3D geological modeling methods, systems, equipment and storage media

Abstract

This application discloses a three-dimensional geological modeling method, system, device, and storage medium. The three-dimensional geological modeling method includes, given the construction of a borehole dataset of the area to be modeled, determining the spatial coordinates of each borehole in the dataset and constructing a lithology database; constructing a lithology sample library based on the lithology database and a three-dimensional spatial cubic mesh model; iteratively training an initial graph attention network model based on the first lithology sample library to obtain a trained graph attention network model; inputting a second lithology sample library into the trained graph attention network model to output the predicted lithology category of each second lithology sample; and constructing a three-dimensional geological model of the area to be modeled based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, thereby improving the accuracy of the three-dimensional geological model construction.

Description

Technical Field

[0001] This application relates to the technical field of three-dimensional geological modeling, and in particular to three-dimensional geological modeling methods, systems, equipment and storage media. Background Technology

[0002] Three-dimensional geological modeling, as a core technology in the field of digital geology, serves as a bridge connecting geological theory and engineering practice, playing an irreplaceable role in mineral resource exploration, geological disaster early warning, and underground space development. Through 3D geological modeling, complex underground strata, faults, and ore bodies can be visually displayed, and exploration decisions can be aided in optimizing resource development plans and reducing exploration risks, thereby achieving the scientific management and sustainable utilization of underground resources. Precisely because of its significant scientific value and engineering implications, improving the accuracy and reliability of 3D geological models has become a critical issue urgently needing to be addressed in the field of Earth sciences.

[0003] Currently, three-dimensional geological modeling methods can be divided into two major systems: explicit modeling and implicit modeling. Regardless of whether it is explicit or implicit modeling, the reliability of the model is essentially limited by the quality and completeness of the borehole data. However, due to the limitations of exploration costs and terrain conditions, the spatial distribution of borehole data is often extremely uneven, which significantly increases the uncertainty of the model and thus affects the accuracy of the three-dimensional geological model. Summary of the Invention

[0004] This application aims to at least address the technical problems existing in the prior art. To this end, this application proposes a three-dimensional geological modeling method, system, device, and storage medium, which can improve the accuracy of three-dimensional geological model construction.

[0005] A first aspect of this application provides a three-dimensional geological modeling method, comprising the following steps:

[0006] In constructing a borehole dataset for the region to be modeled, the spatial coordinates of each borehole data in the borehole dataset are determined, wherein the borehole dataset includes the lithology category of each borehole data.

[0007] A lithology database is constructed based on the spatial coordinates and lithology categories of all the borehole data.

[0008] In constructing a three-dimensional spatial cube mesh model of the region to be modeled, a lithological sample library is constructed based on the lithological database and the three-dimensional spatial cube mesh model. The lithological sample library includes a first lithological sample library and a second lithological sample library. The first lithological sample library includes the lithological category of each first lithological sample, while the second lithological sample library does not include the lithological category of any second lithological sample.

[0009] An initial graph attention network model is constructed, and the initial graph attention network model is iteratively trained based on the first lithology sample library to obtain a trained graph attention network model.

[0010] The second lithological sample library is input into the trained graph attention network model to output the predicted lithological category of each second lithological sample through the trained graph attention network model;

[0011] Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed.

[0012] The three-dimensional geological modeling method according to the embodiments of this application has at least the following beneficial effects:

[0013] This application, by constructing a borehole dataset for the region to be modeled, determines the spatial coordinates of each borehole in the dataset. Based on the spatial coordinates and lithology categories of all borehole data, a lithology database is constructed. Then, by constructing a three-dimensional cubic mesh model of the region to be modeled, a lithology sample library is built based on the lithology database and the three-dimensional cubic mesh model. An initial graph attention network (GNN) model is constructed, and iterative training is performed on the initial GNN model based on the first lithology sample library to obtain a trained GNN model. A second lithology sample library is input into the trained GNN model to output the predicted lithology category for each second lithology sample. Based on the first lithology sample library... This application constructs a three-dimensional geological model of the area to be modeled by integrating all borehole data, constructing a lithology sample library through gridding, and then outputting the predicted lithology category of each second lithology sample through a trained graph attention network model. This solves the problem of uneven spatial distribution of borehole data and improves the accuracy of model prediction. Finally, based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed, which improves the accuracy of three-dimensional geological model construction.

[0014] According to some embodiments of this application, constructing the borehole dataset for the region to be modeled includes:

[0015] Given all the borehole files for the region to be modeled, normalize the format of all the borehole files to obtain a normalized borehole text dataset.

[0016] When the normalized borehole text dataset is divided into a first normalized borehole text dataset and a second normalized borehole text dataset according to a preset ratio, each first normalized borehole text data in the first normalized borehole text dataset is labeled to obtain a labeled borehole text dataset.

[0017] An initial semantic recognition model is constructed. Based on the labeled post-drilling text dataset, the initial semantic recognition model is trained to obtain a trained semantic recognition model. The first normalized post-drilling text dataset and the second normalized post-drilling text dataset are input into the trained semantic recognition model to obtain the recognition result of each first normalized post-drilling text data and the recognition result of each second normalized post-drilling text data output by the trained semantic recognition model.

[0018] The borehole dataset is constructed based on the recognition results of all the first normalized borehole text data and the recognition results of all the second normalized borehole text data.

[0019] According to some embodiments of this application, the borehole dataset further includes the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, depth at the beginning of the lithological interval, and depth at the end of the lithological interval for each borehole data. The spatial coordinates include the spatial coordinates at the beginning and the spatial coordinates at the end. Determining the spatial coordinates of each borehole data in the borehole dataset includes:

[0020] Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and the depth at the beginning of the lithology interval, the spatial coordinates at the beginning of each borehole data are determined, wherein the spatial coordinates at the beginning include the east coordinates, the north coordinates, and the elevation at the beginning;

[0021] Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and depth at the end of the lithological interval, the spatial coordinates at the end of each borehole data are determined, wherein the spatial coordinates at the end include the east coordinate of the end, the north coordinate of the end, and the elevation of the end.

[0022] According to some embodiments of this application, constructing a lithology database based on the spatial coordinates and lithology categories of all the borehole data includes:

[0023] When a line segment is obtained by connecting the spatial coordinates of the starting point and the spatial coordinates of the ending point, the line segment is linearly interpolated according to a preset discontinuity value using a linear interpolation method to obtain linear interpolation point data for each borehole data. The linear interpolation point data includes the east coordinate, north coordinate, elevation, and lithology category of each linear interpolation point, and each linear interpolation point has the same lithology category as its corresponding borehole data.

[0024] The lithology database is constructed based on the spatial coordinates, lithology category, and linear interpolation point data of each borehole.

[0025] According to some embodiments of this application, in the case of constructing a three-dimensional spatial cube mesh model of the region to be modeled, constructing a lithological sample library based on the lithological database and the three-dimensional spatial cube mesh model includes:

[0026] Determine the minimum inter-point distance among all points in the lithology database;

[0027] Using the minimum distance between points as the side length of the cube mesh, a three-dimensional spatial cube mesh model of the region to be modeled is constructed.

[0028] Based on the east coordinates, north coordinates, and elevation of each lithological data in the lithological database, the lithological database is written into the three-dimensional spatial cubic mesh model to obtain the lithological cubic mesh model. The cubic mesh in the lithological cubic mesh model is used as the first cubic mesh, wherein the lithological category of the first cubic mesh is the same as the lithological category of the lithological data falling within the first cubic mesh, and the side length of the first cubic mesh is not included in the first cubic mesh.

[0029] Extract data for each first cube grid, wherein the data for each first cube grid includes the position coordinates and lithology category label of each first cube grid, and the lithology category label is either a lithology category or a null value;

[0030] The data of the first cube grid, which labels all the lithological categories as lithological categories, is used as the first lithological sample library; the data of the first cube grid, which labels all the lithological categories as null values, is used as the second lithological sample library.

[0031] According to some embodiments of this application, the step of iteratively training the initial graph attention network model based on the first lithological sample library to obtain a trained graph attention network model includes:

[0032] Based on the first lithological sample library, the adjacency matrix is ​​determined using the K-nearest neighbor algorithm;

[0033] Based on the first lithological sample library and the adjacency matrix, the initial graph attention network model is iteratively trained until a preset maximum number of iterations is reached, thereby obtaining the trained graph attention network model.

[0034] According to some embodiments of this application, constructing a three-dimensional geological model of the area to be modeled based on the first lithological sample library, the second lithological sample library, and the predicted lithology category of each second lithological sample includes:

[0035] Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed using the moving cube algorithm.

[0036] A second aspect of this application provides a three-dimensional geological modeling system, the three-dimensional geological modeling system comprising:

[0037] The spatial coordinate determination module is used to determine the spatial coordinates of each borehole data in the borehole dataset when constructing a borehole dataset of the area to be modeled, wherein the borehole dataset includes the lithology category of each borehole data;

[0038] A lithology database construction module is used to construct a lithology database based on the spatial coordinates and lithology categories of all the borehole data;

[0039] The lithological sample library construction module is used to construct a lithological sample library based on the lithological database and the three-dimensional spatial cube mesh model of the area to be modeled, wherein the lithological sample library includes a first lithological sample library and a second lithological sample library, the first lithological sample library includes the lithological category of each first lithological sample, and the second lithological sample library does not include the lithological category of any second lithological sample;

[0040] The model training module is used to construct an initial graph attention network model, and iteratively train the initial graph attention network model based on the first lithology sample library to obtain a trained graph attention network model.

[0041] The predicted lithology category output module is used to input the second lithology sample library into the trained graph attention network model, so as to output the predicted lithology category of each second lithology sample through the trained graph attention network model;

[0042] The three-dimensional geological model construction module is used to construct a three-dimensional geological model of the area to be modeled based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample.

[0043] This system, by constructing a borehole dataset for the region to be modeled, determines the spatial coordinates of each borehole in the dataset. Based on the spatial coordinates and lithology categories of all borehole data, a lithology database is built. Then, by constructing a 3D cubic mesh model of the region, a lithology sample library is built based on the lithology database and the 3D cubic mesh model. An initial graph attention network (GNN) model is constructed. Based on the first lithology sample library, the initial GNN model is iteratively trained to obtain a trained GNN model. A second lithology sample library is input into the trained GNN model to output the predicted lithology category for each second lithology sample. Based on the first lithology sample library... This application constructs a three-dimensional geological model of the area to be modeled by integrating all borehole data, constructing a lithology sample library through gridding, and then outputting the predicted lithology category of each second lithology sample through a trained graph attention network model. This solves the problem of uneven spatial distribution of borehole data and improves the accuracy of model prediction. Finally, based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed, which improves the accuracy of three-dimensional geological model construction.

[0044] A third aspect of this application provides a three-dimensional geological modeling electronic device, including at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which are executed by the at least one control processor to enable the at least one control processor to perform the above-described three-dimensional geological modeling method.

[0045] In a fourth aspect, this application provides a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the aforementioned three-dimensional geological modeling method.

[0046] It should be noted that the beneficial effects of the second to fourth aspects of this application with respect to the prior art are the same as the beneficial effects of the aforementioned three-dimensional geological modeling system with respect to the prior art, and will not be described in detail here.

[0047] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0048] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0049] Figure 1This is a flowchart illustrating an embodiment of the three-dimensional geological modeling method provided in this application;

[0050] Figure 2 This is a schematic diagram of the structure of an embodiment of the three-dimensional geological modeling system provided in this application;

[0051] Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application. Detailed Implementation

[0052] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0053] In the description of this application, the use of terms such as "first," "second," etc., is for the purpose of distinguishing technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or the order of the technical features indicated.

[0054] In the description of this application, it should be understood that the orientation descriptions, such as up, down, etc., are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0055] In the description of this application, it should be noted that, unless otherwise explicitly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.

[0056] Three-dimensional geological modeling, as a core technology in the field of digital geology, serves as a bridge connecting geological theory and engineering practice, playing an irreplaceable role in mineral resource exploration, geological disaster early warning, and underground space development. Through 3D geological modeling, complex underground strata, faults, and ore bodies can be visually displayed, and exploration decisions can be aided in optimizing resource development plans and reducing exploration risks, thereby achieving the scientific management and sustainable utilization of underground resources. Precisely because of its significant scientific value and engineering implications, improving the accuracy and reliability of 3D geological models has become a critical issue urgently needing to be addressed in the field of Earth sciences.

[0057] Currently, three-dimensional geological modeling methods can be divided into two major systems: explicit modeling and implicit modeling. Regardless of whether it is explicit or implicit modeling, the reliability of the model is essentially limited by the quality and completeness of the borehole data. However, due to the limitations of exploration costs and terrain conditions, the spatial distribution of borehole data is often extremely uneven, which significantly increases the uncertainty of the model and thus affects the accuracy of the three-dimensional geological model.

[0058] To address the aforementioned technical deficiencies, embodiments of this application provide a three-dimensional geological modeling method, system, device, and storage medium.

[0059] Please see Figure 1 This is a flowchart illustrating a three-dimensional geological modeling method provided in an embodiment of this application. This method is applied to an electronic device, such as a server. Figure 1 As shown, this three-dimensional geological modeling method includes:

[0060] Step S101: In the case of constructing a borehole dataset of the area to be modeled, determine the spatial coordinates of each borehole data in the borehole dataset, wherein the borehole dataset includes the lithology category of each borehole data.

[0061] The aforementioned borehole dataset may also include the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, depth at the beginning of the lithological interval, and depth at the end of the lithological interval for each borehole data.

[0062] The aforementioned spatial coordinates may include the starting spatial coordinates and the ending spatial coordinates.

[0063] The aforementioned lithological categories can be quartz diorite, limestone, or skarn.

[0064] Step S102: Construct a lithology database based on the spatial coordinates and lithology categories of all borehole data;

[0065] Step S103: In the case of constructing a three-dimensional spatial cube mesh model of the area to be modeled, construct a lithological sample library based on the lithological database and the three-dimensional spatial cube mesh model. The lithological sample library includes a first lithological sample library and a second lithological sample library. The first lithological sample library includes the lithological category of each first lithological sample, while the second lithological sample library does not include the lithological category of any second lithological sample.

[0066] Step S104: Construct an initial graph attention network model. Based on the first lithology sample library, iteratively train the initial graph attention network model to obtain a trained graph attention network model.

[0067] Step S105: Input the second lithology sample library into the trained graph attention network model, so as to output the predicted lithology category of each second lithology sample through the trained graph attention network model;

[0068] Step S106: Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, construct a three-dimensional geological model of the area to be modeled.

[0069] This application, by constructing a borehole dataset for the region to be modeled, determines the spatial coordinates of each borehole in the dataset. Based on the spatial coordinates and lithology categories of all borehole data, a lithology database is constructed. Then, by constructing a three-dimensional cubic mesh model of the region to be modeled, a lithology sample library is built based on the lithology database and the three-dimensional cubic mesh model. An initial graph attention network (GNN) model is constructed, and iterative training is performed on the initial GNN model based on the first lithology sample library to obtain a trained GNN model. A second lithology sample library is input into the trained GNN model to output the predicted lithology category for each second lithology sample. Based on the first lithology sample library... This application constructs a three-dimensional geological model of the area to be modeled by integrating all borehole data, constructing a lithology sample library through gridding, and then outputting the predicted lithology category of each second lithology sample through a trained graph attention network model. This solves the problem of uneven spatial distribution of borehole data and improves the accuracy of model prediction. Finally, based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed, which improves the accuracy of three-dimensional geological model construction.

[0070] In some embodiments, step S101 may include steps S201 to S204:

[0071] Step S201: After obtaining all the borehole files of the area to be modeled, normalize the format of all the borehole files to obtain the normalized borehole text dataset.

[0072] All the above-mentioned drilling files can include drilling files in Excel format, PDF format, and text format.

[0073] In step S201, after obtaining all the borehole files of the area to be modeled, the format of all the borehole files is normalized to obtain the normalized borehole text dataset. This normalized dataset can be: data from all the Excel format borehole files of the area to be modeled as the first borehole text data; data from all the PDF format borehole files as the second borehole text data; and data from all the text format borehole files as the third borehole text data. The first, second, and third borehole text data are then normalized to obtain the normalized borehole text dataset.

[0074] Step S202: After dividing the normalized borehole text dataset into a first normalized borehole text dataset and a second normalized borehole text dataset according to a preset ratio, each first normalized borehole text data in the first normalized borehole text dataset is labeled to obtain the labeled borehole text dataset.

[0075] The above preset ratio can be a value set in advance according to actual needs.

[0076] In step S202, after dividing the normalized borehole text dataset into a first normalized borehole text dataset and a second normalized borehole text dataset according to a preset ratio, each piece of text in the first normalized borehole text dataset is labeled to obtain a labeled borehole text dataset. This labeled dataset can be obtained by randomly selecting 20% ​​(preset ratio) of the normalized borehole text dataset as the first normalized borehole text dataset and the remaining 80% as the second normalized borehole text dataset. The LabelStudio tool is used to label the characters in the first normalized borehole text dataset to obtain the labeled borehole text dataset. The labeling can be: B = start of key information, I = inside key information, O = non-key information. For example, for text such as "Borehole number: ZK001." in the first normalized borehole text dataset, the labeling can be as shown in Table 1.

[0077] Table 1

[0078] character Drill hole Editor Number ： Z K 0 0 1 . Label O O O O O B I I I I O

[0079] Step S203: Construct an initial semantic recognition model. Based on the labeled borehole text dataset, train the initial semantic recognition model to obtain a trained semantic recognition model. Input the second normalized borehole text dataset into the trained semantic recognition model to obtain the recognition result of each second normalized borehole text data output by the trained semantic recognition model.

[0080] The initial semantic recognition model mentioned above can be a deep learning semantic recognition model based on BERT-base.

[0081] In step S203, the initial semantic recognition model is constructed based on the labeled borehole text dataset. The initial semantic recognition model is then trained to obtain a trained semantic recognition model. The first and second normalized borehole text datasets are input into the trained semantic recognition model to obtain the recognition results for each first and second normalized borehole text data point. These results may include:

[0082] A domain dictionary is constructed based on human experience. The domain dictionary may include borehole number (which may include a prefix, such as ZK or ZKS), coordinate terms (which may include E / longitude, N / latitude and Z / elevation), skewing terms (which may include skewing depth, dip angle or azimuth), and lithology category (which may include quartz diorite, limestone or skarn).

[0083] A label mapping dictionary is constructed to convert the above BIO labels (BIO labels are the annotations of the labeled borehole text dataset) into numeric IDs. The label mapping dictionary can be: "O":0, "B-Borehole number":1, "I-Borehole number":2, "B-Borehole opening E":3, "I-Borehole opening E":4, "B-Borehole opening N":5, "I-Borehole opening N":6, "B-Borehole opening Z":7, "I-Borehole opening Z":8, "B-Inclination depth":9, "I-Inclination depth":10, "B-Dip angle":11, "I-Dip angle":12, "B-Azimuth":13, "I-Azimuth":14, "B-Lithology starting depth":15, "I-Lithology starting depth":16, "B-Lithology ending depth":17, "I-Lithology ending depth":18, "B-Rock name":19, and "I-Rock name":20.

[0084] The labeled borehole text dataset is transformed according to the label mapping dictionary to obtain the transformed borehole text dataset.

[0085] An initial semantic recognition model is constructed by masking the second normalized drill text dataset according to a preset mask ratio set according to actual needs, resulting in a masked normalized drill text dataset. This masked normalized drill text dataset is then input into the initial semantic recognition model, enabling the initial semantic recognition model to undergo self-supervised task training in semantic recognition, thereby obtaining a preprocessed semantic recognition model with text recognition capabilities. The underlying data of the initial semantic recognition model includes the aforementioned domain dictionary.

[0086] The converted borehole text dataset is input into the preprocessed semantic recognition model for training until the maximum number of iterations is preset according to actual needs, and the trained semantic recognition model is obtained.

[0087] Input the first normalized borehole text dataset and the second normalized borehole text dataset into the trained semantic recognition model to obtain the recognition results of each first normalized borehole text dataset and each second normalized borehole text dataset output by the trained semantic recognition model. The above recognition results can be semantic recognition results generated according to the label mapping dictionary.

[0088] Step S204: Construct a borehole dataset based on the recognition results of all first normalized borehole text data and all second normalized borehole text data.

[0089] The identification results may include borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, depth at the beginning of the lithological interval, depth at the end of the lithological interval, and lithological category.

[0090] In step S204, the above-mentioned borehole dataset is constructed by filling the identification results of all first normalized borehole text data and all second normalized borehole text data into a table with the data content of "borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, depth at the beginning of the lithology interval, depth at the end of the lithology interval, and lithology category".

[0091] This application introduces a semantic recognition model, combined with a domain dictionary and tag system, to achieve automatic extraction of key information from multi-format borehole files. This solves the problem of difficult batch preprocessing of multi-source data, providing more accurate data for subsequent modeling and thus improving the accuracy of 3D geological modeling.

[0092] In some embodiments, step S101 may include steps S301 to S302:

[0093] Step S301: Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and the depth at the beginning of the lithology interval, determine the spatial coordinates at the beginning of each borehole data point. The spatial coordinates at the beginning of ...

[0094] In step S301, the determination of the starting spatial coordinates of each borehole data point based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and the starting depth of the lithological interval can be achieved by sorting all inclination points from the borehole coordinates to the starting depth of the lithological interval in ascending order (i.e., arranged from smallest to largest distance) according to their distance from the borehole coordinates. Then, based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and the starting depth of the lithological interval, the starting spatial coordinates of each borehole data point are calculated using the following formulas 1 to 6:

[0095]

[0096] in, Let be the east coordinate of the starting point of the i-th borehole data. Let be the north coordinate of the starting point of the i-th borehole data. Let be the starting elevation of the i-th borehole data. The depth at the beginning of the lithological interval for the i-th borehole data. This represents the depth difference at the starting point. This represents the total number of all inclinometer points between the borehole coordinates and the depth at the beginning of the lithological interval for the pre-acquired data of the i-th borehole. Let be the east coordinate of the borehole opening coordinate of the i-th borehole data. Let be the north coordinate of the borehole opening of the i-th borehole data. Let be the elevation of the borehole coordinates for the i-th borehole data. Let be the inclination depth of the k-th inclination point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. Let be the apex angle of the inclination measurement point at the k-th inclination measurement point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. Let be the inclination azimuth of the k-th inclination point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. This is the value of the first intermediate parameter. Let m be the inclination depth of the m-th survey point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. Let be the apex angle of the m-th survey point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. Let be the inclination azimuth of the m-th inclination point between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. The inclination depth of the m-th inclination point is added between the borehole coordinates and the depth at the beginning of the lithological interval for the i-th borehole data. For the i-th borehole data, add the inclination apex angle of the m-th inclination point between the borehole coordinates and the depth at the beginning of the lithological interval. The inclination azimuth of the m-th inclination point is added to the depth between the borehole coordinates and the starting point of the lithological interval for the i-th borehole data.

[0097] Step S302: Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and depth at the end of the lithology interval, determine the spatial coordinates at the end of each borehole data point. The spatial coordinates at the end of ...

[0098] In step S302, the determination of the spatial coordinates at the end of each borehole data point based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and lithological interval termination depth can be achieved by sorting all inclination points from the borehole coordinates to the lithological interval termination depth in ascending order (i.e., arranged from smallest to largest distance) according to their distance from the borehole coordinates. Then, based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and lithological interval termination depth, the spatial coordinates at the end of each borehole data point are calculated using the following formulas 7 to 12:

[0099]

[0100] in, Let be the east coordinate of the termination point of the i-th borehole data. Let be the north coordinate of the end point of the i-th borehole data. Let be the elevation at the end of the i-th borehole data. The depth at the end of the lithological interval for the i-th borehole data. The depth difference at the termination point. This represents the total number of all inclinometer points between the borehole coordinates and the depth at the end of the lithological interval for the pre-acquired i-th borehole data. Let be the inclination depth of the b-th inclination point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. Let be the apex angle of the b-th survey point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. Let be the inclination azimuth of the b-th inclination point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. This is the value of the second intermediate parameter. Let be the inclination depth of the nth inclination point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. Let be the apex angle of the inclination measurement point at the nth inclination measurement point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. Let be the inclination azimuth of the nth inclination point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data. The inclination depth of the nth inclination point between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data is calculated. For the i-th borehole data, add the inclination apex angle of the n-th inclination point between the borehole coordinates and the depth at the end of the lithological interval. The inclination azimuth of the nth inclination point is added between the borehole coordinates and the depth at the end of the lithological interval for the i-th borehole data.

[0101] This application improves the accuracy of 3D geological modeling by calculating the spatial coordinates of each borehole data point, thus providing more accurate data for subsequent modeling.

[0102] In some embodiments, step S102 may include steps S401 to S402:

[0103] Step S401: After connecting the spatial coordinates of the starting point and the spatial coordinates of the ending point to obtain a line segment, the line segment is linearly interpolated according to the preset discontinuity value using a linear interpolation method to obtain the linear interpolation point data for each borehole data. The linear interpolation point data includes the east coordinate, north coordinate, elevation and lithology category of each linear interpolation point. Each linear interpolation point has the same lithology category as its corresponding borehole data.

[0104] The aforementioned preset discontinuity values ​​can be pre-set according to actual needs.

[0105] Step S402: Construct a lithology database based on the spatial coordinates, lithology category, and linear interpolation point data of each borehole.

[0106] In step S402, the above-mentioned construction of the lithology database based on the spatial coordinates, lithology category, and linear interpolation point data of each borehole data can be used to construct an initial lithology database. This involves writing the spatial coordinates, lithology category, and linear interpolation point data of each borehole data into the initial lithology database to obtain the lithology database.

[0107] This application, having obtained the linear interpolation point data for each borehole, constructs a lithology database based on the spatial coordinates, lithology category, and linear interpolation point data of each borehole. This database can complete the borehole spatial trajectory data, providing a more accurate data basis for subsequent modeling and thus improving the accuracy of 3D geological modeling.

[0108] In some embodiments, step S103 may include steps S501 to S505:

[0109] Step S501: Determine the minimum inter-point distance among all points in the lithology database;

[0110] In step S501, the determination of the minimum inter-point distance between all points in the lithology database can be achieved by calculating the distance between any two points in the lithology database and selecting the smallest distance value from all the calculated distance values ​​as the minimum inter-point distance.

[0111] Step S502: Construct a three-dimensional spatial cube mesh model of the region to be modeled, using the minimum distance between points as the side length of the cube mesh;

[0112] Step S503: Based on the east coordinates, north coordinates, and elevation of each lithological data in the lithological database, write the lithological database into a three-dimensional spatial cubic mesh model to obtain a lithological cubic mesh model. The cubic mesh in the lithological cubic mesh model is taken as the first cubic mesh. The lithological category of the first cubic mesh is the same as the lithological category of the lithological data falling within the first cubic mesh. The side length of the first cubic mesh is not included in the first cubic mesh.

[0113] In step S503, based on the east coordinates, north coordinates, and elevation of each lithological data in the lithological database, the lithological database is written into a three-dimensional spatial cube grid model. The resulting lithological cube grid model can be obtained by writing each lithological data into the corresponding position in the three-dimensional spatial cube grid model according to the east coordinates, north coordinates, and elevation. The lithological category of the first cube grid in the lithological cube grid model is labeled as the lithological category of the lithological data falling within the first cube grid.

[0114] Specifically, if no lithological data is written into the first cube grid in the lithological cube gridding model, the lithological category of the first cube grid is marked as null. It should be noted that if lithological data is written only on the side length of the first cube grid, it is also considered that no lithological data is written into the first cube grid.

[0115] Step S504: Extract the data for each first cube grid, wherein the data for each first cube grid includes the position coordinates and lithology category label of each first cube grid, and the lithology category label is either the lithology category or a null value;

[0116] Step S505: Use the data of the first cube grid that labels all lithological categories as lithological categories as the first lithological sample library; use the data of the first cube grid that labels all lithological categories as null values ​​as the second lithological sample library.

[0117] This application achieves the transformation of lithological data from discrete single-point storage to three-dimensional regular gridded representation by writing the lithological database into a three-dimensional spatial cubic gridded model. This allows for precise binding of lithological spatial distribution with three-dimensional grid cells. At the same time, the data of cubic grids with all lithological categories labeled as lithological categories are used as the first lithological sample library; the data of cubic grids with all lithological categories labeled as null values ​​are used as the second lithological sample library. This provides a more accurate data basis for subsequent prediction of lithological categories through the model, thereby improving the accuracy of three-dimensional geological modeling.

[0118] In some embodiments, step S104 may include steps S601 to S602:

[0119] Step S601: Based on the first lithological sample library, determine the adjacency matrix using the K-nearest neighbor algorithm;

[0120] In step S601, the above-mentioned determination of the adjacency matrix based on the first lithological sample library using the K-nearest neighbor algorithm may include:

[0121] The lithology category of each first lithology sample in the first lithology sample library is encoded using the one-hot encoding method to obtain the encoded lithology sample library;

[0122] Based on the spatial coordinates of each coded lithological sample and the preset K value set according to actual needs, an adjacency matrix is ​​constructed using the K-nearest neighbor algorithm, where K can be 20.

[0123] Step S602: Based on the first lithology sample library and the adjacency matrix, iteratively train the initial graph attention network model until the preset maximum number of iterations is reached, and obtain the trained graph attention network model.

[0124] The above-mentioned maximum number of iterations can be a value preset according to actual needs.

[0125] In step S602, the initial graph attention network model is iteratively trained based on the first lithology sample library and the adjacency matrix until a preset maximum number of iterations is reached, resulting in a trained graph attention network model. This can be achieved by encoding the lithology category of each first lithology sample in the first lithology sample library using a one-hot encoding method, obtaining an encoded lithology sample library, and then inputting the encoded lithology sample library and the adjacency matrix into the initial graph attention network model for iterative training until a preset maximum number of iterations is reached, thus obtaining a trained graph attention network model.

[0126] This application uses a first lithological sample library and an adjacency matrix to iteratively train an initial graph attention network model until a preset maximum number of iterations is reached, resulting in a well-trained graph attention network model. Relying on the adjacency matrix, the model accurately characterizes the association between lithological samples, allowing it to capture the spatial association features of lithology. This overcomes the limitations of traditional independent sample training and improves the accuracy of model training.

[0127] In some embodiments, step S106 may include step S701:

[0128] Step S701: Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, construct a three-dimensional geological model of the area to be modeled using the moving cube algorithm.

[0129] In step S701, the above-mentioned construction of a three-dimensional geological model of the area to be modeled based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample using the moving cube algorithm can be achieved by extracting the isosurface of each lithology category according to the lithology categories in the first lithology sample library and the predicted lithology category of each second lithology sample using the moving cube algorithm, and constructing a three-dimensional geological model of the area to be modeled based on the isosurface of each lithology category.

[0130] This application achieves accurate reconstruction of isosurfaces through the moving cube algorithm, thereby improving the accuracy of three-dimensional geological models.

[0131] Additionally, refer to Figure 2 One embodiment of this application provides a three-dimensional geological modeling system, including a spatial coordinate determination module 1100, a lithology database construction module 1200, a lithology sample library construction module 1300, a model training module 1400, a predicted lithology category output module 1500, and a three-dimensional geological model construction module 1600, wherein:

[0132] The spatial coordinate determination module 1100 is used to determine the spatial coordinates of each borehole data in the borehole dataset when constructing a borehole dataset of the area to be modeled, wherein the borehole dataset includes the lithology category of each borehole data.

[0133] The lithology database construction module 1200 is used to construct a lithology database based on the spatial coordinates and lithology categories of all borehole data;

[0134] The lithological sample library construction module 1300 is used to construct a lithological sample library based on the lithological database and the three-dimensional spatial cubic mesh model when constructing a three-dimensional spatial cubic mesh model of the area to be modeled. The lithological sample library includes a first lithological sample library and a second lithological sample library. The first lithological sample library includes the lithological category of each first lithological sample, while the second lithological sample library does not include the lithological category of any second lithological sample.

[0135] The model training module 1400 is used to construct an initial graph attention network model. Based on the first lithology sample library, the initial graph attention network model is iteratively trained to obtain a trained graph attention network model.

[0136] The lithology category prediction output module 1500 is used to input the second lithology sample library into the trained graph attention network model, so as to output the predicted lithology category of each second lithology sample through the trained graph attention network model;

[0137] The 3D geological model building module 1600 is used to build a 3D geological model of the area to be modeled based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample.

[0138] This system, by constructing a borehole dataset for the region to be modeled, determines the spatial coordinates of each borehole in the dataset. Based on the spatial coordinates and lithology categories of all borehole data, a lithology database is built. Then, by constructing a 3D cubic mesh model of the region, a lithology sample library is built based on the lithology database and the 3D cubic mesh model. An initial graph attention network (GNN) model is constructed. Based on the first lithology sample library, the initial GNN model is iteratively trained to obtain a trained GNN model. A second lithology sample library is input into the trained GNN model to output the predicted lithology category for each second lithology sample. Based on the first lithology sample library... This application constructs a three-dimensional geological model of the area to be modeled by integrating all borehole data, constructing a lithology sample library through gridding, and then outputting the predicted lithology category of each second lithology sample through a trained graph attention network model. This solves the problem of uneven spatial distribution of borehole data and improves the accuracy of model prediction. Finally, based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed, which improves the accuracy of three-dimensional geological model construction.

[0139] It should be noted that the system embodiments described above are based on the same inventive concept as the method embodiments described above. Therefore, the relevant content of the method embodiments described above is also applicable to the system embodiments described above, and will not be repeated here.

[0140] Figure 3 A schematic diagram of the hardware structure for three-dimensional geological modeling provided in an embodiment of this application is shown.

[0141] The three-dimensional geological modeling equipment may include a processor 301 and a memory 302 storing computer program instructions.

[0142] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0143] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 302 is non-volatile solid-state memory.

[0144] In some embodiments, memory 302 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Thus, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0145] The processor 301 reads and executes computer program instructions stored in the memory 302 to implement any of the three-dimensional geological modeling methods in the above embodiments.

[0146] In one example, the 3D geological modeling device may also include a communication interface 303 and a bus 310. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 310 and complete communication with each other.

[0147] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0148] Bus 310 includes hardware, software, or both, that couples components of a 3D geological modeling device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 310 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0149] This 3D geological modeling equipment can execute the 3D geological modeling method in the embodiments of this application based on a 3D design model, thereby achieving a combination of Figure 1 and Figure 2 The methods and systems for three-dimensional geological modeling are described.

[0150] Furthermore, in conjunction with the three-dimensional geological modeling methods in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the three-dimensional geological modeling methods in the above embodiments.

[0151] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0152] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0153] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0154] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in 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, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0155] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method of three-dimensional geological modeling, characterized by, The three-dimensional geological modeling method includes: In constructing a borehole dataset for the region to be modeled, the spatial coordinates of each borehole data point in the borehole dataset are determined. The borehole dataset includes the lithology category of each borehole data point. Constructing the borehole dataset for the region to be modeled includes: Given all the borehole files for the region to be modeled, normalize the format of all the borehole files to obtain a normalized borehole text dataset. When the normalized borehole text dataset is divided into a first normalized borehole text dataset and a second normalized borehole text dataset according to a preset ratio, each first normalized borehole text data in the first normalized borehole text dataset is labeled to obtain a labeled borehole text dataset. An initial semantic recognition model is constructed. Based on the labeled post-drilling text dataset, the initial semantic recognition model is trained to obtain a trained semantic recognition model. The first normalized post-drilling text dataset and the second normalized post-drilling text dataset are input into the trained semantic recognition model to obtain the recognition result of each first normalized post-drilling text data and the recognition result of each second normalized post-drilling text data output by the trained semantic recognition model. Based on the recognition results of all the first normalized borehole text data and the recognition results of all the second normalized borehole text data, the borehole dataset is constructed; A lithology database is constructed based on the spatial coordinates and lithology categories of all the borehole data. In constructing a three-dimensional spatial cube mesh model of the region to be modeled, a lithological sample library is constructed based on the lithological database and the three-dimensional spatial cube mesh model. The lithological sample library includes a first lithological sample library and a second lithological sample library. The first lithological sample library includes the lithological category of each first lithological sample, while the second lithological sample library does not include the lithological category of any second lithological sample. An initial graph attention network model is constructed, and the initial graph attention network model is iteratively trained based on the first lithology sample library to obtain a trained graph attention network model. The second lithological sample library is input into the trained graph attention network model to output the predicted lithological category of each second lithological sample through the trained graph attention network model; Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed.

2. The method of three-dimensional geological modeling according to claim 1, characterized in that, The borehole dataset also includes the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, depth at the beginning of the lithological interval, and depth at the end of the lithological interval for each borehole data. The spatial coordinates include the spatial coordinates at the beginning and the spatial coordinates at the end. Determining the spatial coordinates of each borehole data in the borehole dataset includes: Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and the depth at the beginning of the lithology interval, the spatial coordinates at the beginning of each borehole data are determined, wherein the spatial coordinates at the beginning include the east coordinates, the north coordinates, and the elevation at the beginning; Based on the borehole number, borehole coordinates, inclination depth, inclination apex angle, inclination azimuth angle, and depth at the end of the lithological interval, the spatial coordinates at the end of each borehole data are determined, wherein the spatial coordinates at the end include the east coordinate of the end, the north coordinate of the end, and the elevation of the end.

3. The method of three-dimensional geological modeling according to claim 2, characterized in that, The lithology database is constructed based on the spatial coordinates and lithology categories of all the borehole data, including: When a line segment is obtained by connecting the spatial coordinates of the starting point and the spatial coordinates of the ending point, the line segment is linearly interpolated according to a preset discontinuity value using a linear interpolation method to obtain linear interpolation point data for each borehole data. The linear interpolation point data includes the east coordinate, north coordinate, elevation, and lithology category of each linear interpolation point, and each linear interpolation point has the same lithology category as its corresponding borehole data. The lithology database is constructed based on the spatial coordinates, lithology category, and linear interpolation point data of each borehole.

4. The method of three-dimensional geological modeling according to claim 3, wherein, In the case of constructing a three-dimensional spatial cube mesh model of the region to be modeled, a lithological sample library is constructed based on the lithological database and the three-dimensional spatial cube mesh model, including: Determine the minimum inter-point distance among all points in the lithology database; Using the minimum distance between points as the side length of the cube mesh, a three-dimensional spatial cube mesh model of the region to be modeled is constructed. Based on the east coordinates, north coordinates, and elevation of each lithological data in the lithological database, the lithological database is written into the three-dimensional spatial cubic mesh model to obtain the lithological cubic mesh model. The cubic mesh in the lithological cubic mesh model is used as the first cubic mesh, wherein the lithological category of the first cubic mesh is the same as the lithological category of the lithological data falling within the first cubic mesh, and the side length of the first cubic mesh is not included in the first cubic mesh. Extract data for each first cube grid, wherein the data for each first cube grid includes the position coordinates and lithology category label of each first cube grid, and the lithology category label is either a lithology category or a null value; The data of the first cube grid, which labels all the lithological categories as lithological categories, is used as the first lithological sample library; the data of the first cube grid, which labels all the lithological categories as null values, is used as the second lithological sample library.

5. A three-dimensional geological modeling method according to claim 4, characterized in that, The step of iteratively training the initial graph attention network model based on the first lithology sample library to obtain a trained graph attention network model includes: Based on the first lithological sample library, the adjacency matrix is ​​determined using the K-nearest neighbor algorithm; Based on the first lithological sample library and the adjacency matrix, the initial graph attention network model is iteratively trained until a preset maximum number of iterations is reached, thereby obtaining the trained graph attention network model.

6. The three-dimensional geological modeling method according to claim 1, characterized in that, The construction of a three-dimensional geological model of the area to be modeled, based on the first lithological sample library, the second lithological sample library, and the predicted lithology category of each second lithological sample, includes: Based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample, a three-dimensional geological model of the area to be modeled is constructed using the moving cube algorithm.

7. A three-dimensional geological modeling system, characterized in that, The three-dimensional geological modeling system includes: A spatial coordinate determination module is used to determine the spatial coordinates of each borehole data in a borehole dataset when constructing a borehole dataset for a region to be modeled. The borehole dataset includes the lithology category of each borehole data. The borehole dataset for constructing the region to be modeled includes: Given all the borehole files for the region to be modeled, normalize the format of all the borehole files to obtain a normalized borehole text dataset. When the normalized borehole text dataset is divided into a first normalized borehole text dataset and a second normalized borehole text dataset according to a preset ratio, each first normalized borehole text data in the first normalized borehole text dataset is labeled to obtain a labeled borehole text dataset. An initial semantic recognition model is constructed. Based on the labeled post-drilling text dataset, the initial semantic recognition model is trained to obtain a trained semantic recognition model. The first normalized post-drilling text dataset and the second normalized post-drilling text dataset are input into the trained semantic recognition model to obtain the recognition result of each first normalized post-drilling text data and the recognition result of each second normalized post-drilling text data output by the trained semantic recognition model. Based on the recognition results of all the first normalized borehole text data and the recognition results of all the second normalized borehole text data, the borehole dataset is constructed; A lithology database construction module is used to construct a lithology database based on the spatial coordinates and lithology categories of all the borehole data; The lithological sample library construction module is used to construct a lithological sample library based on the lithological database and the three-dimensional spatial cube mesh model of the area to be modeled, wherein the lithological sample library includes a first lithological sample library and a second lithological sample library, the first lithological sample library includes the lithological category of each first lithological sample, and the second lithological sample library does not include the lithological category of any second lithological sample; The model training module is used to construct an initial graph attention network model, and iteratively train the initial graph attention network model based on the first lithology sample library to obtain a trained graph attention network model. The predicted lithology category output module is used to input the second lithology sample library into the trained graph attention network model, so as to output the predicted lithology category of each second lithology sample through the trained graph attention network model; The three-dimensional geological model construction module is used to construct a three-dimensional geological model of the area to be modeled based on the first lithology sample library, the second lithology sample library, and the predicted lithology category of each second lithology sample.

8. A three-dimensional geological modeling device, characterized in that, It includes at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which, when executed by the at least one control processor, enable the at least one control processor to perform a three-dimensional geological modeling method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions for causing a computer to perform a three-dimensional geological modeling method as described in any one of claims 1 to 6.

Images