Two-dimensional data generation and modeling method and system for point bar of meandering river

Abstract

The application discloses a kind of two-dimensional data generation and modeling method and system for meandering river point bar side accumulation layer, and the method comprises the following steps: preprocessing the multi-source geological data of meandering river point bar side accumulation layer, to obtain the standardized original two-dimensional data set;The original two-dimensional data set is subjected to data diversity enhancement, and the enhanced data is converted into network training format;According to the training data converted, ProGAN-2D to 3D cross-dimensional modeling network is trained and constructed, and three-dimensional meandering river point bar side accumulation layer model is generated using the ProGAN-2D to 3D cross-dimensional modeling network trained and completed.Based on the above data processing procedure, the application generates high-quality and high-diversity two-dimensional training data set by enriching data sources, designing targeted data enhancement operations and constructing cross-dimensional intelligent modeling application procedures, improves the accuracy and geological rationality of meandering river point bar side accumulation layer intelligent modeling, and provides reliable modeling data support for oil and gas field development.

Description

Technical Field

[0001] This invention relates to the field of oil and gas exploration and development technology, and in particular to a method and system for generating and modeling two-dimensional data of lateral sedimentary layers at point dams in meandering rivers. Background Technology

[0002] Lateral deposits at point bars in meandering rivers are important components of meandering river reservoirs. Their spatial distribution and geometry directly affect reservoir heterogeneity and remaining oil distribution. Accurate geological modeling of lateral deposits is crucial for the efficient development of oil and gas fields. The quality and diversity of the training dataset are key to intelligent modeling (such as generative adversarial network methods).

[0003] Currently, data on lateral deposits of point bars in meandering rivers faces the challenge of obtaining 3D training datasets, and there is also a lack of abundant 2D datasets to support effective training of intelligent models. Furthermore, 2D data on lateral deposits (muddy overburden) of point bars in multi-source meandering rivers remains limited, and there is a lack of dedicated data augmentation methods, which has become a technical challenge in the field of cross-dimensional intelligent modeling (inputting 2D data to generate 3D geological models). Therefore, how to obtain lateral sedimentary layer data of meandering river point dams and how to intelligently construct lateral sedimentary layer models of meandering river point dams based on the data are urgent problems that need to be solved. Summary of the Invention

[0004] This invention provides a two-dimensional data generation and modeling method and system for lateral accretion layers of point dams in meandering rivers. Through the full-process design of multi-source data acquisition, lateral accretion layer data enhancement, and cross-dimensional intelligent modeling application, a two-dimensional dataset suitable for intelligent modeling of lateral accretion layers of point dams in meandering rivers is generated, solving the problems of existing datasets having single sources, insufficient patterns, and poor cross-dimensional application effects.

[0005] Firstly, a method for generating and modeling two-dimensional data for lateral deposition layers at point dams in meandering rivers is provided, including: Obtain multi-source geological data of the lateral deposits at point dams in meandering rivers; The multi-source geological data are preprocessed to obtain a standardized raw two-dimensional dataset; The original two-dimensional dataset is augmented with data diversity, and the augmented data is then converted into a network training format. The ProGAN-2Dto3D cross-dimensional modeling network is constructed based on the transformed training data, and a three-dimensional meandering river point dam lateral accumulation model is generated using the trained ProGAN-2Dto3D cross-dimensional modeling network.

[0006] Secondly, a two-dimensional data generation and modeling system for lateral deposition layers of point dams in meandering rivers is provided, including: The multi-source data acquisition module is used to acquire multi-source geological data of the lateral deposits of point dams in meandering rivers; The preprocessing module is communicatively connected to the multi-source data acquisition module and is used to preprocess the multi-source geological data to obtain a standardized raw two-dimensional dataset. The data augmentation module, communicatively connected to the preprocessing module, is used to enhance the data diversity of the original two-dimensional dataset and convert the augmented data into a network training format; and, The network training module, which is communicatively connected to the data augmentation module, is used to train and construct a ProGAN-2Dto3D cross-dimensional modeling network based on the transformed training data, and to generate a three-dimensional meandering river point dam lateral accumulation model using the trained ProGAN-2Dto3D cross-dimensional modeling network.

[0007] Thirdly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the two-dimensional data generation and modeling method for lateral sedimentary layers of point dams in meandering rivers as described above.

[0008] Fourthly, embodiments of the present invention provide an electronic device, including a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor. When the processor runs the computer program, it implements the two-dimensional data generation and modeling method for lateral sedimentary layers of meandering river point dams as described above.

[0009] Compared with the prior art, the advantages of the present invention are as follows: Through a complete process design involving multi-source data acquisition, lateral accretion layer data enhancement, and cross-dimensional intelligent modeling applications, a two-dimensional dataset suitable for intelligent modeling of lateral accretion layers at point dams in meandering rivers is generated. This solves the problems of existing datasets having a single source, insufficient patterns, and poor cross-dimensional application effects, and significantly improves the accuracy and geological conformity of intelligent modeling of lateral accretion layers at point dams in meandering rivers. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating a two-dimensional data generation and modeling method for lateral deposition layers of point dams in meandering rivers according to the present invention. Figure 2 This is a planar distribution map of lateral depositional layers in meandering rivers based on satellite imagery, as presented in this invention. Figure 3 This is a schematic diagram of the planar facies diagram of the Ng44-2 layer in the second northern area of ​​the Gudao Oilfield and a well-connected section of the inner lateral sedimentary layer of the point dam. Figure 4 This is a schematic diagram of the extraction of a single muddy lateral accumulation layer according to the present invention; Figure 5 This is a schematic diagram illustrating the shortening of the length of the single argillaceous lateral accumulation layer according to the present invention; Figure 6 This is a schematic diagram of the single argillaceous lateral accumulation layer rotation operation of the present invention; Figure 7 This is a schematic diagram of the stretching operation of a single clayey lateral accumulation layer according to the present invention; Figure 8 This is a schematic diagram of the random combination of single argillaceous lateral accumulation layers of the present invention; Figure 9 This is a schematic diagram of the parameter optimization of the ProGAN-2Dto3D cross-dimensional modeling network of the present invention. Detailed Implementation

[0011] Referring now to specific embodiments of the invention, examples of which are illustrated in the accompanying drawings. Although the invention will be described in conjunction with specific embodiments, it will be understood that it is not intended to limit the invention to the described embodiments. Rather, it is intended to cover variations, modifications, and equivalents included within the spirit and scope of the invention as defined by the appended claims. It should be noted that the method steps described herein can be implemented by any functional block or functional arrangement, and any functional block or functional arrangement can be implemented as a physical entity or a logical entity, or a combination of both.

[0012] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0013] Note: The examples described below are merely specific examples and are not intended to limit the embodiments of the present invention to the specific steps, values, conditions, data, order, etc. Those skilled in the art can utilize the concept of the present invention to construct more embodiments not mentioned herein by reading this specification.

[0014] Please see Figure 1 This invention provides a method for generating and modeling two-dimensional data for lateral sedimentary layers at point dams in meandering rivers. The method includes: Step S100: Obtain multi-source geological data of the lateral deposits of the meandering river point dam; The multi-source geological data includes field outcrop data, satellite image data, sedimentary numerical simulation data, flume experiment data, data drawn by geologists, and mature oilfield model data; The process of acquiring the field outcrop data is as follows: Select a typical outcrop profile of the lateral strata of the meandering river point-dam, and use a geological compass to mark the interface, attitude, thickness, dip angle, extension length and contact relationship of the lateral strata in the typical outcrop profile; at the same time, use a digital camera to take continuous profile photos, and generate a complete outcrop profile by image stitching to obtain two-dimensional images and geometric data of the lateral strata.

[0015] The process of acquiring the satellite image data is as follows: satellite images of the meandering river development area are selected, and image preprocessing is completed through radiometric calibration, atmospheric correction, and geometric correction. Lateral accretion layer arc-shaped striations of the satellite images are then extracted. (See details...) Figure 2 As shown.

[0016] The process of acquiring the sedimentary numerical simulation data is as follows: the formation process of meandering river point bars under different sedimentary environments is simulated using numerical simulation software. Different hydrodynamic conditions, sediment concentrations, and sedimentation rates are set in the numerical simulation. Sedimentary facies diagrams and spatial distribution data of lateral sedimentary layers at different time periods are output. Two-dimensional profiles are then extracted from the output results along the three orthogonal directions I, J, and K.

[0017] The process of obtaining the flume experimental data is as follows: Since the flume system can be used to physically simulate the deposition process of point bars in meandering rivers, multiple sets of flume deposition experiments are carried out using simulation experimental parameters of meandering river sediments (such as bottom slope, median sediment grain size, volume concentration, and feed velocity) to obtain data on the migration of point bars and the development of lateral sedimentary layers in meandering rivers. The experimental data are then sliced ​​along the three orthogonal directions I, J, and K.

[0018] The process of acquiring data drawn by geologists is as follows: acquiring plan and cross-sectional views of the lateral deposits of the meandering river point dam drawn by geologists, which can clarify the dip angle, superposition relationship, extension length and distribution range of the lateral deposits, and then scanning, digitizing and unifying the resolution of the drawn maps.

[0019] The process of acquiring mature oilfield model data is as follows: extract the geometric parameters (length, dip angle, thickness) and spatial distribution data of the lateral accumulation layers in the fine dissection model of the point dam of the Quliuhe Gudao oilfield (such as the point dam of the Gudao oilfield Ng32, Ng35-2, and Ng44-2 layers), and extract two-dimensional profiles along the three orthogonal directions I, J, and K of the extraction results.

[0020] Step S200: Preprocess the multi-source geological data to obtain a standardized raw two-dimensional dataset, including: The multi-source geological data are normalized, format unified, and lateral accretion layer geometric parameters are calibrated to obtain a standardized original two-dimensional dataset.

[0021] Specifically, in this embodiment of the invention, the above-mentioned multi-source geological data are preprocessed, including coordinate normalization, format unification, geometric parameter calibration, removal of invalid and distorted data, conversion of all data into a standardized two-dimensional data format suitable for intelligent modeling training, and calibration of the core geometric parameters (length, dip angle, longitudinal height, and distribution direction) of the lateral accumulation layer, forming a standardized original two-dimensional dataset of the lateral accumulation layer of meandering river point dam.

[0022] Step S300: Enhance the data diversity of the original two-dimensional dataset and convert the enhanced data into the tfrecords network training format to form a high-quality two-dimensional dataset that can be directly used for subsequent intelligent modeling training. The data diversity enhancement of the original two-dimensional dataset includes: Step S310: Extract multiple mud lateral accumulation layers from the original two-dimensional dataset, remove background interference information, and establish a basic database for a single mud lateral accumulation layer to provide a foundation for subsequent personalized enhancement operations; Step S320: For each of the muddy lateral deposits, the connection end between the muddy lateral deposit and the point dam lateral deposit surface is taken as the stationary end, and the length of the free end far from the river channel is corrected according to the actual geometric parameter range of the lateral deposit. The length correction includes elongation and shortening operations; The elongation operation includes: extending outward along the dip direction of the argillaceous lateral deposit, with the extension length randomly selected within the actual geometric parameters of the lateral deposit; and the extended section maintaining the same dip angle and thickness as the original argillaceous lateral deposit section. The shortening operation includes: truncation along the dip angle of the silty lateral deposit, with the shortening length randomly selected within the actual geometric parameter range of the lateral deposit; therefore, the effective part connected to the main body of the point dam is retained in the shortening operation, and the minimum shortening length is not less than the minimum thickness of a single silty lateral deposit, ensuring that the corrected length of the silty lateral deposit falls within the actual geometric parameter range of the lateral deposit, forming lateral deposit units with different spread lengths.

[0023] Step S330: Based on the actual dip angle range of the lateral accretion layers, the dip angle of each lateral accretion layer after length correction is adjusted, with the connection end between the argillaceous lateral accretion layer and the main point dam as the rotation center. Specifically, during dip angle adjustment, a small rotation is performed along the dip direction of the argillaceous lateral accretion layer, with the rotation angle strictly controlled within -15° to 15°, consistent with the measured dip angle range of the lateral accretion layers in the Gudao Oilfield. During rotation, the thickness and length of the lateral accretion layers remain unchanged; only their dip azimuth is adjusted to simulate the natural dip differences of lateral accretion layers from different phases.

[0024] Step S340: Set the stretching ratio according to the actual depositional evolution process of the lateral accretion layer, and stretch each muddy lateral accretion layer after dip angle adjustment according to the stretching ratio. Therefore, the differences in thickness and distribution scale of the lateral accretion layer in different depositional periods can be simulated, ensuring that the stretched data conforms to the depositional law of the lateral accretion layer.

[0025] The stretching includes horizontal stretching and vertical stretching; The horizontal stretching is stretching along the extension direction of the muddy lateral deposit; The vertical stretching is performed along the thickness direction of the argillaceous lateral deposit.

[0026] Step S350: Randomly combine multiple stretched muddy lateral deposits.

[0027] Specifically, based on the superposition pattern of lateral accretion layers at point dams in meandering rivers, the enhanced single muddy lateral accretion layers are randomly combined to generate two-dimensional data that conforms to the superposition characteristics of lateral accretion layers inside point dams, further enriching the pattern diversity of the dataset.

[0028] Therefore, considering the geological characteristics of the lateral deposits at point bars in meandering rivers, the above-mentioned multi-step directional data augmentation operation was designed. This operation sequentially involves single argillaceous lateral deposit extraction, argillaceous lateral deposit length correction, rotation, stretching, and random combination of single argillaceous lateral deposits. This enhances the diversity of the geological model of the lateral deposits in three pairwise orthogonal directions: the plane K direction, the direction perpendicular to the source profile I direction, and the direction along the source profile J direction. This ensures that the enhanced data conforms to the sedimentary geological patterns of the lateral deposits at point bars in meandering rivers.

[0029] Step S400: Train and construct the ProGAN-2Dto3D (Progressive Growing of GANs -2Dto3D) cross-dimensional modeling network based on the transformed training data, and use the trained ProGAN-2Dto3D cross-dimensional modeling network to generate a three-dimensional meandering river point dam lateral accumulation layer model.

[0030] The step of training and constructing the ProGAN-2Dto3D cross-dimensional modeling network based on the transformed training data includes: Step S410: The ProGAN-2Dto3D cross-dimensional modeling network includes a generator network and a discriminator network. The generator network includes a first convolutional layer, multiple upsampling and convolutional alternation layers, and a second convolutional layer connected in sequence. The discriminator network includes a third convolutional layer, multiple downsampling and convolutional alternation layers, and two fully connected layers connected in sequence.

[0031] Step S420: Using the generator network to generate a three-dimensional meandering river point dam lateral accumulation layer model based on the format-converted training data; that is, the generator network takes a low-dimensional random latent vector as input and gradually generates a high-resolution three-dimensional point dam lateral accumulation layer model through a progressive growth strategy.

[0032] Step S430: The discriminator network is used to slice the three-dimensional meandering river point dam lateral deposition model in three orthogonal directions (I, J, K), and the three-dimensional meandering river point dam lateral deposition model is scored based on the slice data. Step S440: Iteratively train the generator network and the discriminator network using a composite loss function until the loss function converges to obtain the trained ProGAN-2Dto3D cross-dimensional modeling network.

[0033] Specifically, the process of obtaining the composite loss function includes: setting corresponding weights for three orthogonal directions based on the WGAN-GP loss function. The loss function for a single direction is calculated from the distribution distance between the two-dimensional dataset and the model slice data in the corresponding direction. Weights are assigned according to the complexity of the geological patterns in the I, J, and K directions. WGAN-GP (Wasserstein GAN with Gradient Penalty) (Wasserstein distance) is introduced to ensure training stability and avoid pattern collapse.

[0034] Therefore, by utilizing the generator in the converged ProGAN-2Dto3D cross-dimensional modeling network and inputting random latent vectors, a three-dimensional meandering river point-dam lateral sedimentary layer model that conforms to geological laws can be quickly generated, accurately restoring its spatial distribution, geometric shape, and superposition characteristics. This model can be directly applied to oil and gas field development work such as numerical simulation of meandering river reservoirs and prediction of remaining oil distribution.

[0035] The present invention uses the intelligent modeling of the meandering river point-bar lateral sedimentary layer of the Ng44-2 layer in the North Second District of the Gudao Oilfield as an example to illustrate the following: 1. Determination of basic parameters for the study area See Figure 3 As shown, for the meandering river point bar reservoir of Ng44-2 layer in the Zhongerbei area of ​​Gudao Oilfield, this layer has 26 point bars with an average width of 542-1686m and an average length of 819-1706m. The longitudinal height of the lateral accretion layer is 2-9m and the dip angle is 4-10°. Based on the well network density and reservoir characteristics of the study area, the gridding parameters are set to 20m×20m in the plane and 0.2m in the vertical direction to provide a spatial framework for data processing and modeling.

[0036] 2. Multi-source data acquisition and preprocessing Multiple sources of raw data were acquired: a detailed anatomical model of the Ng44-2 layer point bar in the Gudao oilfield (a mature oilfield model), measured data of the meandering river outcrops (Yellow River Delta outcrops), high-resolution satellite images (remote sensing images of the meandering river sedimentary area), sedimentary numerical simulation results (using Delft3D to simulate the deposition of the meandering river point bar), flume experimental data (indoor physical simulation of meandering river sedimentation), and plan / section maps of the lateral sedimentary layers of the Ng44-2 layer point bar drawn by geologists, totaling six types of raw multi-source data; Data preprocessing: The above raw two-dimensional data were filtered to remove distorted and invalid data; the data coordinate system was unified to the geodetic coordinates of the study area, and the core parameters such as the length, dip angle and thickness of the lateral sedimentary layer were calibrated; all data were converted into a standardized two-dimensional image format of 64×64 pixels to form the raw two-dimensional dataset of the meandering river point dam lateral sedimentary layer of the Ng44-2 layer in the Gudao Oilfield.

[0037] 3. Enhanced directional data of lateral sedimentary layers at point dams in meandering rivers The following data augmentation operations are performed sequentially on the original two-dimensional dataset to enhance the diversity of geological models in three directions.

[0038] See details Figure 4 As shown, single clay lateral accretion layer extraction: 320 independent single clay lateral accretion layer units were extracted from the original two-dimensional dataset to establish a basic database of single clay lateral accretion layers; See details Figure 5 As shown, the length correction of the argillaceous lateral accretion layer: Based on the actual length range (50-500m) of the lateral accretion layer of layer Ng44-2, the length of a single argillaceous lateral accretion layer is corrected to generate lateral accretion layer units with a length gradient of 50m, resulting in a total of 1600 corrected single argillaceous lateral accretion layers. See details Figure 6 As shown, the rotation operation is as follows: the rotation angle range is set to -10° to 10° (matching the actual dip angle range of the lateral accretion layer), and each corrected argillaceous lateral accretion layer is randomly rotated 3 times to obtain a total of 4800 rotated argillaceous lateral accretion layer units. See details Figure 7 As shown, the stretching operation was performed by setting the stretching ratio to 0.95~1.05 and randomly stretching the rotated lateral accretion layer twice, resulting in a total of 9600 stretched muddy lateral accretion layer units. See details Figure 8 As shown, the single clay lateral accretion layer is randomly combined: according to the stacking pattern of the Ng44-2 layer point dam lateral accretion layer, the above-mentioned enhanced single clay lateral accretion layer is randomly combined to generate a total of 20,000 two-dimensional data units that conform to the internal stacking characteristics of the point dam. Format conversion: Convert the enhanced 20,000 two-dimensional data units into the tfrecords file format to generate an optimized two-dimensional dataset suitable for training progressively growing generative networks.

[0039] 4. Network model construction and training, see details below. Figure 9 As shown, Network architecture construction: A ProGAN-2Dto3D cross-dimensional modeling network is constructed. The generator adopts a fully convolutional neural network, starting from a low-dimensional latent vector of 4×4×4×8 and gradually generating a 3D model with a resolution of 64×64×64 through a progressive growth strategy. The discriminator adopts a convolutional + fully connected network to slice the generated 3D model in the I, J, and K directions to achieve real and fake model discrimination. Loss function design: The WGAN-GP loss function is used as the original loss function, and the directional label loss is introduced. The weight of the K direction (plane) is set to 1, and the weights of the I and J directions (profile) are set to 0.8 to construct a composite loss function. Network iterative training: The optimized two-dimensional dataset is input into the discriminator. The learning rate of both the generator and the discriminator is set to 0.0001. The maximum number of iterations is 60,000. The generator is trained once for each training of the discriminator, until the loss function converges.

[0040] 5. Application and Validation of Network Models 3D model generation: Input random latent vectors into the generator after training convergence, and generate 10 sets of 64×64×64 resolution 3D meandering river point dam lateral sedimentary layer models of the Ng44-2 layer of the isolated oilfield within 1 second.

[0041] The generated three-dimensional meandering river point-dam lateral accretion model was compared and verified with the well network analysis results and logging data of the study area. The model showed a consistency of over 90% in terms of lateral accretion geometric parameters and spatial distribution. The model was applied to the numerical simulation of the Ng44-2 reservoir, which effectively improved the accuracy of the prediction of remaining oil distribution and provided reliable geological model support for the optimization of oil and gas field development plans in the study area.

[0042] This invention provides a two-dimensional data generation and modeling system for lateral deposition layers of point dams in meandering rivers, comprising: The multi-source data acquisition module is used to acquire multi-source geological data of the lateral deposits of point dams in meandering rivers; The preprocessing module is communicatively connected to the multi-source data acquisition module and is used to preprocess the multi-source geological data to obtain a standardized raw two-dimensional dataset. The data augmentation module, communicatively connected to the preprocessing module, is used to enhance the data diversity of the original two-dimensional dataset and convert the augmented data into a network training format; and, The network training module, which is communicatively connected to the data augmentation module, is used to train and construct a ProGAN-2Dto3D cross-dimensional modeling network based on the transformed training data, and to generate a three-dimensional meandering river point dam lateral accumulation model using the trained ProGAN-2Dto3D cross-dimensional modeling network.

[0043] Specifically, this embodiment corresponds one-to-one with the above method embodiments. The functions of each module have been described in detail in the corresponding method embodiments, so they will not be repeated here.

[0044] In summary, the beneficial effects of this invention are as follows: through the full-process design of multi-source data acquisition, lateral accretion layer-specific data enhancement, and cross-dimensional intelligent modeling application, a two-dimensional dataset suitable for intelligent modeling of lateral accretion layers of point dams in meandering rivers is generated. This solves the problems of existing datasets having a single source, insufficient patterns, and poor cross-dimensional application effects, and significantly improves the accuracy and geological conformity of intelligent modeling of lateral accretion layers of point dams in meandering rivers.

[0045] Based on the same inventive concept, embodiments of this application also provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements all or part of the method steps of the above method.

[0046] The present invention can implement all or part of the processes in the above methods, or it can be accomplished by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0047] Based on the same inventive concept, embodiments of this application also provide an electronic device, including a memory and a processor. The memory stores a computer program that runs on the processor. When the processor executes the computer program, it implements all or part of the method steps described above.

[0048] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the computer device, connecting all parts of the computer device through various interfaces and lines.

[0049] Memory can be used to store computer programs and / or modules. The processor performs various functions of the computer device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory. Memory can primarily include a program storage area and a data storage area. The program storage area can store the operating system and application programs required for at least one function (e.g., sound playback, image playback, etc.); the data storage area can store data created based on mobile usage (e.g., audio data, video data, etc.). Furthermore, memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital cards (SD cards), flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0050] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, servers, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0051] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), servers, and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0052] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0053] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0054] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for generating and modeling two-dimensional data of lateral sedimentary layers at point dams in meandering rivers, characterized in that, include: Obtain multi-source geological data of the lateral deposits at point dams in meandering rivers; The multi-source geological data are preprocessed to obtain a standardized raw two-dimensional dataset; The original two-dimensional dataset is augmented with data diversity, and the augmented data is then converted into a network training format. The ProGAN-2Dto3D cross-dimensional modeling network is constructed based on the transformed training data, and a three-dimensional meandering river point dam lateral accumulation model is generated using the trained ProGAN-2Dto3D cross-dimensional modeling network.

2. The method for generating and modeling two-dimensional data for lateral sedimentary layers of point dams in meandering rivers as described in claim 1, characterized in that, The multi-source geological data includes field outcrop data, satellite image data, sedimentary numerical simulation data, flume experiment data, data drawn by geologists, and mature oilfield model data; The process of acquiring the field outcrop data is as follows: select a typical outcrop profile of the lateral strata of the meandering river point bar, and mark the interface, attitude, thickness, dip angle, extension length and contact relationship of the lateral strata in the typical outcrop profile; The process of acquiring the satellite image data is as follows: select satellite images of the meandering river development area and extract the lateral accretion arc-shaped texture of the satellite images; The process of acquiring the sedimentary numerical simulation data is as follows: the formation process of meandering river point bars under different sedimentary environments is simulated using numerical simulation software. Sedimentary facies diagrams and spatial distribution data of lateral sedimentary layers at different time periods are output in the numerical simulation, and two-dimensional profiles are extracted from the output results along multiple directions. The process of obtaining the flume experimental data is as follows: multiple sets of flume deposition experiments are carried out using simulation experimental parameters of meandering river deposition to obtain data on the migration of point dams and the development of lateral sedimentary layers in meandering rivers, and two-dimensional profiles are extracted from the experimental data along multiple directions. The process of acquiring the data drawn by the geologists is as follows: acquiring the plan view and cross-sectional view of the lateral deposits of the meandering river point dam drawn by the geologists; The process of acquiring the mature oilfield model data is as follows: extract the geometric parameters and spatial distribution data of the lateral accumulation layer in the fine dissection model of the point dam of the Quliuhe Gudao oilfield, and extract two-dimensional profiles of the extraction results along multiple directions.

3. The method for generating and modeling two-dimensional data for lateral sedimentary layers of point dams in meandering rivers as described in claim 1, characterized in that, The preprocessing of the multi-source geological data to obtain a standardized raw two-dimensional dataset includes: The multi-source geological data are normalized, format unified, and lateral accretion layer geometric parameters are calibrated to obtain a standardized original two-dimensional dataset.

4. The method for generating and modeling two-dimensional data for lateral deposition layers of point dams in meandering rivers as described in claim 1, characterized in that, The data diversity enhancement of the original two-dimensional dataset includes: Multiple mud lateral accumulation layers were extracted from the original two-dimensional dataset; For each of the aforementioned muddy lateral deposits, the connection end between the muddy lateral deposit and the point dam lateral deposit surface is taken as the stationary end, and the length of the free end far from the river channel is corrected according to the actual geometric parameter range of the lateral deposit. Based on the actual dip angle range of the lateral deposits, the dip angle of each of the silty lateral deposits after length correction is adjusted with the connection end between the silty lateral deposits and the main body of the point dam as the rotation center; The stretching ratio is set according to the actual sedimentary evolution process of the lateral accretion layer, and each of the muddy lateral accretion layers after dip angle adjustment is stretched according to the stretching ratio. Randomly combined multiple stretched muddy lateral deposits.

5. The method for generating and modeling two-dimensional data for lateral sedimentary layers of point dams in meandering rivers as described in claim 4, characterized in that, The length correction includes elongation and shortening operations; The elongation operation includes: extending outward along the dip angle of the argillaceous lateral accretion layer, with the extension length randomly selected within the range of the actual geometric parameters of the lateral accretion layer; The shortening operation includes: cutting along the dip angle extension direction of the argillaceous lateral accretion layer, with the shortening length randomly selected within the actual geometric parameter range of the lateral accretion layer; The stretching includes horizontal stretching and vertical stretching; The horizontal stretching is stretching along the extension direction of the muddy lateral deposit layer; The vertical stretching is performed along the thickness direction of the argillaceous lateral deposit.

6. The method for generating and modeling two-dimensional data for lateral sedimentary layers of point dams in meandering rivers as described in claim 1, characterized in that, The step of training and constructing the ProGAN-2Dto3D cross-dimensional modeling network based on the transformed training data includes: The ProGAN-2Dto3D cross-dimensional modeling network includes a generator network and a discriminator network; The generator network is used to generate a three-dimensional meandering river point dam lateral accumulation layer model based on the format-converted training data. The discriminator network is used to slice the three-dimensional meandering river point dam lateral deposition model in three orthogonal directions, and the three-dimensional meandering river point dam lateral deposition model is scored based on the slice data; The generator network and the discriminator network are iteratively trained using a composite loss function until the loss function converges, resulting in the trained ProGAN-2Dto3D cross-dimensional modeling network.

7. The method for generating and modeling two-dimensional data for lateral deposition layers of point dams in meandering rivers as described in claim 6, characterized in that, The generator network includes a first convolutional layer, multiple alternating upsampling and convolutional layers, and a second convolutional layer connected in sequence. The discriminator network includes a third convolutional layer connected in sequence, multiple downsampling and convolutional alternating layers, and two fully connected layers; The process of obtaining the composite loss function includes: setting corresponding weights for three orthogonal directions based on the WGAN-GP loss function.

8. A two-dimensional data generation and modeling system for lateral deposition layers of point dams in meandering rivers, characterized in that, include: The multi-source data acquisition module is used to acquire multi-source geological data of the lateral deposits of point dams in meandering rivers; The preprocessing module is communicatively connected to the multi-source data acquisition module and is used to preprocess the multi-source geological data to obtain a standardized raw two-dimensional dataset. The data augmentation module is communicatively connected to the preprocessing module and is used to enhance the data diversity of the original two-dimensional dataset and convert the augmented data into a network training format. as well as, The network training module, which is communicatively connected to the data augmentation module, is used to train and construct a ProGAN-2Dto3D cross-dimensional modeling network based on the transformed training data, and to generate a three-dimensional meandering river point dam lateral accumulation model using the trained ProGAN-2Dto3D cross-dimensional modeling network.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the two-dimensional data generation and modeling method for lateral sedimentary layers of point dams in meandering rivers as described in any one of claims 1 to 7.

10. An electronic device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that, When the processor runs the computer program, it implements the two-dimensional data generation and modeling method for lateral sedimentary layers of point dams in meandering rivers as described in any one of claims 1 to 7.

Images