CN114994762A - A Sparse Domain Seismic Diffraction Wave Separation Method - Google Patents

Abstract

The invention relates to the technical field of seismic exploration multiple suppression, in particular to a sparse domain seismic diffracted wave separation method. And then, iterative decomposition is carried out on the diffracted wave signals by utilizing a matching tracking algorithm to obtain sparse diffracted wave signal atoms, finally, the least square technology is adopted to solve atomic coefficients of the diffracted wave signals obtained by decomposition, and separated diffracted wave data are obtained by weighted superposition of the diffracted wave signal atoms and the coefficients thereof. The method is simple, high in calculation efficiency, good in noise immunity, good in application effect on low signal-to-noise ratio data, and capable of solving the problem of separation integrity of the top point and the two wings of the diffracted wave better, and the diffracted wave signal is decomposed more thoroughly, so that the method is one of conventional processes for seismic data processing, wide in application range and high in popularization value.

Description

A Sparse Domain Seismic Diffraction Wave Separation Method

technical field

The invention relates to the technical field of multiple wave suppression in seismic exploration, and more particularly, to a method for separating seismic diffraction waves in a sparse domain.

Background technique

With the deepening of exploration and development, the significance of small-scale and high-resolution seismic imaging in high-precision seismic exploration has become increasingly important. The focus of seismic exploration is to study information related to laterally heterogeneous bodies such as faults, channels, fractures, pinch-out points, and dissolved vug-type reservoirs. These target geological bodies have small scales, irregular shapes, and strong spatial heterogeneity. Different from reflected waves, seismic waves encounter these small-scale geological bodies in the process of propagation, forming a secondary source to generate a large number of diffracted waves. Compared with reflected waves, these diffracted waves have different dynamics and kinematics. Significant differences. Therefore, making good use of diffraction wave imaging is of great significance for the exploration of lateral heterogeneity, and at the same time, it can better identify structural development zones and find favorable traps. The amplitude of the diffracted wave is much weaker than that of the reflected wave, and it is usually submerged in the reflected wave information and difficult to identify. Conventional seismic processing mainly takes the primary reflected wave as the target object, and usually produces a certain amount of diffracted waves other than the continuous reflection during the noise suppression process. damage, so a separate imaging technique for diffracted waves was developed.

Diffraction wave separation imaging can be divided into "direct method" and "indirect method" diffraction wave imaging. The former is to directly image the diffracted wave according to the characteristic difference between the reflected wave and the diffracted wave during the migration process or after the migration. , the latter is to separate diffracted waves in the gather wave field before migration, and then perform migration imaging on the separated diffracted waves. The "direct method" has the characteristics of high computational efficiency, while the pre-migration gather diffraction wave separation technology mostly adopts the iterative inversion method, and the amount of data for wave field separation is very large, the diffraction wave separation takes a long time, and the computational efficiency is low. . In addition, the energy of the diffracted waves separated by the diffracted waves of the gathers before migration is relatively weak. In the low signal-to-noise ratio area, the diffracted wave signals are often difficult to identify. How to separate the reliable weak diffracted waves from the submerged strong reflection data? Signal has always been a technical difficulty in diffraction wave separation imaging.

The diffracted wave separation method of gather before migration mainly utilizes the difference of kinematics and dynamic characteristics of reflected wave and diffracted wave. The dip filtering or Radon transform filtering technology of seismic wave travel time difference is difficult to solve the simultaneous separation of the diffraction wave vertex and the two wings. In addition, for low signal-to-noise ratio data, due to the influence of noise, the diffraction wave separation effect is often Not ideal.

For example, the prior art discloses a method for separating diffracted waves based on the sparse Radon transform of the near-track fringe. The high-efficiency sparse Radon transform is used to separate the wave field of the pre-stack gathers containing diffracted waves, and the separated reflected waves are obtained. and diffracted waves, although this separation method can effectively reduce the near-channel edge effect of Radon transform and improve the degree of separability of reflected and diffracted waves in the Radon domain, but the separation method still has low noise resistance and poor separation effect. good question.

SUMMARY OF THE INVENTION

In order to overcome the technical problem that the diffraction wave separation technology described in the above-mentioned prior art is affected by noise and the diffraction wave separation effect is not very ideal, the present invention provides a sparse system with high noise resistance and relatively thorough diffraction wave signal separation. Domain Seismic Diffraction Wave Separation Method.

In order to solve the above-mentioned technical problems, the technical solution adopted in the present invention is: a method for separating seismic diffraction waves in a sparse domain, the method comprising:

Step 1: Use the theory that the time-distance curve of the diffracted wave is a hyperbola, and the amplitude energy is inversely proportional to the propagation distance to construct diffracted wave signals with different curvatures, and generate a diffracted wave signal atom library;

Step 2: Use the matching pursuit algorithm to iteratively decompose and search for signal atoms in the diffraction wave signal atom library obtained in step 1 to obtain sparse diffraction wave signal atoms; and use the least squares method for each iteration in the matching pursuit decomposition process Solve the coefficients of sparse signal atoms;

Step 3: Obtain the separated diffracted wave data by weighting and superimposing the diffracted wave signal atoms and their coefficients obtained in the second step.

The invention adopts a complete diffraction wave signal atom library to decompose the diffraction wave in the sparse domain, and has good anti-noise performance, and has a significant application effect on data with a low signal-to-noise ratio, and can also better solve the separation of the diffraction wave apex and the two wings. For the integrity problem, compared with the traditional matching pursuit technology, the iterative matching pursuit is used to solve the coefficients of the diffracted wave signal atoms, and the diffracted wave signal is decomposed more thoroughly.

Further, in the step 1, before generating the diffraction wave signal atom library, input data and rearrange the data, specifically: input shot collection or common center point gather, and input shot collection or common center point gather. The gathers are re-arranged according to the offset distance to obtain a common offset gather d; meanwhile, the root mean square velocity field v _rms is input.

Further, in the step 1, after the data is input and the data is rearranged, the parameters in the diffractive wave separation process are set, including the curvature range [σ ₁ ～σ ₂ ] of the diffracted wave signal atoms, and the curvature interval Δσ , the least squares solution damping coefficient ε, the minimum acceptable correlation coefficient threshold α _acp between the diffracted wave signal atoms and the actual data, the maximum acceptable residual percentage ζ _acp , and the maximum number of iterations N of matching pursuit.

Further, in the step 1, the atomic library Atom(σ, t ₀ , x) of the diffracted wave signal is generated, wherein,

其中t₀为绕射波顶点的双程反射时间，x为水平方向与双曲线顶点处的距离，i为曲率索引，i＝1,2,...,N。The time-distance curve calculation formula of the diffracted wave signal adopts the hyperbolic formula,

where t ₀ is the two-way reflection time at the vertex of the diffracted wave, x is the distance between the horizontal direction and the vertex of the hyperbola, i is the curvature index, i=1,2,...,N.

Further, in the step 1, the atomic library Atom(σ, t ₀ , x) of the diffracted wave signal is generated, wherein,

其中A₀、v分别为双曲线顶点处的地震振幅及均方根速度，x为水平方向与双曲线顶点处的距离，a,b为与振幅相关的待定系数。The amplitude of the diffracted wave is calculated using the formula

where A ₀ and v are the seismic amplitude and RMS velocity at the apex of the hyperbola, respectively, x is the distance between the horizontal direction and the apex of the hyperbola, and a, b are the undetermined coefficients related to the amplitude.

Further, generating the atomic library Atom(σ,t ₀ ,x) of the diffracted wave signal in the step 1 also includes the following operations:

The least squares fitting is performed according to the searched diffracted wave signal, namely ||A(x)-A _real (x)||→min, A _real (x) is the amplitude value of the actual diffracted wave data.

Further, the specific operation of searching for diffracted wave signal atoms in the second step is:

(I), set input data d;

Initialize residual data r ₀ =d, output diffraction wave seismic data d _s =0, searched diffraction wave atom library Atom _searched =Φ, iteration number k = 0;

Time sample loop processing;

Seismic trace number loop processing;

(II), matching tracking loop: k≤N;

(III), calculate the correlation coefficient between the atomic library Atom(σ,t ₀ ,x) of the diffracted wave signal and the data residual r _k

(IV), calculate the maximum index value m=argmax(C(σ _i )) of the correlation coefficient C(σ i ), i=1,2,...,N, if C(σ _m )>α _acp then enter the step (V), otherwise end the matching tracking loop;

表示移除原子运算。(V), add the searched diffracted wave signal atoms to the searched atom library, Atom _searched = Atom _searched ∪ Atom(σ _m , t ₀ , x), and add the atoms from the next iteration of the atom library exclude,

Represents the removal of atomic operations.

Further, in the second step, the coefficient of solving the sparse signal atom is to solve the global optimization coefficient of the diffracted wave atom, and the equation used is an overdetermined equation.

Further, the specific operation of solving the coefficients of the sparse signal atoms in the second step is:

Let L=Atom _searched , use the least square method to obtain the global optimization coefficient y _searched =(L ^T L+εI) ^-1 L ^T d, where T represents matrix transposition, and I represents unit pair Angular matrix, ε represents the least squares solution damping coefficient, d represents the common offset gather.

Further, the specific operation of the step 3 is:

Reconstructed diffracted seismic data d _s =Ly _searched ;

如果ζ>ζ_acp，令k＝k+1，则进入步骤(III)，否则结束匹配追踪循环；Update the residual r _k+1 =dd _s , calculate the percentage coefficient of the residual

If ζ> _ζacp , let k=k+1, then enter step (III), otherwise end the matching tracking loop;

After the matching tracking loop ends;

The seismic trace cycle processing ends;

The time sample loop processing ends;

Diffraction wave seismic data d _s is output.

Compared with the prior art, the beneficial effects of the present invention are:

1) In the present invention, the present invention uses a complete diffraction wave signal atom library to perform diffraction wave decomposition in the sparse domain, so its anti-noise performance is very good, and it has a significant application effect for low signal-to-noise ratio data, and can also solve the problem well. The integrity of the separation between the diffraction wave apex and the two wings;

2) Compared with the traditional matching tracking technology, the present invention adopts iterative matching tracking to solve the coefficient of the diffracted wave signal atom, and the diffraction wave signal is decomposed more thoroughly;

3) The overall method of the present invention is simple and the calculation efficiency is high. As one of the conventional procedures of seismic data processing, it has a wide range of applications and is worthy of popularization and application.

Description of drawings

Fig. 1 is the flow chart of the present invention's sparse domain seismic diffraction wave separation method;

Fig. 2 is the concrete flow chart of sparse domain seismic diffraction wave separation method of the present invention;

Fig. 3 is a noise-free data separation effect diagram of the sparse domain seismic diffraction wave separation method of the present invention;

Fig. 4 is the separation effect diagram when the diffraction wave signal and the noise signal-to-noise ratio of the sparse domain seismic diffraction wave separation method of the present invention are -6db;

Fig. 5 is a diagram showing the separation effect of the method for separating the diffracted wave in the sparse domain of the present invention when the signal-to-noise ratio of the diffracted wave signal to the noise is -20db.

Detailed ways

The accompanying drawings are for illustrative purposes only, and should not be construed as limitations on this patent; in order to better illustrate the present embodiment, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; for those skilled in the art It is understandable to the artisan that certain well-known structures and descriptions thereof may be omitted from the drawings. The positional relationships described in the drawings are only for exemplary illustration, and should not be construed as a limitation on the present patent.

The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms “upper”, “lower”, “left” and “right” The orientation or positional relationship indicated by "long" and "short" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the indicated device or element must have a specific Therefore, the terms describing the positional relationship in the accompanying drawings are only used for exemplary illustration, and should not be construed as a limitation on this patent. For those of ordinary skill in the art, they can understand according to specific circumstances. The specific meanings of the above terms.

Below by specific embodiment, and in conjunction with accompanying drawing, the technical scheme of the present invention is further described in detail:

Example 1

As shown in Figure 1-Figure 2, a sparse domain seismic diffraction wave separation method includes:

Step 1: Use the theory that the time-distance curve of the diffracted wave is a hyperbola, and the amplitude energy is inversely proportional to the propagation distance to construct diffracted wave signals with different curvatures, and generate a diffracted wave signal atom library;

Step 2: Use the matching pursuit algorithm to iteratively decompose and search for signal atoms in the diffraction wave signal atom library obtained in step 1, and obtain sparse diffraction wave signal atoms; and use the least squares method for each iteration in the matching pursuit decomposition process Solve the coefficients of sparse signal atoms;

Step 3: Obtain the separated diffracted wave data by weighting and superimposing the diffracted wave signal atoms and their coefficients obtained in the second step.

The specific operations of the above-mentioned step 1 are: input data and data rearrangement: re-arrange the input shot sets or common center point gathers according to the offset distance, and obtain the common offset distance gather d, where the shot set and the common center point gather d. The gathers are all existing technical terms, and will not be repeated here; at the same time, the root mean square velocity field v _rms is input; the main purpose of data rearrangement is to make full use of the diffracted wave signal with good hyperbolic properties in the common offset domain It is convenient to use the atomic library of the hyperbolic diffracted wave signal for matching pursuit decomposition, and the root mean square velocity field is used for the subsequent amplitude fitting of the atomic library of the diffracted wave signal;

Set the parameters in the process of diffracted wave separation: the curvature range of the diffracted wave signal atoms [σ ₁ ～σ ₂ ], the curvature interval Δσ, the least squares solution damping coefficient ε, and the minimum compatibility between the diffracted wave signal atoms and the actual data. Acceptance correlation coefficient threshold α _acp , maximum acceptable residual error percentage ζ _acp , due to the influence of the absolute magnitude of the amplitude, the residual error definition of the present invention adopts a more widely adaptable relative percentage to measure the size of the residual error energy, and matches the maximum iteration of tracking times N;

其中t₀为绕射波顶点的双程反射时间，x为水平方向与双曲线顶点处的距离，i为曲率索引，i＝1,2,...,N；由于绕射波信号的能量明显随着距离绕射点的横向距离增大而衰减，传统的基于双曲Radon变换的绕射波分离技术一般不考虑振幅的横向变化，其绕射波能量分离存在一定程度的残留，本发明采用近似的绕射波振幅横向变化关系，绕射波的振幅随距离定点的绝对距离关系采用公式

其中A₀、v分别为双曲线顶点处的地震振幅及均方根速度，a,b为与振幅相关的待定系数，根据搜索到的绕射波信号进行最小二乘拟合，即：||A(x)-A_real(x)||→min，A_real(x)为实际绕射波数据的振幅值；Generate the atomic library Atom(σ,t ₀ ,x) of the diffracted wave signal. Each atom in the atomic library of the diffracted wave signal is composed of a hyperbola travel time and corresponding amplitude, and only has a value at the position of the hyperbola, where , the time-distance curve calculation formula of the diffracted wave signal adopts the hyperbolic formula,

where t ₀ is the two-way reflection time at the vertex of the diffracted wave, x is the distance between the horizontal direction and the vertex of the hyperbola, i is the curvature index, i=1,2,...,N; due to the energy of the diffracted wave signal Obviously, the attenuation decreases with the increase of the lateral distance from the diffraction point. The traditional diffraction wave separation technology based on hyperbolic Radon transform generally does not consider the lateral change of the amplitude, and the diffraction wave energy separation has a certain degree of residue. The present invention The approximate lateral variation relationship of the diffracted wave amplitude is adopted, and the relationship between the amplitude of the diffracted wave and the absolute distance from the fixed point adopts the formula

where A ₀ and v are the seismic amplitude and root mean square velocity at the apex of the hyperbola, respectively, a and b are the undetermined coefficients related to the amplitude, and the least squares fitting is performed according to the searched diffracted wave signal, namely: || A(x)-A _real (x)||→min, A _real (x) is the amplitude value of the actual diffraction wave data;

The specific operation of searching for diffracted wave signal atoms in the above step 2 is as follows:

(I), set the initial parameters:

Initialize residual data r ₀ =d, output data d _s =0, searched diffraction wave atom library Atom _searched =Φ, iteration number k = 0;

Time sample loop processing;

Seismic trace number loop processing;

(II), matching tracking loop: k≤N;

(III), calculate the correlation coefficient between the atomic library Atom(σ,t ₀ ,x) of the diffracted wave signal and the data residual r _k

(IV), calculate the index value with the largest correlation coefficient m=argmax(C(σ _i )), i=1,2,...,N, if C(σ _m )>α _acp , then enter step (V), Otherwise, end the matching tracking loop;

表示移除原子运算，每个原子不会被重复搜索利用。(V), add the searched diffracted wave signal atoms to the searched atom library, Atom _searched = Atom _searched ∪ Atom(σ _m , t ₀ , x), and add the atoms from the next iteration of the atom library exclude,

Indicates that atomic operations are removed, and each atom will not be used by repeated searches.

The specific operations for solving the coefficients of the sparse signal atoms in the above step 2 are:

Let L=Atom _searched , compared to the sparse diffracted wave atoms, the sampling points of the diffracted wave data are much larger than the number, so the equation for solving the global optimization coefficient of the diffracted wave atom is an overdetermined equation, which is obtained by the least squares method The global optimization coefficient solution of the searched diffracted wave atoms is expressed as: y _searched = (L ^T L+εI) ^-1 L ^T d, where T represents matrix transpose, I represents unit diagonal matrix, and ε represents least squares Solve the damping coefficient, d represents the common offset gather;

The specific operations of the third step above are as follows:

Reconstructed diffracted seismic data d _s =Ly _searched ;

如果ζ>ζ_acp，令k＝k+1，则进入步骤(III)，否则结束匹配追踪循环；Update the residual r _k+1 =dd _s , calculate the percentage coefficient of the residual

If ζ> _ζacp , let k=k+1, then enter step (III), otherwise end the matching tracking loop;

The matching tracking loop ends;

The seismic trace cycle processing ends;

The time sample loop processing ends.

Diffraction wave seismic data d _s is output.

The invention adopts the complete diffraction wave signal atom library to decompose the diffraction wave in the sparse domain, and has good anti-noise performance, and has a significant application effect for low signal-to-noise ratio data, and can also better solve the diffraction wave apex and the two wings. The separation integrity problem makes the diffraction wave signal separation more complete. In addition, compared with the traditional matching tracking technology, the present invention adopts iterative matching tracking to solve the coefficients of the diffraction wave signal atoms, and the diffraction wave signal is decomposed more thoroughly. The overall method of the invention is simple and the calculation efficiency is high. As one of the conventional processes of seismic data processing, the invention has a wide range of applications and is worthy of application and promotion.

The superiority of the method of the present invention is described below through specific data experiments, and the specific operations are as follows: the experimental data of the present invention adopts synthetic seismic data, the track spacing of the data is 12.5m, and there are 8 groups of adjacent diffraction waves with a certain spacing in the data, The diffraction spacing changes are 12.5m, 25m, 37.5m, 50m, 62.5m, 75m, 87.5m, and 100m, respectively. The reflection coefficient at the vertex of the left diffracted wave is 0.0023, and the reflection coefficient at the vertex of the right diffracted wave is 0.0028. The data also includes two horizontal layered interfaces, whose reflection coefficients are 0.01 and 0.012, respectively.

In order to test the diffractive wave separation effect of the present invention, this embodiment uses noise-free data for testing. The test results include (A), (B), (C), and (D) four sub-graphs, respectively. The data are: Figure (A) is the input data graph, Figure (B) is the separated diffraction wave data graph, Figure (C) is a graph of data other than the diffracted wave, and Figure (D) is the difference between the separated diffracted wave and the real diffracted wave.

As can be seen in Figure 3, the reflected wave signal is completely separated from the diffracted wave signal, and the separated diffracted wave signal is relatively complete and complete, and the difference from the real diffracted wave signal is very small, making it difficult to identify with the naked eye.

Example 2

This embodiment is Embodiment 2 of a method for separating seismic diffraction waves in a sparse domain. The difference between this embodiment and Embodiment 1 is that in this embodiment, synthetic seismic data is used as the test data, and the data trace spacing is 12.5m , there are 8 groups of diffracted waves with a certain distance adjacent to each other in the data, and the distance changes of diffraction are 12.5m, 25m, 37.5m, 50m, 62.5m, 75m, 87.5m and 100m respectively. The reflection coefficient at the vertex of the left diffracted wave is 0.0023, and the reflection coefficient at the vertex of the right diffracted wave is 0.0028. The data also includes two horizontal layered interfaces, whose reflection coefficients are 0.01 and 0.012, respectively.

In order to test the diffracted wave separation effect of the present invention, data with a signal-to-noise ratio of 12db is used for testing in this embodiment. The test results include (A), (B), (C), and (D) four sub-graphs, respectively. The data are: Figure (A) is the input data graph, Figure (B) is the separated diffraction wave data graph, Figure (C) is a graph of data other than the diffracted wave, and Figure (D) is the difference between the separated diffracted wave and the real diffracted wave.

As can be seen in Figure 4, since the amplitude value of the two wings of the diffracted wave gradually decays with the increase of the distance, the diffracted wave signal becomes difficult to identify relative to the vertex of the diffracted wave. If the separation is based on the slope, due to its The signal-to-noise ratio is low, and diffracted wave signals are often difficult to separate well. However, the method of the present invention still separates the diffracted wave signal completely and completely. From Fig. 4(D), it can also be seen that the difference between the diffracted wave and the real diffracted wave is separated, and there is only a slight difference between the two.

Example 3

This embodiment is Embodiment 3 of a method for separating seismic diffraction waves in a sparse domain. The difference between this embodiment and Embodiment 1 is that in this embodiment, synthetic seismic data is used as the test data, and the data trace spacing is 12.5m , there are 8 groups of diffracted waves with a certain distance adjacent to each other in the data, and the distance changes of diffraction are 12.5m, 25m, 37.5m, 50m, 62.5m, 75m, 87.5m and 100m respectively. The reflection coefficient at the vertex of the left diffracted wave is 0.0023, and the reflection coefficient at the vertex of the right diffracted wave is 0.0028. The data also includes two horizontal layered interfaces, whose reflection coefficients are 0.01 and 0.012, respectively.

In order to test the diffraction wave separation effect of the present invention, data with a signal-to-noise ratio of 6db is used for testing in this embodiment. The test results include (A), (B), (C), and (D) four sub-graphs, respectively. The data are: Figure (A) is the input data graph, Figure (B) is the separated diffraction wave data graph, Figure (C) is a graph of data other than the diffracted wave, and Figure (D) is the difference between the separated diffracted wave and the real diffracted wave.

As can be seen in Figure 5, due to the further increase of noise energy, the signals of the two wings of the diffracted wave are basically unrecognizable, and only the clearer diffracted wave signal can be seen near the vertex of the diffracted wave. The wave separation effect is still ideal. Although the difference between the real diffracted wave and the separated diffracted wave becomes stronger, its energy value is still relatively weak compared to the energy of the real diffracted wave.

Obviously, the above-mentioned embodiments of the present invention are only examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those of ordinary skill in the art, changes or modifications in other different forms can also be made on the basis of the above description. There is no need and cannot be exhaustive of all implementations here. Any modification, equivalent replacement and improvement made within the spirit and principle of the present invention shall be included within the protection scope of the claims of the present invention.

Claims (10)

1. A sparse domain seismic diffracted wave separation method is characterized by comprising the following steps:

the method comprises the following steps: constructing diffracted wave signals with different curvatures by using the theory that a time-distance curve of diffracted waves is a hyperbolic curve and amplitude energy and propagation distance are in inverse proportion to generate a diffracted wave signal atom library;

step two: performing iterative decomposition on searching signal atoms in the diffracted wave signal atom library obtained in the step one by using a matching pursuit algorithm to obtain sparse diffracted wave signal atoms; in the matching pursuit decomposition process, each iteration adopts a least square method to solve the coefficient of sparse signal atoms;

step three: and D, performing weighted superposition on the diffracted wave signal atoms obtained in the step two and the coefficients thereof to obtain separated diffracted wave data.

2. The sparse-domain seismic diffracted wave separation method of claim 1, wherein: in the first step, before generating the diffracted wave signal atom library, data needs to be input and rearranged, specifically: inputting a shot gather or a common-center-point gather, and rearranging the input shot gather or the common-center-point gather according to the offset distance to obtain a common-offset-distance gather d; simultaneous input of root mean square velocity field v _rms 。

3. The sparse domain seismic diffracted wave separation method of claim 2, wherein: in the first step, after data is input and rearranged, parameters in the diffracted wave separation process are set, including the curvature range [ sigma ] of diffracted wave signal atoms ₁ ～σ ₂ ]Curvature interval delta sigma, damping coefficient epsilon of least square solution, minimum acceptable correlation coefficient threshold alpha of diffracted wave signal atoms and actual data _acp Maximum acceptable residual percentage ζ _acp The maximum number of iterations N is matched and tracked.

4. A sparse domain seismic diffracted wave separation method as claimed in claim 3, wherein: generating a diffracted wave signal Atom library Atom (sigma, t) in the first step ₀ And x), wherein,

the time-distance curve calculation formula of the diffracted wave signal adopts a hyperbolic formula,

wherein t is ₀ For the two-way reflection time of the vertex of the diffracted wave, x is the distance from the horizontal to the vertex of the hyperbola, i is the curvature index, and i is 1, 2.

5. The sparse-domain seismic diffracted wave separation method of claim 4, wherein: generating a diffracted wave signal Atom library Atom (sigma, t) in the first step ₀ And x), wherein,

the amplitude of the diffracted wave is calculated by using the formula

Wherein A is ₀ V is the seismic amplitude and the root mean square velocity at the vertex of the hyperbola, x is the distance between the horizontal direction and the vertex of the hyperbola, and a and b are undetermined coefficients related to the amplitude.

6. The sparse-domain seismic diffracted wave separation method of claim 5, wherein: generating a diffracted wave signal Atom library Atom (sigma, t) in the first step ₀ X) further comprises the following operations:

performing least square fitting according to the searched diffracted wave signals, namely | | A (x) -A _real (x)||→min，A _real (x) Is the amplitude value of the actual diffracted wave data.

7. The sparse-domain seismic diffracted wave separation method of claim 1, wherein: the specific operations of searching for diffracted wave signal atoms in the second step are as follows:

(I) setting initial parameters:

initializing residual data r ₀ D, outputting diffracted wave seismic data d _s 0, the searched diffracted wave Atom library Atom _searched Phi, the iteration number k is 0;

circularly processing time sampling points;

circularly processing the number of seismic channels;

(II) matching and tracking circulation: k is less than or equal to N;

(III) calculating Atom library Atom (sigma, t) of diffracted wave signals ₀ X) and data residual r _k Correlation coefficient of

(iv) calculating an index value m with the largest correlation coefficient C (σ) as argmax (C (σ) _i ) 1,2,.., N, if C (σ) _m )>α _acp If not, ending the matching tracking cycle;

(V) adding the searched diffracted wave signal atoms into the searched Atom library, Atom _searched ＝Atom _searched ∪Atom(σ _m ,t ₀ X) while excluding from the atom pool of the next iteration,

representing a remove atom operation.

8. The sparse-domain seismic diffracted wave separation method of claim 7, wherein: and in the second step, the coefficient for solving the sparse signal atoms is a global optimization coefficient for solving diffracted wave atoms, and the adopted equation is an overdetermined equation.

9. The sparse-domain seismic diffracted wave separation method of claim 8, wherein: the second step of solving the coefficients of the sparse signal atoms is specifically operated as follows:

let L be Atom _searched The global optimization coefficient y of the searched diffracted wave atoms is solved by adopting a least square method _searched ＝(L ^T L+εI) ^-1 L ^T d, T denotes a matrix transpose, I denotes a unitAnd (3) a diagonal matrix, wherein epsilon represents a least square solution damping coefficient, and d represents a common offset gather.

10. The sparse-domain seismic diffracted wave separation method of claim 9, wherein: the third step comprises the following specific operations:

reconstructing diffracted wave seismic data d _s ＝Ly _searched ；

Updating residual r _k+1 ＝d-d _s Calculating the percentage coefficient of the residual

If ζ>ζ _acp If k is k +1, then step (III) is entered, otherwise, the matching tracking loop is ended;

the matching pursuit cycle ends;

finishing the seismic channel number cycle processing;

finishing the time sampling point cycle processing;

output diffracted wave seismic data d _s 。

Images