;

determining a total variation regularization matrix on a line dimension according to an identity matrix on a channel dimension and a first-order difference matrix on a line dimension and an identity matrix on a time dimension;

and determining a total variation regularization matrix on the track dimension according to the first-order difference matrix on the track dimension and the unit matrix on the line dimension and the unit matrix on the time dimension.

4. The method of claim 1, wherein converting the constructed second order difference matrices in the time dimension, the line dimension, and the lane dimension into gihonov regularization matrices in the time dimension, the line dimension, and the lane dimension, respectively, comprises:

constructing a second-order difference matrix on a time dimension, a line dimension and a channel dimension;

and respectively converting the second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension according to the second-order difference matrixes in the time dimension, the line dimension and the channel dimension and the unit matrixes in the time dimension, the line dimension and the channel dimension.

5. The method of claim 1, wherein converting the second order difference matrices in the time dimension, the line dimension, and the lane dimension to gikhonov regularization matrices in the time dimension, the line dimension, and the lane dimension, respectively, based on the second order difference matrices in the time dimension, the line dimension, and the lane dimension, and the identity matrices in the time dimension, the line dimension, and the lane dimension, comprises:

determining a Gihonov regularization matrix on a time dimension according to an identity matrix on a channel dimension and an identity matrix on a line dimension and a second-order difference matrix on the time dimension;

determining a Gihonov regularization matrix on a line dimension according to an identity matrix on a channel dimension and a second-order difference matrix on the line dimension and an identity matrix on a time dimension;

and determining a Gihonov regularization matrix on the track dimension according to the second-order difference matrix on the track dimension and the identity matrix on the line dimension and the identity matrix on the time dimension.

6. The method of claim 1, wherein constructing the objective function of the wave impedance from the reflection coefficient vector and the total variation regularization matrix, and the Gihonov regularization matrix comprises:

constructing a common constraint term aiming at the transverse abrupt wave impedance and the transverse gradual wave impedance according to the reflection coefficient vector;

constructing a total variation regularization constraint term aiming at the transverse abrupt change wave impedance according to the total variation regularization matrix;

constructing a Gihonov regularization constraint term aiming at the transverse gradient wave impedance according to the Gihonov regularization matrix;

and constructing an objective function of the wave impedance according to the common constraint term, the total variation regularization constraint term and the Gihonov regularization constraint term.

7. The method of claim 1, wherein determining the wave impedance of the seismic recording from the laterally abrupt wave impedance and the laterally gradual wave impedance at which the objective function of wave impedance assumes a minimum comprises:

determining the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value by utilizing the alternative direction multipliers;

and adding the transverse abrupt wave impedance and the transverse gradual change wave impedance when the objective function of the wave impedance obtains the minimum value to obtain the wave impedance of the seismic record.

8. A blind inversion apparatus, comprising:

9. The apparatus of claim 8, wherein the total variation regularization conversion module comprises:

the first-order difference matrix construction unit is used for constructing a first-order difference matrix on a time dimension, a line dimension and a channel dimension;

and the total variation regularization conversion unit is used for respectively converting the first-order difference matrixes in the time dimension, the line dimension and the channel dimension into total variation regularization matrixes in the time dimension, the line dimension and the channel dimension according to the first-order difference matrixes in the time dimension, the line dimension and the channel dimension and the unit matrixes in the time dimension, the line dimension and the channel dimension.

10. The apparatus of claim 9, wherein the total variation regularization conversion unit comprises:

the time dimension total variation regularization conversion subunit is used for determining a total variation regularization matrix in a time dimension according to the unit matrix in the channel dimension and the unit matrix in the line dimension and a first-order difference matrix in the time dimension;

the line dimension total variation regularization conversion subunit is used for determining a total variation regularization matrix on a line dimension according to the unit matrix on the channel dimension and the first-order difference matrix on the line dimension and the unit matrix on the time dimension;

and the track dimension total variation regularization conversion subunit is used for determining a total variation regularization matrix on the track dimension according to the first-order difference matrix on the track dimension and the unit matrix on the line dimension and the unit matrix on the time dimension.

11. The apparatus of claim 8, wherein the gihonov regularization transformation module comprises:

the second-order difference matrix construction unit is used for constructing a second-order difference matrix in a time dimension, a line dimension and a channel dimension;

and the Gihonov regularization conversion unit is used for converting the second-order difference matrixes in the time dimension, the line dimension and the channel dimension into Gihonov regularization matrixes in the time dimension, the line dimension and the channel dimension respectively according to the second-order difference matrixes in the time dimension, the line dimension and the channel dimension and the unit matrixes in the time dimension, the line dimension and the channel dimension.

12. The apparatus of claim 11, wherein the gihonov regularization conversion unit comprises:

the time dimension Gihonov regularization conversion subunit is used for determining a Gihonov regularization matrix on a time dimension according to an identity matrix on a channel dimension and an identity matrix on a line dimension and a second-order difference matrix on the time dimension;

the line dimension Gihonov regularization conversion subunit is used for determining a Gihonov regularization matrix on a line dimension according to an identity matrix on a channel dimension, a second-order difference matrix on the line dimension and an identity matrix on a time dimension;

and the track dimension Gihonov regularization conversion subunit is used for determining a Gihonov regularization matrix on the track dimension according to the second-order difference matrix on the track dimension and the identity matrix on the line dimension and the identity matrix on the time dimension.

13. The apparatus of claim 8, wherein the objective function construction module comprises:

the common constraint term construction unit is used for constructing common constraint terms aiming at the transverse abrupt change wave impedance and the transverse gradual change wave impedance according to the reflection coefficient vector;

the total variation regularization constraint term construction unit is used for determining a total variation regularization constraint term aiming at the transverse abrupt change wave impedance according to the total variation regularization matrix;

the Gihonov regularization constraint item construction unit is used for determining a Gihonov regularization constraint item aiming at the transverse gradient wave impedance according to the Gihonov regularization matrix;

and the object function constructing unit is used for constructing an object function of the wave impedance according to the common constraint term, the total variation regularization constraint term and the Gihonov regularization constraint term.

14. The apparatus of claim 8, wherein the wave impedance determination module comprises:

a determining unit for determining the transverse abrupt wave impedance and the transverse gradual wave impedance when the objective function of the wave impedance obtains the minimum value by using the alternative direction multiplier;

and the adding unit is used for adding the transverse abrupt change wave impedance and the transverse gradual change wave impedance when the objective function of the wave impedance obtains the minimum value to obtain the wave impedance of the seismic record.

15. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 7 when executing the computer program.

16. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the method of any one of claims 1 to 7.