Reservoir three-dimensional space delineation method based on dual transfer function blocking body imaging

Abstract

The application provides a reservoir three-dimensional space delineation method based on a dual transfer function light-blocking body imaging, comprising the following steps: step 1, calculating seismic body attributes and constructing a kernel matrix K; step 2, calculating eigenvalues λ of the kernel matrix i and eigenvectors v i ; step 3, converting from seismic original data to body fusion attributes with reservoir characteristics, and converting each attribute value into a color; step 4, constructing a light-blocking degree calculation model of a dual transfer function; step 5, calculating a light-blocking body, and performing body space resampling on the light-blocking body value in each light direction; step 6, performing cumulative calculation on the light-blocking body voxel value of each point in the light propagation direction, and completing reservoir three-dimensional space description. The reservoir three-dimensional space delineation method based on the dual transfer function light-blocking body imaging accurately describes the gradual change characteristics of the numerical boundary of the reservoir, accurately displays the geological characteristics in the reservoir three-dimensional space, and improves the reservoir three-dimensional body delineation precision.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oilfield development, and particularly relates to a reservoir three-dimensional space delineation method based on a dual transfer function light-blocking body imaging. BACKGROUND

[0002] At present, most of the old areas in the eastern part of China are in the "mature" exploration area stage with high exploration and high proven degree. The exploration and development targets gradually shift from shallow layers to deep layers, from structural oil and gas reservoirs to complex and concealed oil and gas reservoirs, and the main positions of exploration and development also shift from conventional reservoirs to complex reservoir targets. Therefore, accurate reservoir three-dimensional space delineation technology is needed to describe the underground distribution characteristics of complex reservoirs and improve the success rate of exploration and development.

[0003] As known, due to the poor spatial continuity, large lateral variation of lithology, strong heterogeneity, vertical superposition and other factors of continental reservoirs, the reservoir three-dimensional space delineation precision is low, and the multiple solutions are strong. However, the reservoir three-dimensional space delineation determines the oil reservoir exploitation design and well deployment, and therefore, the accurate description of the reservoir space has always been an extremely important problem in the field of exploration and development. At present, the method of fixed seismic attribute maximum and minimum threshold value is usually used for volume carving. This hard threshold method cannot accurately identify the reservoirs with different seismic attribute value range (the seismic attribute value range of reservoirs at different depths is different), and the method of selecting maximum and minimum value cannot show the gradual change characteristics of the numerical boundary of the reservoir, which does not conform to the geological law of the spatial variation of the reservoir.

[0004] In the Chinese patent application with the application number CN201910794293.8, a thin sand body characterization method and device based on a seismic data volume are disclosed, which relates to the technical field of seismic sedimentology. The method comprises the following steps: determining a layer to be analyzed according to a seismic data volume; converting the seismic data volume into an attribute data volume according to the determined layer to be analyzed; adjusting the time window value and / or light-blocking degree curve in the attribute data volume to obtain the distribution characteristics of at least one thin sand body. The application can predict thin sand bodies by using relatively low-resolution seismic data volumes.

[0005] In the Chinese patent application with the application number CN201910644703.0, a continental thin reservoir interpretation method and system based on local peak frequency are disclosed, which belongs to the technical field of seismic exploration and solves the problem of difficult fine delineation of thin layers. The method comprises the following steps: performing Q compensation on seismic data to obtain Q-compensated seismic data; performing generalized S transform on the Q-compensated seismic data to obtain a time-frequency data volume; obtaining all corresponding time points of local time-frequency peaks of the time-frequency data volume to determine the thin layer position from the all corresponding time points of local time-frequency peaks. The method improves the interpretation accuracy of thin layers.

[0006] In the Chinese patent application No. CN201510779780.9, a method for identifying thin reservoir subtle lithologic oil and gas reservoirs is disclosed, which belongs to the technical field of oil exploration and development. The method comprises the following steps: calculating the quality factor of a formation, performing secondary high-frequency compensation processing on post-stack seismic data by using inverse Q filtering according to the quality factor to obtain ideal seismic data, researching the seismic response characteristics of the target layer by using the obtained ideal seismic data, determining the reservoir waveform characteristics, and making a seismic data skeleton profile according to the reservoir top and bottom waveform reflection characteristics. The planar distribution of the sand body is finely described on the basis of the seismic data skeleton profile, and the sand body boundary is determined by using seismic attribute analysis, so as to determine the thin reservoir subtle oil reservoir. The method is verified by specific examples, and can reliably identify the subtle lithologic oil and gas reservoirs and improve the identification rate of the thin reservoir subtle lithologic oil and gas reservoirs.

[0007] The above prior art is quite different from the present application, and cannot solve the technical problems we want to solve. Therefore, we have invented a new reservoir three-dimensional space description method based on double transfer function light blocking body imaging. SUMMARY

[0008] The purpose of the present application is to provide a reservoir three-dimensional space description method based on double transfer function light blocking body imaging, which can accurately describe the gradual change characteristics of the numerical boundary of the reservoir, accurately display the geological characteristics in the three-dimensional space of the reservoir, and improve the three-dimensional body description precision of the reservoir.

[0009] The purpose of the present application can be achieved by the following technical measures: the reservoir three-dimensional space description method based on double transfer function light blocking body imaging comprises the following steps:

[0010] Step 1, calculating seismic body attributes and constructing a kernel matrix K;

[0011] Step 2, calculating the eigenvalue λ of the kernel matrix i and the eigenvector v i of the kernel matrix;

[0012] Step 3, converting the seismic original data to the body fusion attribute with reservoir characteristics, and converting each attribute value to a color;

[0013] Step 4, constructing a light blocking degree calculation model of double transfer function;

[0014] Step 5, calculating the light blocking degree body, and performing body space resampling on the light blocking body value in each light direction;

[0015] Step 6, accumulating and calculating the light blocking voxel value of each point in the light propagation direction to complete the description of the three-dimensional space of the reservoir.

[0016] The purpose of the present application can also be achieved by the following technical measures:

[0017] In step 1, seismic volume attributes are calculated from the original seismic data volume u(x, y, z), resulting in a plurality of volume attributes, including:

[0018] S1(x, y, z) - a mean square amplitude volume attribute

[0019] S2(x, y, z) - a wave shape variation area volume attribute

[0020] S3(x, y, z) - an energy half-time volume attribute

[0021] S4(x, y, z) - a composite envelope difference volume attribute.

[0022] In step 1, the above seismic volume attributes to be fused are normalized, and a polynomial function is selected as a kernel function to construct a kernel matrix K, and the mathematical model of the kernel function is as follows:

[0023] k(S i , S j ) = (aS i T S j +c) n

[0024] In the formula: k is a polynomial kernel function; a and c are polynomial coefficients, constants; s i , S j are volume attributes to be fused; T is a transpose calculation; n is a power number.

[0025] In step 2, the eigenvalues λ i and eigenvectors v i of the kernel matrix are calculated, and the λ i and v i values are arranged from large to small, and the first q eigenvalues λ i with larger contribution rates and the corresponding eigenvectors v i are taken.

[0026] In step 2, the formula for arranging λ i and v i values from large to small is as follows:

[0027] ∧ q = diag(λ1, λ2, λ3......λ q , )

[0028] V q = diag(v1, v2, v3......v q , )

[0029] In the formula: ∧ qis a diagonal matrix composed of eigenvalues;

[0030] V q is a set composed of eigenvectors corresponding to eigenvalues.

[0031] In step 3, after removing redundant data, the seismic attribute volume fusion calculation is performed, and the above to-be-fused volume attributes are fused into one attribute volume with reservoir characteristics, completing the conversion from the seismic original data to the volume fusion attribute with reservoir characteristics.

[0032] In step 3, the seismic attribute volume fusion calculation formula is:

[0033]

[0034] In the formula, L is an arbitrary orthogonal matrix;

[0035] Q is an empirical load matrix;

[0036] Z is the fused seismic volume attribute;

[0037] T is the matrix transpose; K is the kernel matrix; and J represents the Jacobian matrix.

[0038] In step 3, color matching normalization processing is performed on each point in the three-dimensional space of the fused volume attribute Z, and each attribute value is converted into a color:

[0039]

[0040] In the formula, c(x, y, z) is the normalized corresponding color value (between 0 and 255); Z(x, y, z) i is the attribute value of an arbitrary point in space; Z(x, y, z) min is the minimum attribute value in the attribute volume; z(x, y, z) max is the maximum attribute value in the attribute volume.

[0041] In step 4, first, a parameter variable template t is established according to the maximum value and the minimum value of the volume attribute, and then an opacity model is established according to t and the transfer function. The establishment methods are a cubic spline transfer function opacity model and a Gaussian transfer function opacity model. The Gaussian transfer function opacity model controls the imaging of large attribute value voxels, and the cubic spline transfer function opacity model is used to control the imaging of small attribute value voxels.

[0042] In step 4, the cubic spline transfer function opacity model is:

[0043] Q i (t)=Q i (t 3 -3t 2 +1)+Q i+1 (-2t3 +3t z )+R i (t 3 -2t 2 +t)+R i+1 (t 3 -t 2 )

[0044] The light extinction model of the Gaussian transfer function is:

[0045]

[0046] In the formula, Q represents the position vector at the control point, i.e., the control point coordinates;

[0047] R represents the tangent vector at the control point, i.e., the tangent line or the first derivative at the control point;

[0048] t represents a parameter variable constructed by the maximum and minimum fusion attribute values to build the transfer function;

[0049] σ represents the variance value of the parameter variable t. i The value of the Gaussian function parameter variable t at any point, a i represents the weighting coefficient varying with t. i

[0050] In step 5, according to the light extinction calculation model in step 4 and the data C after the color matching conversion in step 3, the light extinction of each point in the space is calculated, the small attribute value is selected to calculate the cubic spline light extinction model, and the large attribute value is selected to calculate the Gaussian light extinction model.

[0051] In step 5, the formula for calculating the light extinction voxel f(x, y, z) is:

[0052]

[0053]

[0054] In the formula, M is the light extinction model selector, and P is the conversion function.

[0055] Z(x, y, z) midle is the average value.

[0056] In step 5, when performing light tracing and data body space resampling calculation, the light source point position is fixed, and light tracing and projection are performed from the light source point to the light extinction body f. Since the angle between each light and the light extinction body is different, it is necessary to perform body space resampling on the light extinction value in the direction of each light, i.e., projection along the path of each light.

[0057] In step 5, the resampling calculation formula is as follows: ​

[0058]

[0059] In the formula, P is a point in the body space, i.e. a point of light incidence;

[0060] f(P) is a value after resampling of the point P;

[0061] f(P1) is the light blocking value at P1;

[0062] f(P2) is the light blocking value at P2;

[0063] y P , y P1 , y P2 are the spatial y coordinate values of the three points, respectively.

[0064] In step 6, the top and bottom horizon surfaces of the reservoir space are selected to generate a mask data body, and only the effective mask voxels are subjected to light projection and accumulation in the light tracing projection, and finally the light blocking voxel value of each point in the light propagation direction is accumulated and calculated on the basis of the above, so that the three-dimensional space description of the reservoir is completed.

[0065] In step 6, the formula for accumulating and calculating the light blocking voxel value of each point is as follows:

[0066]

[0067] In the formula, β i is a light ray weighting coefficient related to the distance;

[0068] f(i) is the light blocking value of each point calculated in step 5;

[0069] D out is the accumulated light blocking voxel value in the light direction.

[0070] The three-dimensional space description method of the reservoir based on the dual transfer function light blocking body imaging in the application relates to a three-dimensional geological body or reservoir space description method based on the dual transfer function light blocking body imaging in the field of three-dimensional geological body or reservoir space description, and is suitable for three-dimensional space description of sandstone and sandy gravel reservoirs in terrestrial reservoirs.

[0071] The application is based on three-dimensional light-blocking body imaging technology of double transfer function, combines the polynomial kernel function body fusion method with the light-blocking voxel imaging technology based on double transfer function, uses the seismic attribute body fusion method to mine the reservoir characteristics, and uses the light-blocking voxel imaging method to further 'brighten' the reservoir, so that the spatial gradual change characteristics of the reservoir can be more accurately described, the three-dimensional carving effect of strengthening the reservoir and weakening the non-reservoir is achieved, more intuitive and reliable seismic data for reservoir modeling is provided, and the means for reservoir body spatial description and evaluation is more comprehensive.

[0072] The application first extracts multiple seismic body attributes such as amplitude, spectrum and sequence from the original seismic data body, selects a polynomial kernel function for nonlinear relationship mapping between the seismic body attributes on the basis, mines the common characteristics between the multiple body attributes, constructs a body attribute fusion relationship based on the polynomial kernel function, and fuses the multiple seismic attribute bodies into a body attribute with reservoir characteristics. Then, the light-blocking voxel imaging method based on double transfer function is used to depict the numerical boundary gradual change characteristics of the reservoir, different transfer functions are selected according to different value ranges, the problem that the numerical boundary accuracy of the reservoir is low in the prior art is overcome, the geological characteristics in the three-dimensional space of the reservoir are accurately displayed, and the three-dimensional body carving accuracy of the reservoir is improved. BRIEF DESCRIPTION OF DRAWINGS

[0073] Figure 1 Fig. 1 is a schematic diagram of an original three-dimensional seismic data body;

[0074] Figure 2 Fig. 2 is a three-dimensional seismic attribute data body after the multi-attribute body fusion in an embodiment of the application;

[0075] Figure 3 Fig. 3 is a schematic diagram of body space resampling in an embodiment of the application;

[0076] Figure 4 Fig. 4 is a reservoir three-dimensional space depiction diagram based on the light-blocking body imaging of double transfer function in the application;

[0077] Figure 5 Fig. 5 is a flowchart of the reservoir three-dimensional space depiction method based on the light-blocking body imaging of double transfer function in the application. DETAILED DESCRIPTION

[0078] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application belongs.

[0079] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, and / or combinations thereof.

[0080] The present invention provides a method for three-dimensional spatial characterization of reservoirs based on dual transfer function light-blocking imaging, comprising:

[0081] Step 1: Calculate the seismic body properties and construct the kernel matrix K;

[0082] Step 2, calculate the eigenvalues ​​λ of the kernel matrix. i With feature vector v i ;

[0083] Step 3: Perform a volume fusion attribute conversion from the raw seismic data to reservoir characteristics, and convert each attribute value into a color.

[0084] Step 4: Construct a light blocking calculation model with dual transfer functions;

[0085] Step 5: Calculate the light-blocking volume and resample the light-blocking volume value in each light direction in the volume space.

[0086] Step 6: Accumulate and calculate the voxel value of light blocking at each point along the direction of light propagation to complete the three-dimensional spatial description of the reservoir.

[0087] The following are several specific embodiments of the application of the present invention.

[0088] Example 1

[0089] In a specific embodiment 1 of the present invention, such as Figure 5 As shown, Figure 5 This is a flowchart of the reservoir three-dimensional spatial characterization method based on dual transfer function light-blocking imaging according to the present invention. The reservoir three-dimensional spatial characterization method based on dual transfer function light-blocking imaging includes:

[0090] Step 101: Calculate the seismic volume properties based on the original seismic data volume u(x, y, z).

[0091] Step 102: Based on the above seismic body attributes to be fused, perform normalization calculations on them, select a polynomial function as the kernel function, and construct the kernel matrix K.

[0092] Step 103: Calculate the eigenvalues ​​λ of the kernel matrix again. i With feature vector v i , and according to λ i and v iThe values are arranged from large to small (the following formula), and the first q eigenvalues λ with larger eigenvalue contribution rate are taken i and the corresponding eigenvectors v i .

[0093] Step 104: After removing redundant data, the seismic attribute volume fusion calculation is carried out by the following formula, the above to-be-fused volume attributes are fused into a volume attribute with reservoir characteristics, and the conversion from the original seismic data to the volume fusion attribute with reservoir characteristics is completed.

[0094] Step 105: Color matching normalization processing is performed on each point in the fused volume attribute Z three-dimensional space, and each attribute value is converted into a color.

[0095] Step 106: A lightness calculation model of the double transfer function is constructed, and the first thing to do this step is to establish a parameter variable template t according to the maximum value and the minimum value of the volume attribute, and then to establish a lightness model according to t and the transfer function.

[0096] Step 107: Calculate the lightness volume, according to the lightness calculation model of step 6 and the color conversion data C of step 5, calculate the lightness of each point in the space, select the cubic spline lightness model for calculation for small attribute values, and select the Gaussian lightness model for calculation for large attribute values, and obtain the lightness voxel f(x, y, z).

[0097] Step 108: Light ray tracing and data volume space resampling calculation, fix the light source point position, and perform light ray tracing and projection from the light source point to the lightness volume f. Since the angle of each light ray with the lightness volume is different, the lightness volume value in the direction of each light ray needs to be resampled in the volume space.

[0098] Step 109: Select the top and bottom horizon surfaces of the reservoir space to generate a mask data volume, and only the effective mask voxels are subjected to light projection and accumulation during light tracing and projection. Finally, the lightness voxel value of each point in the light propagation direction is accumulated and calculated, and the three-dimensional space description of the reservoir is completed.

[0099] Example 2

[0100] In a specific embodiment 2 of the application, the reservoir three-dimensional space description method based on the double transfer function lightness volume imaging includes the following steps:

[0101] Step 1: According to Figure 1 The calculation of the seismic volume attribute (the calculation method of the volume attribute is various, which will not be described in detail here) of the original seismic data volume u(x, y, z) can obtain a plurality of volume attributes, and the following several seismic volume attributes are selected as examples for description.

[0102] S1(x, y, z) is a mean square amplitude volume attribute

[0103] S2(x, y, z) - wave shape variable volume attribute

[0104] S3(x, y, z) - energy half-time volume attribute

[0105] S4(x, y, z) - composite envelope difference volume attribute.

[0106] Step 2: According to the above to be fused seismic volume attributes, a normalization calculation is performed, and a polynomial function is selected as a kernel function to construct a kernel matrix K, and the mathematical model of the kernel function is as follows:

[0107] k(s i , S j ) = (as i T S j +c) n

[0108] In the formula: k is a polynomial kernel function; a and c are polynomial coefficients (constants); S i , S j are volume attributes to be fused; T is a transpose calculation; and n is a power number.

[0109] Step 3: The eigenvalues λ i and eigenvectors v i of the kernel matrix are calculated again, and the λ i and v i values are arranged from large to small (the following formula), and the first q eigenvalues λ i with larger contribution rates and the corresponding eigenvectors v i are taken.

[0110] A q = diag(λ1, λ2, λ3... λ q , )

[0111] V q = diag(v1, v2, v3... v q , )

[0112] In the formula: A q is a diagonal matrix composed of eigenvalues;

[0113] V q is a set composed of eigenvectors corresponding to the eigenvalues.

[0114] Step 4: After removing redundant data, the seismic attribute volume fusion calculation is performed through the following formula, the above to be fused volume attributes are fused into an attribute volume with reservoir characteristics, the conversion from the original seismic data to the volume fusion attribute with reservoir characteristics is completed, and the calculation result is as shown in Figure 2 .

[0115]

[0116] where L is an arbitrary orthogonal matrix;

[0117] Q is an empirical load matrix;

[0118] Z is the fused seismic attribute volume;

[0119] T is the matrix transpose; K is the kernel matrix; J represents the Jacobian matrix.

[0120] Step 5: Color matching normalization processing is performed on each point in the three-dimensional space of the fused volume attribute Z, and each attribute value is converted into a color.

[0121]

[0122] where C(x, y, z) is the corresponding color value after normalization (between 0 and 255); z(x, y, z) t is the attribute value of any point in space; Z(x, y, z) min is the minimum attribute value in the attribute volume; z(x, y, z) max is the maximum attribute value in the attribute volume.

[0123] Step 6: Construct a lightness calculation model of the double transfer function, which first needs to establish a parameter variable template t according to the maximum and minimum values of the volume attribute, and then establish a lightness model according to t and the transfer function. The establishment methods are the cubic spline transfer function lightness model and the Gaussian transfer function lightness model. The Gaussian transfer function lightness model controls the imaging of large attribute value voxels, and the cubic spline transfer function lightness model is used to control the imaging of small attribute value voxels. The key formulas are as follows:

[0124] Cubic spline transfer function lightness model:

[0125] Q i (t)=Q i (t 3 -3t 2 +1)+Q i+1 (-2t 3 +3t 2 )+R i (t 3 -2t 2 +t)+R i+1 (t 3 -t 2 )

[0126] Gaussian transfer function lightness model:

[0127]

[0128] Q represents the position vector at the control point (i.e. the control point coordinates);

[0129] R represents the tangent vector at the control point (i.e. the tangent line or the first derivative at the control point);

[0130] t represents a parameter variable for constructing the transfer function from the maximum and minimum blending attribute values;

[0131] σ represents the variance value of the parameter variable t; t i the value of the Gaussian function parameter variable t at any point, a i represents the weighting coefficient varying with t. i

[0132] Step 7: Calculate the opacity volume. According to the opacity calculation model in step 6 and the data C after color matching conversion in step 5, calculate the opacity of each point in the space. For small attribute values, select a cubic spline opacity model for calculation. For large attribute values, select a Gaussian opacity model for calculation. Obtain the opacity voxel f(x, y, z) through the following calculation.

[0133]

[0134]

[0135] In the formula: M is an opacity model selector; P is a conversion function;

[0136] Z(x, y, z) midle is the average value.

[0137] Step 8: Light tracing and data volume space resampling calculation. Fix the position of the light source point. Perform light tracing and projection from the light source point to the opacity volume f. Since the angle between each light ray and the opacity volume is different, it is necessary to perform volume space resampling on the opacity value in the direction of each light ray, i.e. projection along each light ray path (such as Figure 3 ). The resampling calculation formula is as follows:

[0138]

[0139] In the formula: P is a certain point inside the volume space (the light incident point);

[0140] f(P) is the value after resampling of point P;

[0141] f(P1) is the opacity value at P1;

[0142] f(P2) is the opacity value at P2;

[0143] y P , y​P1 , y P2 are the spatial y coordinate values for the three points respectively.

[0144] Step 9: Select the top and bottom horizon surface of the reservoir space to generate a mask data volume, only perform light projection and accumulation on the effective mask voxels during light tracing projection, and finally accumulate and calculate the light blocking value of each point in the light propagation direction on the basis of the above, that is, the three-dimensional space description of the reservoir can be completed. Figure 4 ).

[0145]

[0146] In the formula: β i Light ray weighting coefficient related to distance;

[0147] f(i) is the light blocking value of each point calculated in step 8;

[0148] D out The light blocking value accumulated in the light direction.

[0149] Finally, it should be noted that the above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent replacements to some technical features. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0150] In addition to the technical features described in the specification, they are known to those skilled in the art.

Claims

1. A method for reservoir 3D space characterization based on dual transfer function optical density volume imaging, characterized in that, The reservoir three-dimensional space delineation method based on the dual transfer function opacity volume imaging comprises: Step 1, calculating seismic volume attributes and constructing a kernel matrix K; Step 2, compute eigenvalues λ of the kernel matrix i with eigenvector v i ; Step 3, converting from seismic original data to volume fusion attributes with reservoir characteristics, and converting each attribute value into color; Step 4, constructing a dual transfer function opacity volume calculation model; Step 5, calculating the opacity volume, and performing volume space resampling on the opacity volume value in each light direction; Step 6, performing cumulative calculation on the opacity voxel value of each point in the light propagation direction to complete the reservoir three-dimensional space description; In step 5, according to the opacity volume calculation model in step 4 and the data C after color conversion in step 3, the opacity volume of each point in the space is calculated, and the small attribute value selects the cubic spline opacity volume model for calculation, and the large attribute value selects the Gaussian opacity volume model for calculation; In step 5, the formula for calculating the opacity voxel f(x, y, z) is: where: M is a lightness model selector; P(c) is a transfer function; Z(x,y,z) is the attribute value at any point in space; Z(x,y,z) midle is the average value of Z(x,y,z). In step 5, when performing light tracing and data volume space resampling calculation, the light source point position is fixed, and light tracing and projection are performed from the light source point to the opacity volume f. Since the angle between each light and the opacity volume is different, it is necessary to perform volume space resampling on the opacity volume value in each light direction, that is, to project along each light path; In step 5, the resampling calculation formula is as follows: In the formula, P is a point inside the volume space, that is, the light incidence point; f(P) is the value after resampling of P; f(P1) is the opacity volume value at P1; f(P2) is the opacity volume value at P2; y P 、y P1 、y P2 are the spatial y coordinate values of the three points p, p1, p2, respectively.

2. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 1, characterized in that, In step 1, according to the original seismic data volume u(x, y, z), the seismic volume attributes are calculated to obtain a plurality of volume attributes, including: S1(x, y, z) is the mean square amplitude volume attribute S2(x, y, z) is the waveform variable area volume attribute S3(x, y, z) is the energy half-time volume attribute S4(x, y, z) is the composite envelope difference volume attribute.

3. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 2, characterized in that, In step 1, according to the above seismic volume attributes to be fused, normalized calculation is performed, and a polynomial function is selected as the kernel function to construct the kernel matrix K, and the kernel function mathematical model is as follows: k(S i , S j ) = (aS i T S j +c) n In the formula, k is a polynomial kernel function; a and c are polynomial coefficients, constants; S i , S j is the body property to be fused; T is the transpose computation; n is the power number.

4. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 1, characterized in that, In step 2, the eigenvalues λ of the kernel matrix are calculated i with the eigenvectors v i , and the eigenvalues λ i and the eigenvectors v i are arranged in descending order of value, and the first q eigenvalues λ i with the largest contribution rate and the corresponding eigenvectors v i are taken.

5. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 4, characterized in that, At step 2, by λ i and v i The formula is arranged from large to small value: q = diag(λ1, λ2, λ3...λ q )​ ∨ q = diag(V1, V2, V3...V q ) wherein: ∧ q is a diagonal matrix composed of eigenvalues; V q A set of eigenvectors corresponding to the eigenvalues.

6. The method of claim 1, wherein the method is a dual transfer function optical density volume imaging method for reservoir 3D space characterization. In step 3, after removing redundant data, seismic attribute volume fusion calculation is performed, and the above volume attributes to be fused are fused into an attribute volume with reservoir characteristics, and conversion from seismic original data to volume fusion attributes with reservoir characteristics is completed.

7. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 6, characterized in that, In step 3, the color normalization processing is performed on each point in the fused seismic volume attribute Z three-dimensional space, and each attribute value is converted into color: where C(x, y, z) is the corresponding normalized color value; Z(x, y, z) min is the minimum attribute value in the attribute volume; Z(x, y, z) max is the maximum attribute value in the attribute volume.

8. The method of claim 1, wherein the method is a dual transfer function optical density volume imaging method for reservoir 3D space characterization. In step 4, first, the parameter variable t of the transfer function is established according to the maximum and minimum fused volume attribute values to establish the opacity volume model; the establishment method is a cubic spline transfer function opacity volume model and a Gaussian transfer function opacity volume model, the Gaussian transfer function opacity volume model controls the imaging of large attribute value voxels, and the cubic spline transfer function opacity volume model is used for controlling the imaging of small attribute value voxels.

9. The method for reservoir 3D space characterization based on dual transfer function optical density volume imaging according to claim 8, characterized in that, In step 4, the cubic spline transfer function opacity volume model is: Q i (t) = Q i (t 3 - 3t 2 + 1) + Q i+1 (-2t 3 + 3t 2 ) + R i (t 3 - 2t 2 + t) + R i+1 (t 3 - t 2 ) The Gaussian transfer function opacity volume model is: In the formula, Q represents a position vector at a control point, i.e. a control point coordinate; R represents a tangent vector at the control point, i.e. a tangent line or a first derivative at the control point; σ denotes a variance value of the parameter variable t; t i the value of the Gaussian function parameter variable t at any point, a i denotes a weighting coefficient that varies with t i .

10. The method of claim 1, wherein the method is a dual transfer function optical density volume imaging method for reservoir 3D space characterization. In step 6, a mask data volume is generated by selecting top and bottom horizon surfaces of a reservoir space, and only effective mask voxels are subjected to light projection and accumulation in light tracing projection, and finally, each point opacity voxel value in a light propagation direction is calculated by accumulation on the basis of the above, so that the three-dimensional space description of the reservoir is completed.

11. The method of claim 10, wherein the method is a dual transfer function opacity volume imaging based reservoir 3D space characterization method. In step 6, the formula for calculating each point opacity voxel value by accumulation is as follows: where: β i is a distance-dependent light weighting factor; f(i) is each point opacity value calculated in step 5; D out is the accumulated opacity value in the direction of the light ray.

Images