Rock reservoir intelligent three-dimensional modeling method and device based on well drilling and rock core data

Abstract

The invention provides a rock reservoir intelligent three-dimensional modeling method and device based on well drilling and rock core data, and the method comprises the steps: segmenting a rock reservoir feature image and a microstructure feature image based on a color gradation difference value, and obtaining a structural feature under an engineering scale and a microstructure feature under a microscopic scale; constructing a sum training diffusion generation model based on the structural features under the engineering scale and the microstructure features under the microscopic scale, and generating a preparation model of the target area through the trained diffusion generation model; and under the condition that the matching rate of the reservoir parameter characteristics of the preparatory model and the reservoir parameter characteristics of the target area is greater than a preset matching rate threshold value, obtaining a rock reservoir three-dimensional model of the engineering scale of the target area based on the preparatory model. According to the method, scale limitation is broken, artificial reservoir model feature selection is reduced, and intelligent construction of the deep rock reservoir is realized.

Description

Technical Field

[0001] The present invention relates to the field of deep learning technology, and in particular to a method and device for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data. Background Art

[0002] Carrying out deep rock reservoir engineering projects to exploit deep resources (such as shale gas, geothermal energy and natural hydrates, etc.) to solve the energy crisis and CO 2 Geological storage plays a very important role, and the research results have important application and economic value for the healthy and sustainable development of the national economy.

[0003] At present, the difficulty of deep rock reservoir engineering projects lies mainly in how to break through the "black box" structure of deep rock reservoirs and establish an accurate 3D model of rock reservoirs, so as to evaluate the resource content of deep rock reservoirs, determine resource exploitation methods, and maintain the stability and safety of long-term exploitation of reservoir resources. The existing deep rock reservoir 3D modeling technology mainly includes three categories of 3D modeling technology: geological statistics (deterministic modeling and stochastic modeling technology), sedimentary simulation (rule modeling and comprehensive geological modeling) and analysis-driven (data-driven reservoir modeling and concept-driven modeling). Although the existing 3D modeling technology can establish a 3D model of rock reservoirs, there are still a series of problems in the existing 3D modeling technology in data accuracy, scale up-downgrading strategy and modeling accuracy.

[0004] Among them, the existing geological statistics 3D modeling technology cannot face the problem of missing intermediate drilling data, resulting in low data accuracy; the existing sedimentary simulation 3D modeling technology cannot be applied to 3D reservoir models with complex strata distribution, resulting in limitations in its technical application; the existing analysis-driven 3D modeling technology is overly dependent on a large amount of basic exploration data, and the lack of reservoir strata basic data will lead to low accuracy of the established 3D model; the existing 3D modeling technology requires model scale dimensionality increase and dimensionality reduction processing, and the fusion and nesting of models of different scales still cannot accurately and effectively characterize rock reservoir properties. At the same time, the existing 3D modeling technology requires a lot of manual intervention and has a low degree of intelligence.

[0005] In other words, existing modeling technologies still have defects in data accuracy and intelligent modeling. Summary of the invention

[0006] The present invention provides a method and device for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data, so as to solve the defects of the modeling basis in the prior art in terms of insufficient data accuracy and intelligent modeling, and realize a three-dimensional modeling method of rock reservoirs with higher modeling accuracy and intelligence.

[0007] The present invention provides a method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data, comprising: Use ultrasonic imaging to obtain the rock reservoir characteristic image of the target area, and obtain the microscopic structure characteristic image based on the core specimens extracted from the target area; Segment the rock reservoir characteristic image and the microscopic structure characteristic image based on the color level difference to obtain the structural characteristics at the engineering scale and the microscopic structure characteristics at the microscopic scale respectively; Establish and train a diffusion generation model based on the structural characteristics at the engineering scale and the microscopic structure characteristics at the microscopic scale, and generate a preliminary model of the target area through the trained diffusion generation model; When the matching rate between the reservoir parameter characteristics of the preliminary model and the reservoir parameter characteristics of the target area is greater than the preset matching rate threshold, obtain the three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model.

[0008] According to an intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the present invention, the steps of using ultrasonic imaging to obtain the rock reservoir characteristic image of the target area and obtaining the microscopic structure characteristic image based on the core specimens extracted from the target area specifically include: Determine a central point drilling position and four boundary point drilling positions in the target area based on engineering requirements and geological exploration data; At the four boundary point drilling positions, use the method of ultrasonic imaging while drilling in the direction from the boundary point drilling position to the central point drilling position to collect the rock reservoir characteristic image of the target area; Based on the cores obtained during drilling at the four boundary point drilling positions and the central point drilling position, prepare core specimens, and use high-resolution imaging technology to obtain the microscopic structure characteristic image.

[0009] According to an intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the present invention, the steps of segmenting the rock reservoir characteristic image and the microscopic structure characteristic image based on the color level difference to obtain the structural characteristics at the engineering scale and the microscopic structure characteristics at the microscopic scale respectively specifically include: Convert the rock reservoir characteristic image and the microscopic structure characteristic image of the target area into rock color images; Based on the color level difference between different structural characteristics, divide the converted rock reservoir characteristic image of the target area, and obtain the structural characteristics at the engineering scale based on the division result; Based on the color level difference between different microscopic structure characteristics, divide the converted microscopic structure characteristic image, and obtain the microscopic structure characteristics at the microscopic scale based on the division result.

[0010] A method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data according to the present invention, the structural features at the engineering scale include formation boundaries, large-scale fractures, large-scale cavities, and fluids; the microstructural features at the microscale include pores, microcracks, micro minerals, and pore fluids of reservoir rocks in different formations.

[0011] A method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data according to the present invention, the steps of establishing and training a diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscale specifically include: Construct a noise distribution model at the engineering scale based on the Weibull distribution, and construct a noise distribution model at the microscale based on the Lorentz distribution; Use the fusion model composed of the noise distribution model at the engineering scale and the noise distribution model at the microscale as the noise model used by the diffusion generation model during the diffusion process.

[0012] A method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data according to the present invention, before the step of obtaining a three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold, further includes: Calculate the first difference between the reservoir characteristic parameters of the preliminary model and the reservoir characteristic parameters of the target area; Use the ratio between the first difference and the reservoir characteristic parameters of the target area as the matching rate of the reservoir parameter features.

[0013] The present invention also provides an intelligent three-dimensional modeling device for rock reservoirs based on drilling and core data, including: An acquisition module, configured to obtain a rock reservoir characteristic image of the target area by using ultrasonic imaging, and obtain a microstructural characteristic image based on the core specimens extracted from the target area; A determination module, configured to segment the rock reservoir characteristic image and the microstructural characteristic image based on the color level difference, and respectively obtain the structural features at the engineering scale and the microstructural features at the microscale; A generation module, configured to establish and train a diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscale, and generate a preliminary model of the target area through the trained diffusion generation model; A reconstruction module, configured to obtain a three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data as described in any one of the above.

[0015] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data as described in any one of the above.

[0016] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data as described in any one of the above.

[0017] The intelligent three-dimensional modeling method and device of the rock reservoir based on drilling and core data provided by the present invention realize the reconstruction of the three-dimensional model of the engineering-scale rock reservoir based on the structural characteristics at the engineering scale and the microstructural characteristics at the microscale of the target area by constructing a diffusion generation model, break the scale limitation, reduce the artificial reservoir model feature selection, realize the intelligent determination of the reservoir model feature parameters, and realize the intelligent construction of the deep rock reservoir. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a schematic flowchart of the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data provided by the present invention; Figure 2 In (a) of is the divided rock reservoir feature image in the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data provided by the present invention; Figure 2 In (b) of is the divided microstructural feature image in the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data provided by the present invention; Figure 3 In (a) of is the reconstructed three-dimensional model of the engineering-scale rock reservoir in the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data provided by the invention; Figure 3Among them, (b) is a three-dimensional model of a fractured rock mass of a certain formation in the three-dimensional model of a rock reservoir at the engineering scale reconstructed in the intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the invention; Figure 3 Among them, (c) is a microscopic scale model of the core in the fractured rock mass in the three-dimensional model of a rock reservoir at the engineering scale reconstructed in the intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the invention; Figure 4 is a schematic structural diagram of the rock reservoir intelligent three-dimensional modeling device provided by the present invention; Figure 5 is a schematic structural diagram of the electronic device provided by the present invention. Detailed implementation manners

[0020] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0021] The following combines Figures 1 to 3 to introduce the intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data of the present invention. As Figure 1 shown, it includes: Step 101, obtaining a characteristic image of a rock reservoir in a target area by ultrasonic imaging, and obtaining a microscopic structure characteristic image based on core specimens extracted from the target area; Ultrasonic imaging generates an image of the internal structure of the rock reservoir in the target area through the characteristics of ultrasonic waves when propagating in the rock reservoir in the target area. Since ultrasonic waves can propagate in all directions, it can be understood that the characteristic image of the rock reservoir in the target area obtained by ultrasonic imaging is a three-dimensional image representing the engineering scale structural characteristics of the rock reservoir in the target area.

[0022] Drill and take cores at different positions and / or depths in the target area, and prepare core specimens based on the extracted core samples, so as to obtain a number of core specimens that can represent the structural characteristics of different formations in the target area.

[0023] Use high-precision imaging technology to image the prepared core specimens to obtain a microscopic structure characteristic image for representing the microscopic structure characteristics of the target area.

[0024] Step 102, segmenting the rock reservoir characteristic image and the microscopic structure characteristic image based on the color level difference to respectively obtain the structural characteristics at the engineering scale and the microscopic structure characteristics at the microscopic scale; Optionally, convert the rock reservoir feature image and the microstructure feature image into a rock color image.

[0025] In the rock color image of the rock storage feature image and the microstructure feature image, different structures correspond to different color scale values.

[0026] Therefore, based on the color scale value of each pixel point in the rock color image of the rock reservoir feature image, divide the rock reservoir feature image, determine the pixel points corresponding to each type of structure in the rock reservoir feature image, so as to obtain the structural features at the engineering scale, such as Figure 2 shown in (a) of

[0027] Similarly, divide in the microstructure feature image, determine the pixel points corresponding to each type of microstructure, so as to obtain the microstructure features at the microscale, such as Figure 2 shown in (b) of

[0028] Step 103, establish and train a diffusion generation model based on the structural features at the engineering scale and the microstructure features at the microscale, and generate a preliminary model of the target area through the trained diffusion generation model; Optionally, fit a noise distribution model at the engineering scale based on the structural features at the engineering scale, and fit a noise distribution model at the microscale based on the microstructure features at the microscale.

[0029] Add noise to the blank image based on the two fitted noise distribution models, so that the image with added noise can simultaneously represent the structural features at the engineering scale and the microstructure features at the microscale.

[0030] During the training process, compare the structural features at the engineering scale and the microstructure features at the microscale of the image with added noise with the real structural features at the engineering scale and the microstructure features at the microscale of the obtained target area, and optimize the hyperparameters of the diffusion generation model based on the difference to obtain the optimal diffusion generation model as the trained diffusion generation model.

[0031] Optionally, in the training stage, adjust hyperparameters such as the number of training times and the learning rate, construct a data set based on the structural features at the engineering scale and the microstructure features at the microscale obtained, and use 30% of them for training to obtain a diffusion generation model.

[0032] It can be understood that the preliminary model generated by the trained diffusion generation model can simultaneously reflect the structural features at the engineering scale and the microstructure features at the microscale.

[0033] Step 104, when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than the preset matching rate threshold, a three-dimensional model of the rock reservoir at the engineering scale of the target area is obtained based on the preliminary model.

[0034] Optionally, 70% of the remaining data in the dataset is used as the input of the trained diffusion generation model to generate several preliminary models, and the reservoir parameter feature matching rate between the reservoir parameter features of each preliminary model and the reservoir parameter features of the target area is calculated.

[0035] When there is a preliminary model with a reservoir parameter feature matching rate greater than the preset matching rate threshold, it is considered that the reservoir feature parameters of this preliminary model are similar to the reservoir feature parameters of the target area. Therefore, a three-dimensional model of the rock reservoir at the engineering scale of the target area can be constructed based on this preliminary model.

[0036] Among them, the preset matching rate threshold is an empirical value determined based on the calculation method of the matching rate.

[0037] It can be understood that the preliminary model is actually feature data containing the generated structural features at the engineering scale and microstructural features at the microscale. Based on the matched preliminary model, the Meshgrid function of MATLAB software is used for 3D visualization display, and finally a three-dimensional model of the rock reservoir at the engineering scale of the target area is obtained.

[0038] Optionally, the three-dimensional model of the rock reservoir at the engineering scale of the target area is exported in the format of ".stl" using MATLAB software and imported into other numerical software, and then the reconstructed model can be used for numerical simulation analysis of the rock reservoir in the target area.

[0039] The present invention realizes the reconstruction of a three-dimensional model of an engineering-scale rock reservoir based on the structural features at the engineering scale and the microstructural features at the microscale of the target area by constructing a diffusion generation model, breaks the scale limitation, reduces the artificial selection of reservoir model features, realizes the intelligent determination of reservoir model feature parameters, and realizes the intelligent construction of deep rock reservoirs.

[0040] In the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data of the present invention, the steps of obtaining the rock reservoir feature image of the target area by ultrasonic imaging and obtaining the microstructural feature image based on the core specimens extracted from the target area specifically include: Determine a central point drilling position and four boundary point drilling positions in the target area based on engineering requirements and geological exploration data; At the four boundary point drilling positions, the rock reservoir feature image of the target area is collected by using the method of ultrasonic imaging while drilling in the direction from the boundary point drilling position to the central point drilling position; Based on the cores obtained during drilling at the four boundary point drilling positions and the central point drilling position, core specimens are prepared, and high-resolution imaging technology is used to obtain the microscopic structure feature images.

[0041] According to the engineering requirements and geological exploration data, a representative central point drilling position and four boundary point drilling positions are selected in the target area, and the area determined by the connection lines of the four boundary point drilling positions is used as the engineering research area.

[0042] In this embodiment, a research area with a length of 100 meters, a width of 100 meters, and a depth of 1000 meters is determined. On this basis, the length direction of the research area is taken as the X-axis direction, the width direction as the Y-axis direction, and the depth direction as the Z-axis direction, and a research area coordinate system is constructed with any one of the boundary point drilling positions as the origin.

[0043] Therefore, at the central point drilling position and the four boundary point drilling positions, drillings with a depth from 0 to 1000 meters are respectively carried out along the Z-axis direction.

[0044] Optionally, for each boundary point drilling position, the high-precision ultrasonic imaging technology while drilling is adopted, the imaging resolution is set to 1 cm, and fine high-precision imaging is carried out in the direction pointing to the central point drilling position to obtain the rock reservoir feature images. In this embodiment, four rock reservoir feature images are correspondingly obtained.

[0045] Optionally, after obtaining the rock reservoir feature images, the drilling cores are extracted and X-ray tomography (XCT) imaging standard core specimens are prepared, and high-resolution XCT imaging technology is used to capture the microscopic structure feature images of the core specimens at a resolution of 2 μm of the resolution.

[0046] In this embodiment, core specimens with a diameter of 10 mm and a height of 20 mm are prepared.

[0047] Through the above method, the rock reservoir feature images that can characterize the structural features of the target area and the microscopic structure feature images that can characterize the microscopic structure features of the target area can be conveniently obtained in the target area.

[0048] In the intelligent three-dimensional modeling method of rock reservoir based on drilling and core data of the present invention, the step of segmenting the rock reservoir feature images and the microscopic structure feature images based on the color scale difference to respectively obtain the structural features at the engineering scale and the microscopic structure features at the microscopic scale specifically includes: Converting the rock reservoir feature images and microscopic structure feature images of the target area into rock color images; Based on the color scale difference between different structural features, dividing the converted rock reservoir feature images of the target area, and obtaining the structural features at the engineering scale based on the division results; Based on the color level difference between different microstructural features, the converted microstructural feature images are divided, and the microstructural features at the microscale are obtained based on the division results.

[0049] In order to be able to extract the features of different microstructures in the rock reservoir feature image and the microstructural feature image of the target area, in this embodiment, the Rock Color Image Digital Difference Order Segmentation Algorithm (RCI-DDLS) is adopted to achieve the fine division of the microstructural features of rocks in different formation reservoirs.

[0050] Specifically, first, the Watershed algorithm and the Label2RGB function in MATLAB software are used to convert the rock reservoir feature image and the microstructural feature image of the target area into a rock color image. .

[0051] On this basis, the following formula is used to achieve the division of structural features based on the color level difference: ; ; ; In the formula, is a color pixel unit, is the pixel color level value, which is between [0, 255]; is the numerical value of the depth boundary of the rock reservoir at the engineering scale; is the large-scale fracture image; is the large-scale cavity image; is the fluid image; is the p th color pixel unit.

[0052] is the micro-pore image of rocks in different formation reservoirs; is the micro-crack image; is the micro-pore fluid image; is the micro-mineral image; is the p th color pixel representing the q -7 ( q ≥8) sub-mineral phases.

[0053] Generally speaking, the selected structural features at the engineering scale in this embodiment include the formation boundary, large-scale fractures, large-scale cavities and fluids. Therefore, the areas corresponding to large-scale fractures, large-scale cavities and fluids are divided in the rock reservoir feature image through color level segmentation, and the formation boundary is determined based on the division results. Among them, the formation boundary is marked with height values in the constructed coordinate system.

[0054] The selected microstructural features at the microscale include pores, microcracks, micro minerals, and pore fluids in reservoir rocks of different strata. Therefore, regions corresponding to pores, microcracks, micro minerals, and pore fluids of each layer are demarcated in the microstructural feature map.

[0055] It can be understood that since the structural features of the target region are different, and the structural features at the selected engineering scale are different from the microstructural features at the microscale, the above formula can be adjusted accordingly.

[0056] In the intelligent three-dimensional modeling method of rock reservoirs based on drilling and core data of the present invention, the step of establishing and training a diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscale specifically includes: Construct a noise distribution model at the engineering scale based on the Weibull distribution, and construct a noise distribution model at the microscale based on the Lorentz distribution; Use the fusion model composed of the noise distribution model at the engineering scale and the noise distribution model at the microscale as the noise model used by the diffusion generation model during the diffusion process.

[0057] In order to reconstruct the three-dimensional model of the rock reservoir at the engineering scale based on the structural features at the engineering scale and the microstructural features at the microscale obtained by partitioning, in this embodiment, a deep learning diffusion generation model (DL-DGM) based on the Weibull distribution noise at the engineering scale and the Lorentz distribution noise at the microscale is constructed and trained.

[0058] Specifically, in the DL-DGM model of this embodiment, the noise distribution model therein adopts a fusion noise model composed of Weibull distribution noise and Lorentz distribution noise; the DL-DGM model includes two processes: forward noise superposition and reverse reservoir model generation, and their corresponding expressions are respectively: ; ; In the formula, represents the Weibull probability density function, represents the Lorentz probability density function, represents the fusion noise model; α is the shape parameter of the Weibull distribution, m is the scale parameter, is the location parameter of the Weibull distribution.

[0059] is the reservoir feature image data set, is the predicted reservoir feature image data set, is the hyperparameter that controls the noise increment at each step, To conform to the overall noise model of a given rock reservoir dataset, is the predicted noise model of the rock reservoir dataset.

[0060] Among them, the reservoir feature image dataset is composed of the reservoir structure feature map and the microstructure feature map after feature division.

[0061] That is to say, at the engineering scale, based on the reservoir structure feature map after division, various parameters of the Weibull distribution are determined, so that the Weibull noise distribution model can reflect the structural characteristics of the target area at the engineering scale; at the microscale, based on the microstructure feature map after division, various parameters of the Lorentz distribution are determined, so that the Lorentz noise distribution model can reflect the microstructure characteristics of the target area at the microscale.

[0062] Based on the above two types of noise distribution models, a fusion noise model used in the diffusion process of the diffusion generation model is constructed to add noise at each step of the diffusion process.

[0063] In the above way, hyperparameters such as the number of training times and the learning rate are adjusted to obtain the optimized DL-DGM model for generating the preliminary model.

[0064] In this embodiment, the specific size of the center-boundary point division area is set, the engineering scale (CBP-REM) model range of the center-boundary point area is determined, and the CBP-REM model is generated using the DL-DGM model.

[0065] In the intelligent three-dimensional modeling method of the rock reservoir based on drilling and core data of the present invention, before the step of obtaining the three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model when the matching rate of the reservoir parameter characteristics of the preliminary model and the reservoir parameter characteristics of the target area is greater than the preset matching rate threshold, it further includes: Calculating a first difference between the reservoir feature parameters of the preliminary model and the reservoir feature parameters of the target area; Taking the ratio between the first difference and the reservoir feature parameters of the target area as the matching rate of the reservoir parameter characteristics.

[0066] It can be understood that when the reservoir parameter characteristics of the generated preliminary model are similar to the real reservoir parameter characteristics of the target area, it can be considered that the preliminary model can accurately reflect the reservoir parameter characteristics of the target area. Therefore, in this embodiment, a calculation method for the matching rate CMr of the reservoir parameter characteristics is provided: ; In the formula, represents the matching rate of the reservoir parameter characteristics, Denote the reservoir prediction parameters, which are determined by the preliminary model; Denote the true reservoir parameters.

[0067] Optionally, the reservoir characteristic parameters include fracture rate, size, etc. at the engineering scale and microscale.

[0068] That is to say, Namely the first difference.

[0069] Based on the above calculation method of CMr, in this embodiment, the preset matching rate threshold is determined to be 90%. That is to say, when the matching rate of reservoir parameter characteristics is greater than 90%, a three-dimensional model of the rock reservoir at the engineering scale of the target area can be reconstructed based on the corresponding preliminary model, as Figure 3 shown.

[0070] Next, a three-dimensional modeling device for rock reservoirs based on drilling and core data provided by the present invention will be described. The three-dimensional modeling device for rock reservoirs based on drilling and core data described below can be correspondingly referred to the intelligent three-dimensional modeling method for rock reservoirs based on drilling and core data described above.

[0071] As Figure 4 shown, the three-dimensional modeling device for rock reservoirs based on drilling and core data includes an acquisition module 401, a determination module 402, a generation module 403, and a reconstruction module 404; The acquisition module 401 is configured to acquire a characteristic image of the rock reservoir in the target area by using ultrasonic imaging, and acquire a microstructural characteristic image based on the core specimens extracted from the target area; Ultrasonic imaging generates an image of the internal structure of the rock reservoir in the target area through the characteristics of ultrasonic waves propagating in the rock reservoir in the target area. Since ultrasonic waves can propagate in all directions, it can be understood that the characteristic image of the rock reservoir in the target area obtained based on ultrasonic imaging is a three-dimensional image representing the engineering-scale structural characteristics of the rock reservoir in the target area.

[0072] Drill and take cores at different positions and / or depths in the target area, and prepare core specimens based on the extracted core samples, so as to obtain a number of core specimens that can represent the structural characteristics of different strata in the target area.

[0073] Image the prepared core specimens by using high-precision imaging technology to obtain a microstructural characteristic image for representing the microstructural characteristics of the target area.

[0074] The determination module 402 is configured to segment the rock reservoir characteristic image and the microstructural characteristic image based on the color scale difference, and respectively obtain the structural characteristics at the engineering scale and the microstructural characteristics at the microscale; Optionally, convert the rock reservoir feature image and the microstructure feature image into a rock color image.

[0075] In the rock color image of the rock storage feature image and the microstructure feature image, different structures correspond to different color scale values.

[0076] Therefore, based on the color scale value of each pixel in the rock color image of the rock reservoir feature image, divide the rock reservoir feature image, determine the pixels corresponding to each type of structure in the rock reservoir feature image, so as to obtain the structural features at the engineering scale, such as Figure 2 shown in (a) of

[0077] Similarly, divide in the microstructure feature image, determine the pixels corresponding to each type of microstructure, so as to obtain the microstructure features at the microscale, such as Figure 2 shown in (b) of

[0078] The generation module 403 is used to establish and train a diffusion generation model based on the structural features at the engineering scale and the microstructure features at the microscale, and generate a preliminary model of the target area through the trained diffusion generation model; Optionally, fit a noise distribution model at the engineering scale based on the structural features at the engineering scale, and fit a noise distribution model at the microscale based on the microstructure features at the microscale.

[0079] Add noise to the blank image based on the two fitted noise distribution models, so that the image with added noise can simultaneously represent the structural features at the engineering scale and the microstructure features at the microscale.

[0080] During the training process, compare the structural features at the engineering scale and the microstructure features at the microscale of the image with added noise with the real structural features at the engineering scale and the microstructure features at the microscale of the obtained target area, and optimize the hyperparameters of the diffusion generation model based on the difference to obtain the optimal diffusion generation model as the trained diffusion generation model.

[0081] Optionally, in the training stage, adjust hyperparameters such as the number of training times and the learning rate, construct a data set based on the structural features at the engineering scale and the microstructure features at the microscale obtained, and use 30% of them for training to obtain a diffusion generation model.

[0082] It can be understood that the preliminary model generated by the trained diffusion generation model can simultaneously reflect the structural features at the engineering scale and the microstructure features at the microscale.

[0083] A reconstruction module 404, configured to obtain a three-dimensional model of a rock reservoir at the engineering scale of the target area based on the preliminary model when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold.

[0084] Optionally, use the remaining 70% of the dataset as the input to the trained diffusion generation model to generate a number of preliminary models, and calculate the reservoir parameter feature matching rate between the reservoir parameter features of each preliminary model and the reservoir parameter features of the target area.

[0085] In the case where there is a preliminary model with a reservoir parameter feature matching rate greater than the preset matching rate threshold, it is considered that the reservoir feature parameters of this preliminary model are similar to the reservoir feature parameters of the target area. Therefore, a three-dimensional model of a rock reservoir at the engineering scale of the target area can be constructed based on this preliminary model.

[0086] Among them, the preset matching rate threshold is an empirical value determined based on the calculation method of the matching rate.

[0087] It can be understood that the preliminary model is actually feature data containing the generated structural features at the engineering scale and microstructural features at the microscale. Based on the matched preliminary model, the Meshgrid function of MATLAB software is used for 3D visualization display, and finally a three-dimensional model of a rock reservoir at the engineering scale of the target area is obtained.

[0088] Optionally, use MATLAB software to export the three-dimensional model of the rock reservoir at the engineering scale of the target area in the ".stl" format and import it into other numerical software, then the reconstructed model can be used for numerical simulation analysis of the rock reservoir in the target area.

[0089] The present invention realizes the reconstruction of a three-dimensional model of a rock reservoir at the engineering scale based on the structural features at the engineering scale and the microstructural features at the microscale of the target area by constructing a diffusion generation model, breaks the scale limitation, reduces the artificial selection of reservoir model features, realizes the intelligent determination of reservoir model feature parameters, and realizes the intelligent construction of deep rock reservoirs.

[0090] Figure 5 Illustrates a schematic diagram of the physical structure of an electronic device, such as Figure 5As shown in the figure, the electronic device may include: a processor 510, a communications interface 520, a memory 530, and a communication bus 540. Among them, the processor 510, the communications interface 520, and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logical instructions in the memory 530 to execute an intelligent three-dimensional modeling method for rock reservoirs based on drilling and core data. The method includes: obtaining a rock reservoir feature image of a target area by using ultrasonic imaging, and obtaining a microscopic structure feature image based on a core specimen extracted from the target area; segmenting the rock reservoir feature image and the microscopic structure feature image based on the color level difference to respectively obtain a structural feature at the engineering scale and a microscopic structure feature at the microscopic scale; establishing and training a diffusion generation model based on the structural feature at the engineering scale and the microscopic structure feature at the microscopic scale, and generating a preliminary model of the target area through the trained diffusion generation model; and obtaining a three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold.

[0091] In addition, when the logical instructions in the above-mentioned memory 530 are implemented in the form of software functional units and sold or used as an independent product, they may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc that can store program codes.

[0092] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the above-mentioned various methods. The method includes: obtaining a rock reservoir feature image of a target area by using ultrasonic imaging, and obtaining a microscopic structure feature image based on core specimens extracted from the target area; segmenting the rock reservoir feature image and the microscopic structure feature image based on the color level difference to respectively obtain a structural feature at the engineering scale and a microscopic structure feature at the microscopic scale; establishing and training a diffusion generation model based on the structural feature at the engineering scale and the microscopic structure feature at the microscopic scale, and generating a preliminary model of the target area through the trained diffusion generation model; when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold, obtaining a three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model.

[0093] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the intelligent three-dimensional modeling method of a rock reservoir based on drilling and core data provided by the above-mentioned various methods. The method includes: obtaining a rock reservoir feature image of a target area by using ultrasonic imaging, and obtaining a microscopic structure feature image based on core specimens extracted from the target area; segmenting the rock reservoir feature image and the microscopic structure feature image based on the color level difference to respectively obtain a structural feature at the engineering scale and a microscopic structure feature at the microscopic scale; establishing and training a diffusion generation model based on the structural feature at the engineering scale and the microscopic structure feature at the microscopic scale, and generating a preliminary model of the target area through the trained diffusion generation model; when the matching rate between the reservoir parameter features of the preliminary model and the reservoir parameter features of the target area is greater than a preset matching rate threshold, obtaining a three-dimensional model of the rock reservoir at the engineering scale of the target area based on the preliminary model.

[0094] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor.

[0095] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solutions, in essence, or the parts that contribute to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. And these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data, characterized in that: include: Ultrasonic imaging is used to obtain rock reservoir characteristic images in the target area, and microstructure characteristic images are obtained based on core specimens extracted from the target area; Segmenting the rock reservoir characteristic image and the microstructure characteristic image based on color scale difference to obtain structural characteristics at an engineering scale and microstructure characteristics at a microscale, respectively; Establishing and training a diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscopic scale, and generating a preliminary model of the target area through the trained diffusion generation model; When the matching rate between the reservoir parameter characteristics of the preparation model and the reservoir parameter characteristics of the target area is greater than a preset matching rate threshold, a three-dimensional rock reservoir model of the target area at an engineering scale is obtained based on the preparation model.

2. The intelligent three-dimensional modeling method of rock reservoir based on drilling and core data according to claim 1 is characterized in that: The step of using ultrasonic imaging to obtain a rock reservoir characteristic image of a target area and obtaining a microstructure characteristic image based on a core specimen extracted from the target area specifically includes: Determine a central drilling location and four boundary drilling locations in the target area based on engineering requirements and geological survey data; At the four boundary drilling positions, ultrasonic imaging while drilling is used in a direction from the boundary drilling position to the center drilling position to acquire a rock reservoir characteristic image of the target area; Based on the rock cores obtained during drilling at the four boundary point drilling positions and the center point drilling position, the rock core specimens are prepared, and the microstructure characteristic images are obtained by using high-resolution imaging technology.

3. The intelligent three-dimensional modeling method of rock reservoir based on drilling and core data according to claim 1 is characterized in that: The step of segmenting the rock reservoir characteristic image and the microstructure characteristic image based on the color scale difference to obtain the structural characteristics at the engineering scale and the microstructure characteristics at the microscale, respectively, specifically includes: Converting the rock reservoir characteristic image and the microstructure characteristic image of the target area into a rock color image; Based on the color scale difference between different structural features, the rock reservoir characteristic image of the target area after conversion is divided, and the structural features at the engineering scale are obtained based on the division result; Based on the color scale difference between different microstructure features, the converted microstructure feature image is divided, and the microstructure features at the microscopic scale are obtained based on the division result.

4. The intelligent three-dimensional modeling method of rock reservoir based on drilling and core data according to claim 3 is characterized in that: The structural features at the engineering scale include stratum boundaries, large-scale cracks, large-scale cavities and fluids; the microstructural features at the microscopic scale include pores, microscopic cracks, microscopic minerals and pore fluids in reservoir rocks of different strata.

5. The intelligent three-dimensional modeling method for rock reservoirs based on drilling and core data according to any one of claims 1 to 4, characterized in that: The step of establishing and training the diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscopic scale specifically includes: The noise distribution model at the engineering scale is constructed based on the Weibull distribution, and the noise distribution model at the microscopic scale is constructed based on the Lorentz distribution. The fusion model composed of the engineering-scale noise distribution model and the micro-scale noise distribution model is used as the noise model used by the diffusion generation model in the diffusion process.

6. The intelligent three-dimensional modeling method for rock reservoirs based on drilling and core data according to any one of claims 1 to 4, characterized in that: In the case where the matching rate between the reservoir parameter characteristics of the preparation model and the reservoir parameter characteristics of the target area is greater than a preset matching rate threshold, before the step of obtaining the engineering-scale rock reservoir three-dimensional model of the target area based on the preparation model, the method further includes: Calculating a first difference between the reservoir characteristic parameters of the preparation model and the reservoir characteristic parameters of the target area; The ratio between the first difference and the reservoir characteristic parameter of the target area is used as the reservoir parameter characteristic matching rate.

7. An intelligent three-dimensional modeling device for rock reservoirs based on drilling and core data, characterized in that: include: An acquisition module is used to obtain a rock reservoir characteristic image of a target area by using ultrasonic imaging, and to obtain a microstructure characteristic image based on a core specimen extracted from the target area; A determination module, used for segmenting the rock reservoir characteristic image and the microstructure characteristic image based on the color scale difference, and obtaining the structural characteristics at the engineering scale and the microstructure characteristics at the microscale respectively; A generation module, used to establish and train a diffusion generation model based on the structural features at the engineering scale and the microstructural features at the microscale, and to generate a preliminary model of the target area through the trained diffusion generation model; A reconstruction module is used to obtain a three-dimensional rock reservoir model of the target area at an engineering scale based on the preparation model when the matching rate between the reservoir parameter characteristics of the preparation model and the reservoir parameter characteristics of the target area is greater than a preset matching rate threshold.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the intelligent three-dimensional modeling method of rock reservoir based on drilling and core data as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for intelligent three-dimensional modeling of rock reservoirs based on drilling and core data as described in any one of claims 1 to 6 is implemented.

Images