Inversion method and device based on omnibearing seismic data, electronic equipment and medium

Abstract

The invention discloses an inversion method and device based on omnibearing seismic data, electronic equipment and a medium. The method comprises the following steps: determining variables representing a longitudinal wave reflection coefficient, a representation isotropy gradient, a representation anisotropy gradient and a representation fracture development direction; establishing a matching target functional of the observation gather and the omnibearing forward modeling gather according to a plurality of variables; respectively establishing constraint models of the variable representing the longitudinal wave reflection coefficient and the variable representing the isotropic gradient; respectively establishing constraint models of the variable representing the anisotropy gradient and the variable representing the fracture development direction; and in the range of the constraint model, carrying out optimization solution on the matching target functional through a random algorithm, and realizing inversion of all-dimensional seismic data. According to the method, the longitudinal wave reflection coefficient, the isotropic gradient, the anisotropic gradient and the fracture development direction are inverted from the omnibearing angle gather through omnibearing waveform inversion, and important data are provided for application of omnibearing seismic data and prediction of fracture distribution.

Description

Technical Field

[0001] The present invention relates to the field of petroleum geophysical exploration, and more specifically, to an inversion method, device, electronic equipment and medium based on omnidirectional seismic data. Background Art

[0002] Fracture distribution and development information plays an important guiding role in the exploration and development of oil and gas. During the fracturing process, only when underground rocks continuously generate various forms of fracture networks can oil and gas wells achieve higher production capacity. In order to effectively identify fractures, geophysicists have proposed various fracture detection methods, such as stress field analysis and discontinuity detection based on post-stack seismic data, and AVO and anisotropy analysis based on pre-stack seismic data. In the application of post-stack data, fracture prediction based on stress field analysis is suitable for fractures caused by structures. Various fracture prediction methods based on discontinuity detection, such as coherence analysis, generalized Hilbert transform and image edge detection, can effectively delineate the boundary between irregular fracture-cavity development zones and seismic wave fields of homogeneous media, but are greatly affected by noise. With the application of wide-azimuth (or omnidirectional) seismic exploration, how to mine more abundant underground medium information from omnidirectional seismic data has become an important research hotspot. Since fractured reservoirs often have obvious HTI anisotropy, which is mainly manifested as changes in amplitude, velocity, reflection waveform and phase with the azimuth of the survey line, the change of P-wave information with azimuth can be used to predict fractures in the application of wide-azimuth pre-stack data. For example, azimuth travel time, azimuth attenuation, azimuth velocity (VVAZ, velocity changes with azimuth) and azimuth amplitude change (AVAZ, amplitude changes with azimuth) can be used to predict fractures. In terms of using P-wave azimuth anisotropy characteristics to predict fractures, Chinese scholars have carried out a lot of application research. For example, Cheng Bingjie et al. used the converted wave AVAZ fracture detection method based on wide-azimuth converted wave seismic data to detect fractures in deep fractured gas reservoirs in western Sichuan; Sun Wei studied the optimization of seismic attributes for P-wave azimuth anisotropic fracture prediction, and verified the reliability and accuracy of the prediction results using fracture information on imaging logging and core data. The current method is to extract attributes or perform seismic inversion on seismic data at different azimuth angles, and then perform anisotropic ellipse fitting on the attributes at different azimuth angles. The ratio of the long and short axes of the ellipse and the direction of the long and short axes are used to characterize the intensity and direction of fracture development, thereby predicting the distribution characteristics of reservoir fractures. The seismic attributes used for ellipse fitting can be AVO attributes or wave impedance, etc. The existing technical solution of first dividing the azimuth and then extracting the all-round seismic data is still carried out in a single azimuth and is affected by the energy balance of data in different azimuths.

[0003] At present, there is still a need to develop an inversion method based on all-round seismic data.

[0004] The information disclosed in the background technology section of the present invention is only intended to deepen the understanding of the general background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0005] The present invention proposes an inversion method, device, electronic equipment and medium based on omnidirectional seismic data, which can realize the inversion of longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction from omnidirectional angle gathers through omnidirectional waveform inversion, and provide important information for the application of omnidirectional seismic data and the prediction of fracture distribution.

[0006] In a first aspect, the present disclosure provides an inversion method based on omnidirectional seismic data, comprising:

[0007] Determine multiple variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing an anisotropic gradient, and a variable representing a crack development direction;

[0008] According to the plurality of variables, a matching target functional between the observed gathers and the omnidirectional forward modeling gathers is established;

[0009] Constraint models of variables representing longitudinal wave reflection coefficient and variables representing isotropic gradient are established respectively;

[0010] Constraint models for variables representing anisotropy gradient and variables representing crack development direction are established respectively;

[0011] Within the scope of the constraint model, the matching target functional is optimized and solved by a random algorithm to achieve inversion of all-round seismic data.

[0012] As a specific implementation of the embodiment of the present disclosure, the matching target functional is:

[0013] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0014] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

[0015] As a specific implementation of the embodiment of the present disclosure, the omnidirectional forward modeling gathers are:

[0016] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0017] Where w is the seismic wavelet and * represents the convolution operation.

[0018] As a specific implementation of the embodiment of the present disclosure, the constraint models of the variables characterizing the longitudinal wave reflection coefficient and the variables characterizing the isotropic gradient are respectively established, including:

[0019] Extracting prestack attributes for the observation gathers to obtain intercept attributes P and gradient attributes G;

[0020] The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

[0021] As a specific implementation of the embodiment of the present disclosure, the constraint model of the variable characterizing the longitudinal wave reflection coefficient is:

[0022] Amodel=[(1-α)Ps,(1+α)Ps]

[0023] The constraint model of the variable characterizing the isotropic gradient is:

[0024] Bmodel=[(1-β)Gs,(1+β)Gs]

[0025] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

[0026] As a specific implementation of the embodiment of the present disclosure, the constraint model of the variable characterizing the anisotropic gradient is:

[0027] Cmodel∈[C min ,C max ]

[0028] The constraint model of the variables that characterize the direction of crack development is:

[0029] Dmodel∈[D min ,D max ]

[0030] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, Dmin is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

[0031] As a specific implementation of the embodiment of the present disclosure, the random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0032] In a second aspect, the present disclosure also provides an inversion device based on omnidirectional seismic data, including:

[0033] A variable determination module determines multiple variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing anisotropic gradient, and a variable representing a crack development direction;

[0034] A target functional building module is used to build a matching target functional between the observed gathers and the omnidirectional forward modeling gathers according to the multiple variables;

[0035] The first constraint module establishes constraint models of variables representing the longitudinal wave reflection coefficient and variables representing the isotropic gradient respectively;

[0036] The second constraint module establishes constraint models for variables representing anisotropic gradients and variables representing crack development directions;

[0037] The calculation module optimizes and solves the matching target functional within the scope of the constraint model through a random algorithm to achieve inversion of all-round seismic data.

[0038] As a specific implementation of the embodiment of the present disclosure, the matching target functional is:

[0039] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0040] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

[0041] As a specific implementation of the embodiment of the present disclosure, the omnidirectional forward modeling gathers are:

[0042] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0043] Where w is the seismic wavelet and * represents the convolution operation.

[0044] As a specific implementation of the embodiment of the present disclosure, the constraint models of the variables characterizing the longitudinal wave reflection coefficient and the variables characterizing the isotropic gradient are respectively established, including:

[0045] Extracting prestack attributes for the observation gathers to obtain intercept attributes P and gradient attributes G;

[0046] The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

[0047] As a specific implementation of the embodiment of the present disclosure, the constraint model of the variable characterizing the longitudinal wave reflection coefficient is:

[0048] Amodel=[(1-α)Ps,(1+α)Ps]

[0049] The constraint model of the variable characterizing the isotropic gradient is:

[0050] Bmodel=[(1-β)Gs,(1+β)Gs]

[0051] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

[0052] As a specific implementation of the embodiment of the present disclosure, the constraint model of the variable characterizing the anisotropic gradient is:

[0053] Cmodel∈[C min ,C max ]

[0054] The constraint model of the variables characterizing the direction of crack development is:

[0055] Dmodel∈[D min ,D max ]

[0056] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, D min is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

[0057] As a specific implementation of the embodiment of the present disclosure, the random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0058] In a third aspect, an embodiment of the present disclosure further provides an electronic device, the electronic device comprising:

[0059] A memory storing executable instructions;

[0060] A processor runs the executable instructions in the memory to implement the inversion method based on omnidirectional seismic data.

[0061] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the inversion method based on omnidirectional seismic data is implemented.

[0062] Its beneficial effects are:

[0063] The present invention starts directly from omnidirectional seismic data, takes omnidirectional seismic data as its basis, comprehensively utilizes seismic data at various azimuths, and uses the P-wave azimuthal anisotropy analysis principle to perform waveform inversion. The P-wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction are inverted from the omnidirectional seismic data, providing important analytical technical means for fracture-type reservoir prediction. The inverted anisotropic gradient and fracture direction can directly characterize the fracture development intensity and direction.

[0064] The methods and apparatus of the present invention have other features and advantages that will be apparent from, or will be described in detail in, the accompanying drawings and subsequent detailed descriptions incorporated herein, which together serve to explain the specific principles of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The above and other objects, features and advantages of the present invention will become more apparent through a more detailed description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, wherein like reference numerals generally represent like components throughout the exemplary embodiments of the present invention.

[0066] Figure 1 A flow chart showing the steps of an inversion method based on omnidirectional seismic data according to an embodiment of the present invention.

[0067] Figure 2 A schematic diagram of an azimuth angle gather according to an embodiment of the present invention is shown.

[0068] Figure 3A schematic diagram showing the superposition display of the P-wave reflection coefficient and the isotropic gradient according to an embodiment of the present invention is shown.

[0069] Figure 4 A schematic diagram showing the combined display of the P-wave reflection coefficient and the anisotropy gradient according to an embodiment of the present invention is shown.

[0070] Figure 5a , Figure 5b , Figure 5c , Figure 5d Schematic diagrams of the P-wave reflection coefficient A, the isotropic gradient B, the anisotropic gradient C, and the layer-wise slices in the crack direction according to an embodiment of the present invention are respectively shown.

[0071] Figure 6 A block diagram of an inversion device based on omnidirectional seismic data according to an embodiment of the present invention is shown.

[0072] Description of reference numerals:

[0073] 201, variable determination module; 202, target functional establishment module; 203, first constraint module; 204, second constraint module; 205, calculation module. DETAILED DESCRIPTION

[0074] The preferred embodiments of the present invention will be described in more detail below. Although the preferred embodiments of the present invention are described below, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0075] To facilitate understanding of the solutions and effects of the embodiments of the present invention, six specific application examples are given below. Those skilled in the art should understand that the examples are only for facilitating understanding of the present invention, and any specific details thereof are not intended to limit the present invention in any way.

[0076] Example 1

[0077] Figure 1 A flow chart showing the steps of an inversion method based on omnidirectional seismic data according to an embodiment of the present invention.

[0078] like Figure 1As shown, the inversion method based on omnidirectional seismic data includes: step 101, determining multiple variables, including variables representing the longitudinal wave reflection coefficient, variables representing the isotropic gradient, variables representing the anisotropic gradient and variables representing the fracture development direction; step 102, establishing matching target functionals of the observation track set and the omnidirectional forward modeling track set according to the multiple variables; step 103, respectively establishing constraint models of the variables representing the longitudinal wave reflection coefficient and the variables representing the isotropic gradient; step 104, respectively establishing constraint models of the variables representing the anisotropic gradient and the variables representing the fracture development direction; step 105, within the scope of the constraint model, optimizing the matching target functional by a random algorithm to achieve the inversion of omnidirectional seismic data.

[0079] In one example, the matching target functional is:

[0080] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0081] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

[0082] In an example, the omni-directional forward gathers are:

[0083] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0084] Where w is the seismic wavelet and * represents the convolution operation.

[0085] In one example, establishing constraint models for variables representing longitudinal wave reflection coefficients and variables representing isotropic gradients respectively includes:

[0086] Extract prestack attributes for observation gathers to obtain intercept attribute P and gradient attribute G;

[0087] The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

[0088] In one example, the constraint model for the variable characterizing the longitudinal wave reflection coefficient is:

[0089] Amodel=[(1-α)Ps,(1+α)Ps]

[0090] The constraint model for the variables characterizing the isotropic gradient is:

[0091] Bmodel=[(1-β)Gs,(1+β)Gs]

[0092] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

[0093] In one example, the constraint model for the variable characterizing the anisotropic gradient is:

[0094] Cmodel∈[C min ,C max ]

[0095] The constraint model of the variables that characterize the direction of crack development is:

[0096] Dmodel∈[D min ,D max ]

[0097] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, D min is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

[0098] In one example, the random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0099] Specifically, a matching target functional of the observation gather and the omnidirectional forward modeling gather is established. The specific calculation formula of the matching target functional of the observation gather and the omnidirectional forward modeling gather is as follows:

[0100] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0101] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward gather. The observation gather GatherIn(φ,θ) is the input data. The omnidirectional forward gather GatherData(φ,θ) is obtained by forward modeling using seismic wavelets and omnidirectional seismic reflection coefficients. The calculation method of the omnidirectional forward gather is:

[0102] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0103] Among them, θ is the incident angle, φ is the azimuth, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction, w is the seismic wavelet, and * represents the convolution operation.

[0104] A constraint model of variables A and B is established. The constraint model is established as follows:

[0105] The pre-stack attribute extraction is performed on the observation gather GatherIn(φ,θ) to obtain the intercept attribute P and the gradient attribute G. The attributes P and G are smoothed to obtain Ps and Gs respectively. The constraint models of variables A and B are obtained according to the following formulas:

[0106] Amodel=[(1-α)Ps,(1+α)Ps]

[0107] Bmodel=[(1-β)Gs,(1+β)Gs]

[0108] Among them, Amodel and Bmodel represent the constraint models of variables A and B, α and β are constraint coefficients, and their value ranges are 0.1≤α≤0.4, 0.1≤β≤0.4; the smoothing process is median filtering.

[0109] A constraint model of variables C and D is established. The constraint model of variables C and D is expressed as follows:

[0110] Cmodel∈[C min ,C max ]

[0111] Dmodel∈[D min ,D max ]

[0112] Among them, C min ,C max ,D min ,Dmax They represent the minimum and maximum values ​​of variables C and D, respectively, obtained from imaging logging data and seismic data.

[0113] The matching target functional of the above steps is optimized and solved by using a random algorithm within the scope of the constraint model. The random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0114] Example 2

[0115] The present invention also provides an inversion device based on omnidirectional seismic data, comprising:

[0116] A variable determination module determines multiple variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing anisotropic gradient, and a variable representing a crack development direction;

[0117] The target functional building module builds the matching target functional between the observed gathers and the omnidirectional forward modeling gathers based on multiple variables;

[0118] The first constraint module establishes constraint models of variables representing the longitudinal wave reflection coefficient and variables representing the isotropic gradient respectively;

[0119] The second constraint module establishes constraint models for variables representing anisotropic gradients and variables representing crack development directions;

[0120] The calculation module, within the scope of the constraint model, uses a random algorithm to optimize the matching target functional and realize the inversion of all-round seismic data.

[0121] In one example, the matching target functional is:

[0122] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0123] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

[0124] In an example, the omni-directional forward gathers are:

[0125] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0126] Where w is the seismic wavelet and * represents the convolution operation.

[0127] In one example, establishing constraint models for variables representing longitudinal wave reflection coefficients and variables representing isotropic gradients respectively includes:

[0128] Extract prestack attributes for observation gathers to obtain intercept attribute P and gradient attribute G;

[0129] The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

[0130] In one example, the constraint model for the variable characterizing the longitudinal wave reflection coefficient is:

[0131] Amodel=[(1-α)Ps,(1+α)Ps]

[0132] The constraint model for the variables characterizing the isotropic gradient is:

[0133] Bmodel=[(1-β)Gs,(1+β)Gs]

[0134] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

[0135] In one example, the constraint model for the variable characterizing the anisotropic gradient is:

[0136] Cmodel∈[C min ,C max ]

[0137] The constraint model of the variables that characterize the direction of crack development is:

[0138] Dmodel∈[D min ,D max ]

[0139] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, D min is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

[0140] In one example, the random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0141] Specifically, a matching target functional of the observation gather and the omnidirectional forward modeling gather is established. The specific calculation formula of the matching target functional of the observation gather and the omnidirectional forward modeling gather is as follows:

[0142] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0143] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward gather. The observation gather GatherIn(φ,θ) is the input data. The omnidirectional forward gather GatherData(φ,θ) is obtained by forward modeling using seismic wavelets and omnidirectional seismic reflection coefficients. The calculation method of the omnidirectional forward gather is:

[0144] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0145] Among them, θ is the incident angle, φ is the azimuth, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction, w is the seismic wavelet, and * represents the convolution operation.

[0146] A constraint model of variables A and B is established. The constraint model is established as follows:

[0147] The pre-stack attribute extraction is performed on the observation gather GatherIn(φ,θ) to obtain the intercept attribute P and the gradient attribute G. The attributes P and G are smoothed to obtain Ps and Gs respectively. The constraint models of variables A and B are obtained according to the following formulas:

[0148] Amodel=[(1-α)Ps,(1+α)Ps]

[0149] Bmodel=[(1-β)Gs,(1+β)Gs]

[0150] Among them, Amodel and Bmodel represent the constraint models of variables A and B, α and β are constraint coefficients, and their value ranges are 0.1≤α≤0.4, 0.1≤β≤0.4; the smoothing process is median filtering.

[0151] A constraint model of variables C and D is established. The constraint model of variables C and D is expressed as follows:

[0152] Cmodel∈[C min ,C max ]

[0153] Dmodel∈[D min ,D max ]

[0154] Among them, C min ,C max ,D min ,D max They represent the minimum and maximum values ​​of variables C and D, respectively, obtained from imaging logging data and seismic data.

[0155] The matching target functional of the above steps is optimized and solved by using a random algorithm within the scope of the constraint model. The random algorithm includes a genetic algorithm, a simulated annealing algorithm, and an ant colony algorithm.

[0156] Example 3

[0157] Based on omnidirectional seismic data, comprehensive use of seismic data at various azimuths, and the principle of P-wave azimuthal anisotropy analysis are used to perform waveform inversion of omnidirectional seismic data. The P-wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction are inverted from the omnidirectional seismic data, providing important analytical technical means for fracture-type reservoir prediction.

[0158] First, the seismic wavelet and omnidirectional seismic reflection coefficient are used for forward modeling to obtain omnidirectional forward gathers, and then the matching target functional of the omnidirectional forward gathers and the observation gathers is established. Then, the waveform inversion of omnidirectional seismic data is performed using the optimization algorithm. Through omnidirectional waveform inversion, the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction are inverted from omnidirectional angle gathers. The anisotropic gradient and fracture direction obtained by inversion can directly characterize the fracture development intensity and direction.

[0159] Taking actual seismic data of a certain area as an example, the method is used to invert the P-wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction from the omnidirectional seismic data, thereby illustrating the effect of the present invention.

[0160] Figure 2 A schematic diagram of an azimuth angle gather according to an embodiment of the present invention is shown, given as input data. Figure 2 The middle is an azimuth angle gather, which is composed of several angle gathers, and each angle gather corresponds to a different azimuth. Figure 2 The azimuth gather consists of five azimuths (0, 36, 72, 108, and 144 degrees), and each azimuth corresponds to a set of angle gathers (incident angle gathers, with the incident angle in the figure ranging from 0 to 45 degrees).

[0161] by Figure 2 The azimuth angle gathers are used as input data to perform waveform inversion of all-directional seismic data, and the variables A, B, C, and D representing the P-wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction are obtained;

[0162] Figure 3 A schematic diagram showing the superposition display of the P-wave reflection coefficient and the isotropic gradient according to an embodiment of the present invention is shown.

[0163] Figure 4 A schematic diagram showing the combined display of the P-wave reflection coefficient and the anisotropy gradient according to an embodiment of the present invention is shown.

[0164] Figure 5a , Figure 5b , Figure 5c , Figure 5d Schematic diagrams of the P-wave reflection coefficient A, the isotropic gradient B, the anisotropic gradient C, and the layer-wise slices in the crack direction according to an embodiment of the present invention are respectively shown.

[0165] The forward modeling is performed using seismic wavelets and omnidirectional seismic reflection coefficients to obtain omnidirectional forward modeling gathers, and then the matching target functional between the omnidirectional forward modeling gathers and the observation gathers is established. The waveform inversion of the omnidirectional seismic data is then performed using an optimization algorithm.

[0166] This embodiment is an implementation example of the present invention for inverting the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient and fracture development direction from omnidirectional seismic data. The steps of a seismic inversion method and system based on omnidirectional seismic data are as follows:

[0167] 1. Given omnidirectional seismic data, in the present invention, given in the form of azimuth angle gathers;

[0168] 2. Perform routine processing on the input data to obtain the constraint model of temporary variables (A, B, C, D), and then establish the matching target functional between the observed gathers and the omnidirectional forward modeling gathers;

[0169] 3. Use the genetic algorithm in the random algorithm to solve the above step 2 until the termination condition is met. The termination condition is that the objective function error reaches a smaller value (such as 0.1).

[0170] Example 4

[0171] Figure 6 A block diagram of an inversion device based on omnidirectional seismic data according to an embodiment of the present invention is shown.

[0172] like Figure 6 As shown, the inversion device based on omnidirectional seismic data includes:

[0173] The variable determination module 201 determines a plurality of variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing anisotropic gradient, and a variable representing a crack development direction;

[0174] A target functional building module 202 builds a matching target functional between the observed gathers and the omnidirectional forward modeling gathers according to a plurality of variables;

[0175] The first constraint module 203 establishes constraint models for variables representing longitudinal wave reflection coefficient and variables representing isotropic gradient respectively;

[0176] The second constraint module 204 establishes constraint models for variables representing anisotropic gradients and variables representing fracture development directions respectively;

[0177] The calculation module 205 optimizes the matching target functional within the scope of the constraint model through a random algorithm to achieve inversion of all-round seismic data.

[0178] As an alternative, the matching target functional is:

[0179] Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2

[0180] Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

[0181] As an optional solution, the full-scale forward modeling gathers are:

[0182] GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w

[0183] Where w is the seismic wavelet and * represents the convolution operation.

[0184] As an optional solution, the constraint models of the variables representing the longitudinal wave reflection coefficient and the variables representing the isotropic gradient are established respectively, including:

[0185] Extract prestack attributes for observation gathers to obtain intercept attribute P and gradient attribute G;

[0186] The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

[0187] As an alternative, the constraint model of the variable characterizing the longitudinal wave reflection coefficient is:

[0188] Amodel=[(1-α)Ps,(1+α)Ps]

[0189] The constraint model for the variables characterizing the isotropic gradient is:

[0190] Bmodel=[(1-β)Gs,(1+β)Gs]

[0191] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

[0192] As an alternative, the constraint model for the variable characterizing the anisotropic gradient is:

[0193] Cmodel∈[C min ,C max ]

[0194] The constraint model of the variables that characterize the direction of crack development is:

[0195] Dmodel∈[D min ,D max ]

[0196] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, D min is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

[0197] As an alternative, random algorithms include genetic algorithms, simulated annealing algorithms, and ant colony algorithms.

[0198] Example 5

[0199] This embodiment provides an electronic device, which includes: a memory storing executable instructions; and a processor, which runs the executable instructions in the memory to implement the above-mentioned inversion method based on omnidirectional seismic data.

[0200] An electronic device according to an embodiment of the present disclosure includes a memory and a processor.

[0201] The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0202] The processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In an embodiment of the present disclosure, the processor is used to run the computer-readable instructions stored in the memory.

[0203] Those skilled in the art should understand that, in order to solve the technical problem of how to obtain good user experience effects, this embodiment may also include well-known structures such as communication buses, interfaces, etc., and these well-known structures should also be included in the protection scope of the present disclosure.

[0204] For a detailed description of this embodiment, reference may be made to the corresponding descriptions in the foregoing embodiments, and details will not be repeated here.

[0205] Example 6

[0206] This embodiment provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the inversion method based on omnidirectional seismic data is implemented.

[0207] A computer-readable storage medium according to an embodiment of the present disclosure stores non-transitory computer-readable instructions thereon. When the non-transitory computer-readable instructions are run by a processor, all or part of the steps of the methods of the foregoing embodiments of the present disclosure are executed.

[0208] The above-mentioned computer-readable storage media include but are not limited to: optical storage media (such as CD-ROM and DVD), magneto-optical storage media (such as MO), magnetic storage media (such as magnetic tape or mobile hard disk), media with built-in rewritable non-volatile memory (such as memory card), and media with built-in ROM (such as ROM cartridge).

[0209] Those skilled in the art should understand that the purpose of the above description of the embodiments of the present invention is only to exemplarily illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any given examples.

[0210] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. An inversion method based on omnidirectional seismic data, It is characterized in that include: Determine multiple variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing an anisotropic gradient, and a variable representing a crack development direction; According to the plurality of variables, a matching target functional between the observed gathers and the omnidirectional forward modeling gathers is established; Constraint models of variables representing longitudinal wave reflection coefficient and variables representing isotropic gradient are established respectively; Constraint models for variables representing anisotropy gradient and variables representing crack development direction are established respectively; Within the scope of the constraint model, the matching target functional is optimized and solved by a random algorithm to achieve inversion of all-round seismic data.

2. The inversion method based on omnidirectional seismic data according to claim 1, in, The matching target functional is: Fobj(A,B,C,D)=||GatherIn(φ,θ)-GatherData(φ,θ)|| 2 Among them, A, B, C, and D are variables that characterize the longitudinal wave reflection coefficient, isotropic gradient, anisotropic gradient, and fracture development direction. θ is the incident angle, φ is the azimuth, GatherIn(φ,θ) is the observation gather, and GatherData(φ,θ) is the omnidirectional forward modeling gather.

3. The inversion method based on omnidirectional seismic data according to claim 2, in, The omnidirectional forward modeling gathers are: GatherData(φ,θ)={A+[B+Ccos(2φ)+Dsin(2φ)]sin 2 θ}*w Where w is the seismic wavelet and * represents the convolution operation.

4. The inversion method based on omnidirectional seismic data according to claim 1, in, The constraint models for establishing variables representing the longitudinal wave reflection coefficient and the variables representing the isotropic gradient respectively include: Extracting prestack attributes for the observation gathers to obtain intercept attributes P and gradient attributes G; The attributes P and G are smoothed to obtain Ps and Gs, respectively, and then constraint models of variables representing the longitudinal wave reflection coefficient and the isotropic gradient are established respectively.

5. The inversion method based on omnidirectional seismic data according to claim 4, in, The constraint model of the variable characterizing the longitudinal wave reflection coefficient is: Amodel=[(1-α)Ps,(1+α)Ps] The constraint model of the variable characterizing the isotropic gradient is: Bmodel=[(1-β)Gs,(1+β)Gs] Among them, Amodel is the constraint model of the variables characterizing the longitudinal wave reflection coefficient, Bmodel is the constraint model of the variables characterizing the isotropic gradient, and α and β are constraint coefficients.

6. The inversion method based on omnidirectional seismic data according to claim 1, in, The constraint model for the variables characterizing the anisotropic gradient is: Cmodel∈[C min ,C max ] The constraint model of the variables that characterize the direction of crack development is: Dmodel∈[D min ,D max ] Among them, Cmodel is the constraint model of the variable characterizing the anisotropic gradient, Dmodel is the constraint model of the variable characterizing the fracture development direction, and C min is the minimum value of the variable characterizing the anisotropic gradient, C max is the minimum value of the variable characterizing the anisotropic gradient, D min is the minimum value of the variable that characterizes the direction of crack development, D max is the minimum value of the variable that characterizes the direction of crack development.

7. The inversion method based on omnidirectional seismic data according to claim 1, in, The random algorithms include genetic algorithms, simulated annealing algorithms, and ant colony algorithms.

8. An inversion device based on omnidirectional seismic data, It is characterized in that include: A variable determination module determines multiple variables, including a variable representing a longitudinal wave reflection coefficient, a variable representing an isotropic gradient, a variable representing anisotropic gradient, and a variable representing a crack development direction; A target functional building module is used to build a matching target functional between the observed gathers and the omnidirectional forward modeling gathers according to the multiple variables; The first constraint module establishes constraint models of variables representing the longitudinal wave reflection coefficient and variables representing the isotropic gradient respectively; The second constraint module establishes constraint models for variables representing anisotropic gradients and variables representing crack development directions; The calculation module optimizes and solves the matching target functional within the scope of the constraint model through a random algorithm to achieve inversion of all-round seismic data.

9. An electronic device, It is characterized in that The electronic device comprises: A memory storing executable instructions; A processor, wherein the processor runs the executable instructions in the memory to implement the inversion method based on omnidirectional seismic data as described in any one of claims 1-7.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the inversion method based on omnidirectional seismic data described in any one of claims 1 to 7 is implemented.

Images