A method for three-dimensional modeling of a geological profile

Abstract

The present application relates to the technical field of geological image processing, and specifically discloses a three-dimensional modeling method for a geological profile, which is used to solve the problem of how to accurately identify and hide irrelevant geological areas in geological three-dimensional modeling and improve modeling efficiency, and comprises data acquisition and preprocessing, geological model constraint, implicit modeling and machine learning optimization, three-dimensional modeling and graphics rendering optimization, model correction and optimization, dynamic interaction and analysis, and four analysis formulas based on hiding requirements are proposed; the present application sets up hiding requirement judgment formulas for coal seams, rock strata, mined areas and non-active faults, screens geological areas irrelevant to target research, optimizes the geological three-dimensional modeling process, improves the accuracy and efficiency of the model, reduces irrelevant data interference, ensures that the mineral resource evaluation result is more in line with the actual geological conditions, and effectively improves the accuracy and reliability of mineral resource prediction.

Description

Technical Field

[0001] This invention relates to the field of geological image processing, and more specifically, to a three-dimensional modeling method for geological profiles. Background Technology

[0002] In existing technologies, 3D lithology modeling constraints are obtained from surface models in 2D geoscientific information databases and interpreted through human-computer interaction based on geophysical inversion and statistical laws of rock properties. 3D fracture modeling, a major component of 3D lithology modeling, uses fracture boundaries on geological maps and profiles as geological elements and human-computer interaction interpretation based on various geophysical information gradient zones as geophysical elements. In recent years, machine learning and deep learning algorithms have been widely applied in 3D lithology modeling, mainly including unsupervised clustering and supervised classification algorithms. While unsupervised clustering can improve modeling efficiency, it does not consider existing geological constraints. Supervised classification algorithms are faster and can consider existing geological constraints to complete geophysical interpretations, which is very advantageous for implicit 3D lithology modeling. However, when there is an imbalance in rock lithology categories, the prediction accuracy of the supervised classification algorithm will be affected if the training samples and algorithm are not adjusted. To address this, those skilled in the art have proposed a cost-sensitive positive sample unlabeled learning algorithm based on the "bagging method" to solve the extreme imbalance problem of having only a small number of positive samples in quantitative prediction of mineral resources. This algorithm is mainly applied to the quantitative prediction of gold mineral resources. However, compared to gold mines, coal mines have regional distribution stability, with relatively small thickness variations but a wide extension range. Coal-bearing strata are multi-layered with frequent thickness variations, layered sedimentary characteristics, and rapidly changing lithological differences. This leads to existing machine learning models also facing problems during training, such as lithological imbalance, large differences in the proportion of coal seam or interbedded gangue data samples, and missing coal seam borehole data in some areas. At the same time, coal-bearing strata may have gentle dips in some areas, dense fault areas, and combined fault and folding effects, with a greater focus on issues such as coal seam thickness changes, coal seam fracturing, and gas enrichment caused by fault displacement. Therefore, how to improve upon the machine learning methods in the published literature to solve the problems of current 3D geological profile modeling of coal seams using machine learning is a current challenge. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of existing technologies, this invention provides a three-dimensional modeling method for geological profiles. By setting a hidden demand judgment formula for coal seams, rock strata, mined areas, and inactive faults, it accurately filters out geological areas unrelated to the target study, optimizes the geological three-dimensional modeling process, improves the accuracy and efficiency of the model, reduces interference from irrelevant data, ensures that mineral resource assessment results are more consistent with actual geological conditions, and effectively improves the accuracy and reliability of mineral resource prediction.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A three-dimensional modeling method for geological profiles includes the following steps:

[0006] Step 1, Data Acquisition and Preprocessing: Combining geological information data source databases and geophysical data, machine learning methods are used to extract geological structural constraints, and human-computer interactive interpretation is carried out by combining two-dimensional geoscientific information and geophysical information gradient zones.

[0007] Step 2, Geological Model Constraints: Using geological structure data from a two-dimensional geoscience database, a three-dimensional geological model is constructed using linear model constraints. This model models the lithology, combines stratigraphic data and physical properties, and employs a surface model. Planar geological contact surfaces are established using lithological data, and fracture surfaces are constrained using fracture boundaries on geological profiles. Fractured sections that do not affect the structure are removed during the visualization stage.

[0008] Step 3, Implicit Modeling and Machine Learning Optimization: Unsupervised clustering algorithm is used to initially classify geological bodies into hierarchical categories, automatically monitor the spatial distribution of lithology categories, and supervised classification algorithm is used in combination with labeled borehole data to further improve the classification model;

[0009] Step 4, 3D modeling and graphics rendering optimization: Voxel technology is used to create volumetric rendering of underground lithology. Volumetric data is used to represent the transition and contact relationships between different underground rock layers. Different lithological layers are visualized through color mapping and transparency settings. Algorithms based on viewpoint selection and geometric visibility are used to delete the parts that need to be hidden during the 3D image generation process.

[0010] Step 5, Model calibration and optimization: Adjust the three-dimensional lithological layers and fault surfaces of the model according to the actual geological profile data to ensure that they conform to the geological conditions. Based on the geophysical inversion results, calibrate the rock physical property data in the model.

[0011] Step Six, Dynamic Interaction and Analysis: Using modern virtual reality technology, dynamic interaction is performed on the three-dimensional geological model. Through user-interactive control, geological profiles at different depths can be viewed in real time, and the distribution information of ore bodies can be analyzed.

[0012] As a further aspect of the present invention, the data acquisition and preprocessing process in step one includes:

[0013] Step 21, Data Acquisition: Obtain geological profile data, geological maps and geological structure maps, groundwater and soil information, and production exploration data from the geological information data source database. Geological profile data includes, but is not limited to, borehole data, stratigraphic thickness, lithological information, and mineral composition. Geological maps and geological structure maps include fault location, folds, and lithological distribution data. Groundwater and soil information includes, but is not limited to, hydrogeology and groundwater flow. Production exploration data includes, but is not limited to, mining information and ore body analysis information. Obtain gravity data, magnetic data, electromagnetic data, and seismic data from the geophysical database. Seismic data includes, but is not limited to, lithology and fault analysis information obtained through seismic wave velocity analysis.

[0014] Step 22, Data Storage and Integration: Establish a unified data repository, use a geographic information system to store and manage spatial data, integrate data from different sources, ensure spatial consistency between data, and convert it into a unified format (a unified format requires normalization and standardization of the acquired data to facilitate data processing and analysis).

[0015] As a further aspect of the present invention, in step 22, it is necessary to unify the format of data from different sources into numerical format, and then perform k-means clustering algorithm to standardize the data after it has been formatted into the same format, and then use the minimum-maximum normalization method to normalize the standardized data.

[0016] As a further aspect of the present invention, the specific implementation process of step two, which involves constraining the geological model, is as follows:

[0017] Step 31, Feature Extraction and Structural Constraints: Extract the stratum thickness, lithological information, location, strike and dip angle of geological faults, mineral composition and spatial distribution based on geological maps and profiles. Extract data related to gravity anomalies and underground density distribution, magnetic anomalies and rock magnetism, seismic wave velocity and lithology, geological structure, and electrical conductivity based on gravity, magnetic, and seismic wave data. Combine geological profiles and geophysical data, constrain the spatial distribution of different lithological regions based on the physical characteristics of different rock layers and seismic wave and gravity anomaly data. Infer the geometric shape of faults based on their strike and dip angle and seismic reflection data. Construct the geometry of the stratigraphic contact surface based on mineral composition, lithology, and fault information using seismic wave reflection data.

[0018] Step 32, Human-computer interactive explanation: Use Web GIS to provide geological experts with an intuitive data viewing and editing interface, supporting experts to annotate, edit and confirm unclear areas of data, optimize the machine learning model based on expert feedback, and update the data source and model of the machine learning model based on the new annotations added during the human-computer interaction process;

[0019] Step 33, Data Integration and Output: Merge geological and geophysical data, integrate all constraints based on machine learning models and expert feedback, and form the final input data.

[0020] As a further aspect of the present invention, in step three, the execution process for implicit modeling and machine learning optimization includes:

[0021] Step 41, Unsupervised algorithm for preliminary hierarchical classification: The K-means clustering algorithm is used to divide the data points into K clusters, where the center of each cluster is the average value of the data points. Geological profile data, lithological data, and geophysical data are input into the clustering algorithm. The clustering algorithm automatically discovers the hierarchical structure of the geological bodies and outputs the cluster identifier of each geological body point, thus performing preliminary hierarchical classification.

[0022] Step 42: Further optimize the model using a supervised classification algorithm: Combine the labeled borehole data, select features that affect lithology classification, train a random forest model using the labeled borehole data to enable the model to learn the lithology classification pattern, evaluate the model performance using cross-validation, adjust the model's hyperparameters based on accuracy and recall, optimize the classification results, and output the classification results of the geological body based on the probability values ​​output by the classification model.

[0023] Step 43, combine unsupervised and supervised classification results: combine the preliminary results of unsupervised classification with the classification results of supervised learning, and use a weighted average method to obtain the classification result for each geological body.

[0024] In practical geological modeling, the quality and depth of coal seams are crucial for mineral resource assessment. It's necessary to consider both coal seam quality and depth to determine whether low-quality or excessively deep coal seams need to be hidden, ensuring that the 3D model only includes the portions that meet the research requirements. In ore cluster modeling, the primary focus is on the coal seams, while unrelated rock strata can interfere with the modeling results. Explicitly hiding irrelevant rock strata effectively focuses the research objective, reduces model complexity, and improves data processing efficiency and subsequent analysis accuracy. Mined areas within the mining area should be hidden during modeling to avoid misjudging or interfering with unmined portions. Using mining depth and a cutting threshold, areas that have been fully mined can be identified and removed from the 3D model. Inactive faults can be misleading during mining area modeling, especially when they do not affect the distribution and assessment of mineral resources; therefore, inactive faults that do not affect ore body prediction should be hidden. Manually screening and intervening in the above data is inefficient and limited by the expertise and knowledge of the screening personnel, resulting in uncertainty and poor accuracy.

[0025] As a further aspect of the present invention, in step four, the deletion of the identified parts that need to be hidden during the three-dimensional image generation process is achieved in the following way:

[0026] Step 11, Calculate the normal vector: Obtain the unit vector of the resulting vector by taking the cross product of the two vectors formed by the key points of the stratigraphic profile on each geological profile and their two nearest neighboring points;

[0027] Step 12, generate shadows: Based on the above key points, determine whether there is a shadow by the angle between the light source and the normal vector of the key point. If the angle is 0, there is no shadow; if the angle is not zero, there is a shadow. The maximum value between the dot product of the normal vector of the key point and the unit vector from the light source to the key point and 0 is used as the illumination intensity of the key point.

[0028] Step 13, Remove hidden parts: For the above key points, if the distance from the point to the surrounding set area is less than a preset threshold, the set area is regarded as the geological area that needs to be hidden, and the geological area that needs to be hidden is removed.

[0029] Step 14, Generate Rendering: Combining the shadow intensity with the information that needs to be hidden in Step 13, generate the final visualized geological profile rendering.

[0030] As a further aspect of the present invention, in step 13, the needs to be hidden include, but are not limited to, the hidden needs of non-target coal seams, the hidden needs of rock strata related to non-coal seam research, the hidden needs of mined areas, and the hidden needs of inactive faults. The determination of each hidden need is based on the numerical values ​​calculated by the hidden need determination formulas for non-target coal seams, non-coal seam research-related rock strata, mined areas, and inactive faults.

[0031] As a further aspect of this invention, in the formula for judging the hidden demand of non-target coal seams, based on the quality and depth of the coal seam, a Gaussian distribution is used to measure the depth difference of the coal seam, and a weighted activation function is used to process the difference between the coal seam depth and the maximum depth to analyze the mineability of the coal seam. The formula for judging the hidden demand of non-target coal seams is as follows:

[0032]

[0033] In the formula: P is the key point of the geological profile, H c (P) represents the hidden demand analysis value of the non-target coal seam at point P, Q P Let Q be the coal seam quality at point P. min D represents the minimum quality threshold for the coal seam. P Let D be the depth of the coal seam at point P. maxα1 is the maximum exploitable depth of the coal seam, σ1 is the coal seam quality correction coefficient, and σ1 is the standard deviation of the sensitivity to the influence of coal seam depth.

[0034] H c (P) If the threshold is exceeded, the coal seam is considered an unimportant coal seam, and the important coal seam is deleted using the profile method.

[0035] As a further aspect of this invention, in the formula for determining the concealment requirement of non-coal seam-related rock strata, the type, thickness, and location of the rock strata are used to determine whether these non-coal seam-related rock strata should be concealed. A threshold is used to determine whether they are rock strata unrelated to coal seam research. Furthermore, the concealment determination is dynamically adjusted based on the rock strata thickness and the distance between the rock strata and the research area. The formula for determining the concealment requirement of non-coal seam-related rock strata is as follows:

[0036] In (T) R ≠k c )∧(T P >T max When H nc (P) = 1;

[0037] If (T) is not satisfied R ≠k c )∧(T P >T max )hour,

[0038] In the formula: T R k is the lithological analysis value. c The numerical representative value of the coal seam is obtained through preset calculation, T P T represents the thickness of the rock strata. max H represents the maximum thickness of the rock strata. nc (P) represents the hidden demand analysis value of the relevant rock strata in non-coal seam research, and D... P,max Let D be the maximum depth of the coal seam at point P. pro R represents the distance between the rock strata and the study area. min α2 is the minimum distance between the rock strata and the study area, σ2 is the influence control factor of the coal seam depth at point P, and sigmoid(·) is the sigma function.

[0039] H nc If (P) is greater than the preset threshold, the area is considered a non-coal seam study area and is removed by the region segmentation algorithm.

[0040] As a further aspect of the present invention, the hidden demand judgment formula for the already mined area is based on the depth of the already mined area, the mining progress, and its relative position to the current research area. The hidden demand judgment formula for the already mined area is as follows:

[0041]

[0042] In the formula: H Mining (P) represents the hidden demand assessment value for the already mined area at point P, and L... expl (P) represents the mining progress of the coal seam where point P is located, and L expl,min α3 is the minimum display threshold for coal seam mining progress, and D is the correction coefficient for coal seam mining progress. expl (P) represents the distance between the mined area of ​​the coal seam where point P is located and the study area, β2 represents the correction coefficient for the apparent influence of the distance between the mined area and the study area, and σ3 represents the control parameter for the apparent influence of the distance between the mined area and the study area.

[0043] H Mining If (P) is greater than a preset threshold, the area is considered a mining area and is masked using masking technology.

[0044] As a further aspect of the present invention, the formula for determining the need to hide inactive faults is based on the activity, location, and magnitude of influence of the faults to determine whether these faults should be hidden. The formula for determining the need to hide inactive faults is as follows:

[0045]

[0046] In the formula: H fault (P) represents the hidden demand assessment value for inactive faults, F fa (P) represents the fault activity analysis value, obtained through seismic activity and slippage analysis; α4 is the indicative influence coefficient of coal seam activity; f fa,min D is the minimum value of the fault activity analysis. TCA (P) represents the distance between the fault region where point P is located and the study area, and D TCA,max β3 represents the maximum permissible display distance between the fault region and the study area, and β3 is the adjustment factor for the influence of the fault region distance.

[0047] H fault If (P) is greater than the preset threshold, the area is considered an inactive fault. The inactive fault is then rendered with a transparency effect using segmented rendering technology.

[0048] As a further aspect of the present invention, in step five, model calibration and optimization includes the following steps:

[0049] Step 51, Correcting the lithological layers and fault surfaces in the model: Based on the geological profile data, adjust the three-dimensional lithological layers and fault surfaces. Adjust the boundaries, thickness, and spatial distribution of the lithological layers in the three-dimensional model, especially for the contact surfaces between different lithologies. Use interpolation to fill in the data gaps to ensure the continuity of the lithological layers throughout the region. Adjust each lithological layer in the three-dimensional model according to the changes in lithology in the actual profile, especially for abrupt changes, to ensure that the lithological layers conform to the geological structural constraints in three-dimensional space. Use the location information of fault lines in the geological profile data and geological maps to obtain the strike, dip angle, and displacement of faults. Adjust the fault surfaces in the three-dimensional model according to the location and properties of the actual fault surfaces to ensure that they conform to the actual structural characteristics. For known active faults, emphasize the influence of their displacement. Use geometric constraints (such as plane equations) to correct the fault surfaces to ensure that the shape of the fault surfaces is consistent with the actual profile data.

[0050] Step 52, Correcting rock property data based on geophysical inversion results: The seismic velocity model obtained based on reflected wave data is used to estimate underground lithology and porosity. The density distribution of underground materials is inverted through geophysical measurement data. The underground electrical parameters obtained from electromagnetic wave detection results are used. The seismic velocity model and geophysical inversion results are used to compare the inverted data with the original property data through regression analysis to obtain an optimized rock property model.

[0051] Step 53, Verify the corrected 3D geological model: Compare the corrected 3D geological model with the actual geological profile, strictly check the accuracy of lithological distribution and fault location, calculate the error of the model before and after correction, including but not limited to lithological distribution error, fault location error, and physical property data error, and use standard error for evaluation.

[0052] To address the technical problems raised in the background, this invention proposes a three-dimensional modeling method for geological profiles. The technical effects include: by integrating three-dimensional implicit modeling, machine learning optimization, and geological constraint correction, this invention combines geological information databases and geophysical data, utilizes machine learning methods to extract geological structural constraints, constructs a three-dimensional structural model, and achieves lithology and fault surface modeling through surface models and fault boundary constraints from two-dimensional geoscientific information. An unsupervised clustering algorithm is used for preliminary classification of geological bodies, and combined with a supervised classification algorithm, the model is further optimized using borehole data. This overcomes the challenge of lithological class imbalance and improves modeling accuracy. In terms of efficiency, during the modeling process, the algorithm extracts lithological distribution, ore body characteristics, and fault information from known borehole data and geological profiles to ensure that the model structure matches the actual geological conditions. Geophysical inversion and statistical laws of rock properties are combined with human-computer interactive interpretation to effectively correct deviations in the lithological model. By establishing three-dimensional lithological layers and fault surfaces, non-target coal seams, non-coal seams, mined areas, and inactive faults are hidden based on preset formulas. The algorithm automatically judges and processes the data that needs to be hidden, reducing the complexity of manual intervention and data processing, and clearly displaying the underground structure and resource distribution of the mining area, providing a reliable basis for quantitative prediction of mineral resources. Attached Figure Description

[0053] Figure 1 This is a flowchart of a three-dimensional modeling method for geological profiles proposed in this invention;

[0054] Figure 2 This is a flowchart illustrating the specific implementation of step four of the method proposed in this invention;

[0055] Figure 3 This is a flowchart illustrating the specific implementation of step one of the method proposed in this invention;

[0056] Figure 4 This is a flowchart illustrating the specific implementation of step two of the method proposed in this invention;

[0057] Figure 5 This is a flowchart illustrating the specific implementation of step three of the method proposed in this invention;

[0058] Figure 6 This is the algorithm control panel interface of the application platform for the method proposed in this invention;

[0059] Figure 7 This is a screenshot of the data processing interface of the application platform for the method proposed in this invention.

[0060] Figure 8 This is a classification result diagram of the application platforms of the method proposed in this invention;

[0061] Figure 9This is a diagram of the three-dimensional geological profile model control interface of the application platform for the method proposed in this invention. Detailed Implementation

[0062] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0063] like Figures 1 to 5 As shown, the present invention proposes a three-dimensional modeling method for geological profiles, comprising the following steps:

[0064] Step 1, Data Acquisition and Preprocessing: Combining geological information data source databases and geophysical data, machine learning methods are used to extract geological structural constraints, and human-computer interactive interpretation is carried out by combining two-dimensional geoscientific information and geophysical information gradient zones.

[0065] Step 2, Geological Model Constraints: Using geological structure data from a two-dimensional geoscience database, a three-dimensional geological model is constructed using linear model constraints. This model models the lithology, combines stratigraphic data and physical properties, and employs a surface model. Planar geological contact surfaces are established using lithological data, and fracture surfaces are constrained using fracture boundaries on geological profiles. Fractured sections that do not affect the structure are removed during the visualization stage.

[0066] Step 3, Implicit Modeling and Machine Learning Optimization: Unsupervised clustering algorithm is used to initially classify geological bodies into hierarchical categories, automatically monitor the spatial distribution of lithology categories, and supervised classification algorithm is used in combination with labeled borehole data to further improve the classification model;

[0067] Step 4, 3D modeling and graphics rendering optimization: Voxel technology is used to create volumetric rendering of underground lithology. Volumetric data is used to represent the transition and contact relationships between different underground rock layers. Different lithological layers are visualized through color mapping and transparency settings. Algorithms based on viewpoint selection and geometric visibility are used to delete the parts that need to be hidden during the 3D image generation process.

[0068] Step 5, Model calibration and optimization: Adjust the three-dimensional lithological layers and fault surfaces of the model according to the actual geological profile data to ensure that they conform to the geological conditions. Based on the geophysical inversion results, calibrate the rock physical property data in the model.

[0069] Step Six, Dynamic Interaction and Analysis: Using modern virtual reality technology, dynamic interaction is performed on the three-dimensional geological model. Through user-interactive control, geological profiles at different depths can be viewed in real time, and the distribution information of ore bodies can be analyzed.

[0070] It should be noted that the data acquisition and preprocessing process in step one includes:

[0071] Step 21, Data Acquisition: Obtain geological profile data, geological maps and geological structure maps, groundwater and soil information, and production exploration data from the geological information data source database. Geological profile data includes, but is not limited to, borehole data, stratigraphic thickness, lithological information, and mineral composition. Geological maps and geological structure maps include fault location, folds, and lithological distribution data. Groundwater and soil information includes, but is not limited to, hydrogeology and groundwater flow. Production exploration data includes, but is not limited to, mining information and ore body analysis information. Obtain gravity data, magnetic data, electromagnetic data, and seismic data from the geophysical database. Seismic data includes, but is not limited to, lithology and fault analysis information obtained through seismic wave velocity analysis.

[0072] Step 22, Data Storage and Integration: Establish a unified data repository, use a geographic information system to store and manage spatial data, integrate data from different sources, ensure spatial consistency between data, convert the data to a unified format, and perform data preprocessing (in Step 22, it is necessary to unify the format of data from different sources into numerical format, and then use the k-means clustering algorithm to standardize the data after it is in the same format, and then use the min-max normalization method to normalize the standardized data).

[0073] It should be noted that the specific implementation process for step two, which involves constraining the geological model, is as follows:

[0074] Step 31, Feature Extraction and Structural Constraints: Extract the stratum thickness, lithological information, location, strike and dip angle of geological faults, mineral composition and spatial distribution based on geological maps and profiles. Extract data related to gravity anomalies and underground density distribution, magnetic anomalies and rock magnetism, seismic wave velocity and lithology, geological structure, and electrical conductivity based on gravity, magnetic, and seismic wave data. Combine geological profiles and geophysical data, constrain the spatial distribution of different lithological regions based on the physical characteristics of different rock layers and seismic wave and gravity anomaly data. Infer the geometric shape of faults based on their strike and dip angle and seismic reflection data. Construct the geometry of the stratigraphic contact surface based on mineral composition, lithology, and fault information using seismic wave reflection data.

[0075] Step 32, Human-computer interactive explanation: Use Web GIS to provide geological experts with an intuitive data viewing and editing interface, supporting experts to annotate, edit and confirm unclear areas of data, optimize the machine learning model based on expert feedback, and update the data source and model of the machine learning model based on the new annotations added during the human-computer interaction process;

[0076] Step 33, Data Integration and Output: Merge geological and geophysical data, integrate all constraints based on machine learning models and expert feedback, and form the final input data.

[0077] The data acquisition and preprocessing process in step one involves obtaining various types of data from geological information databases and geophysical databases, including geological profiles, geological maps, geological structure maps, groundwater and soil information, production exploration data, and gravity, magnetic, electromagnetic, and seismic data. This provides accurate input for subsequent modeling. Furthermore, by establishing a unified data repository and using a geographic information system to manage and integrate data from different sources, spatial consistency and accuracy of the data are ensured, providing high-quality data support for 3D geological modeling.

[0078] It should be noted that the execution flow for implicit modeling and machine learning optimization in step three includes:

[0079] Step 41, Unsupervised algorithm for preliminary hierarchical classification: The K-means clustering algorithm is used to divide the data points into K clusters, where the center of each cluster is the average value of the data points. Geological profile data, lithological data, and geophysical data are input into the clustering algorithm. The clustering algorithm automatically discovers the hierarchical structure of the geological bodies and outputs the cluster identifier of each geological body point, thus performing preliminary hierarchical classification.

[0080] Step 42: Further optimize the model using a supervised classification algorithm: Combine the labeled borehole data, select features that affect lithology classification, train a random forest model using the labeled borehole data to enable the model to learn the lithology classification pattern, evaluate the model performance using cross-validation, adjust the model's hyperparameters based on accuracy and recall, optimize the classification results, and output the classification results of the geological body based on the probability values ​​output by the classification model.

[0081] Step 43, combine unsupervised and supervised classification results: combine the preliminary results of unsupervised classification with the classification results of supervised learning, and use a weighted average method to obtain the classification result for each geological body.

[0082] Step 3 optimizes the hierarchical and classification of geological bodies by combining unsupervised K-means clustering and supervised random forest classification algorithms. K-means clustering automatically identifies the preliminary hierarchical structure of geological bodies and performs preliminary classification of geological data based on the clustering results. Subsequently, a random forest model is trained using labeled borehole data to further optimize the classification. Hyperparameters are adjusted through cross-validation to improve the accuracy and robustness of the model. Finally, a weighted average method is used to fuse the results of both methods, thereby achieving a more accurate classification of geological bodies and providing more accurate data support for 3D modeling.

[0083] In practical geological modeling, the accuracy and efficiency of models can be effectively improved by setting specific judgment formulas and thresholds to address the need for concealing different areas such as coal seams, rock strata, mined areas, and inactive faults. First, a comprehensive assessment of coal seam quality and depth is conducted to conceal low-quality or excessively deep seams, ensuring the model only includes the coal seam portion that meets the research requirements, thus avoiding interference from irrelevant data. Second, in the modeling of mineral clusters, rock strata unrelated to coal seam research are explicitly concealed. By screening relevant data, the focus is on the coal seam research objective, reducing unnecessary data processing and improving modeling efficiency and the accuracy of subsequent analysis. Third, by setting mining depth and cutting thresholds, mined areas are concealed to avoid misjudgment and interference with unmined bodies. Finally, for inactive faults, manual screening and intervention remove inactive faults that do not affect the distribution and assessment of mineral resources, further improving the accuracy and rationality of the modeling. These operations effectively optimize data processing, model simplification, and accurate prediction in the modeling process, providing more reliable support for mineral resource assessment.

[0084] In practical geological modeling, the quality and depth of coal seams are crucial for mineral resource assessment. It's necessary to consider both coal seam quality and depth to determine whether low-quality or excessively deep coal seams need to be hidden, ensuring that the 3D model only includes the portions that meet the research requirements. In ore cluster modeling, the primary focus is on the coal seams, while unrelated rock strata can interfere with the modeling results. Explicitly hiding irrelevant rock strata effectively focuses the research objective, reduces model complexity, and improves data processing efficiency and subsequent analysis accuracy. Mined areas within the mining area should be hidden during modeling to avoid misjudging or interfering with unmined portions. Using mining depth and a cutting threshold, areas that have been fully mined can be identified and removed from the 3D model. Inactive faults can be misleading during mining area modeling, especially when they do not affect the distribution and assessment of mineral resources; therefore, inactive faults that do not affect ore body prediction should be hidden. Manually screening and intervening in the above data is inefficient and limited by the expertise and knowledge of the screening personnel, resulting in uncertainty and poor accuracy.

[0085] It should be noted that in step four, the removal of the identified parts that need to be hidden during the 3D image generation process is achieved in the following way:

[0086] Step 11, Calculate the normal vector: The unit vector of the vector obtained by the cross product of the two vectors formed by the key point of the stratigraphic profile on each geological profile and its two nearest neighboring points is the normal vector of that key point.

[0087] Step 12, generate shadows: Based on the above key points, determine whether there is a shadow by the angle between the light source and the normal vector of the key point. If the angle is 0, there is no shadow; if the angle is not zero, there is a shadow. The maximum value between the dot product of the normal vector of the key point and the unit vector from the light source to the key point and 0 is used as the illumination intensity of the key point.

[0088] Step 13, Remove hidden parts: For the above key points, if the distance from the point to the surrounding set area is less than a preset threshold, the set area is regarded as the geological area that needs to be hidden, and the geological area that needs to be hidden is removed.

[0089] Step 14, Generate Rendering: Combining the shadow intensity with the information that needs to be hidden in Step 13, generate the final visualized geological profile rendering.

[0090] It should be noted that in step 13, the needs to be hidden include, but are not limited to, the hidden needs of non-target coal seams, the hidden needs of rock strata related to non-coal seam research, the hidden needs of mined areas, and the hidden needs of inactive faults. The determination of each hidden need is based on the numerical values ​​calculated by the hidden need determination formulas for non-target coal seams, non-coal seam research related rock strata, mined areas, and inactive faults.

[0091] It should be noted that in the formula for judging the hidden demand of non-target coal seams, based on the quality and depth of the coal seam, a Gaussian distribution is used to measure the depth difference of the coal seam, and a weighted activation function is used to process the difference between the coal seam depth and the maximum depth to analyze the mineability of the coal seam. The formula for judging the hidden demand of non-target coal seams is as follows:

[0092]

[0093] In the formula: P is the key point of the geological profile, H c (P) represents the hidden demand analysis value of the non-target coal seam at point P, Q P Let Q be the coal seam quality at point P. min D represents the minimum quality threshold for the coal seam. P Let D be the depth of the coal seam at point P. max α1 is the maximum exploitable depth of the coal seam, σ1 is the dimensional balance coefficient, and σ1 is the standard deviation of the sensitivity to the influence of coal seam depth.

[0094] H c (P) If the threshold is exceeded, the coal seam is considered an unimportant coal seam, and the important coal seam is deleted using the profile method.

[0095] In this formula, By using a Gaussian activation function to process the difference between coal seam quality and a minimum quality threshold, coal seams below the minimum quality threshold are considered unmineable and thus hidden. The Gaussian distribution is used to measure the difference between the coal seam depth and the maximum exploitable depth. Coal seams that are deeper than the maximum exploitable depth are considered unmineable and need to be hidden.

[0096] Example 1

[0097] The specific execution process of the above formulas will be explained with specific examples:

[0098] After extracting multi-source data from borehole records and profiles, the data was converted into a unified numerical format and stored in a database, ensuring that each point had two fields: coal seam quality and coal seam depth. The acquired data contained five data points, whose original characteristics were coal seam quality and coal seam depth. Specifically, point P1 had a coal seam quality score of 60 and a depth of 500 meters; point P2 had a coal seam quality score of 45 and a depth of 700 meters; point P3 had a coal seam quality score of 55 and a depth of 550 meters; point P4 had a coal seam quality score of 50 and a depth of 650 meters; and point P5 had a coal seam quality score of 65 and a depth of 480 meters.Preliminary classification was performed using k-means clustering. The data points were divided into two categories based on the similarity of coal seam quality and depth, with k=2. Cluster A contained data points P1, P3, and P5, and cluster B contained data points P2 and P4. Local standardization (Z-score) was performed within each cluster. For cluster A (P1, P3, P5), the mean and standard deviation of coal seam quality were calculated to be 60 and 4.08, respectively, and the mean and standard deviation of coal seam depth were calculated to be 510 and 29.44, respectively. Therefore, the coal seam quality of data points P1, P3, and P5 was standardized using Z-score. The standardized values ​​were 0, -1.225, and 1.225, respectively. The Z-score standardized values ​​for coal seam depth were -0.34, 1.36, and -1.02, respectively. Similarly, the mean and standard deviation of the data were calculated for the three points in cluster B. The mean and standard deviation of coal seam quality were 47.5 and 2.5, respectively, and the mean and standard deviation of coal seam depth were 675 and 25, respectively. The Z-score standardized values ​​for coal seam quality at data points P2 and P4 were -1 and 1, respectively, and the Z-score standardized values ​​for coal seam depth were 1 and -1, respectively. After standardization within each cluster, the values ​​for all... The standardized values ​​of the data are subjected to global minimum and maximum normalization, mapping each feature to [0,1]. For coal seam quality, the summarized values ​​P1, P3, P5, P2, and P4 are 0, -1.225, 1.225, -1, and 1, respectively. The global minimum value is -1.225, and the global maximum value is 1.225. After processing using the global minimum-maximum normalization formula, P1, P3, P5, P2, and P4 are 0.5, 0, 1, 0.0918, and 0.9082, respectively. For coal seam depth, the summarized values ​​P1, P3, P5, P2, and P4 are -0.34 and 1.36, respectively. The values ​​are -1.02, 1, and -1. The global minimum value is -1.02, and the global maximum value is 1.36. After the global minimum-maximum normalization formula, P1, P3, P5, P2, and P4 are 0.2857, 1, 0, 0.8487, and 0.0084, respectively. Then, using the non-target coal seam hidden demand judgment formula, under the premise that the minimum quality threshold and the maximum mining depth threshold are set to 0.5 and 0.5, respectively, and taking α1 and σ1 to be 0.2 and 0.2, respectively, the data processing for points P1, P3, P5, P2, and P4 is as follows (decimals are rounded to four decimal places):

[0099] (1) For P1:

[0100]

[0101] Among them H c (P) = 0.5 * 0.563 = 0.2815;

[0102] (2) For P3:

[0103]

[0104] Among them H c (P) = 0.525 * 0.044 = 0.0231;

[0105] (3) For P5:

[0106]

[0107] Among them H c (P) = 0.475 * 0.044 = 0.0209;

[0108] (4) For P2:

[0109]

[0110] Among them H c (P) = 0.52 * 0.221 = 0.1148;

[0111] (5) For P4:

[0112]

[0113] Among them H c (P)=0.4797*0.049=0.0235;

[0114] Finally, when the preset hiding threshold is 0.05 (the actual value is adjusted based on experience and verification data), according to H... c Formula determination of (P):

[0115] P1: H c (P) = 0.2815 > 0.05, hidden;

[0116] P2: H c (P) = 0.1148 > 0.05, hidden;

[0117] P3: H c (P) = 0.0235 ≤ 0.05, display;

[0118] P4: H c (P) = 0.0235 ≤ 0.05, display;

[0119] P5: H c (P) = 0.0209 ≤ 0.05, display;

[0120] The normalized and processed data obtained above, and the H values ​​of each key point are then processed. c (P) The data and the final "show" and "hide" markers are stored in the GIS database and the 3D modeling system. During the rendering process, the system selects to render the entire area or to hide it based on the display status of the key points corresponding to each area.

[0121] It should be noted that in the formula for determining the concealment requirement of non-coal seam related rock strata, the type, thickness, and location of the rock strata are used to determine whether these non-coal seam related rock strata are concealed. A threshold is used to determine whether they are rock strata unrelated to coal seam research. The concealment determination is further dynamically adjusted based on the rock strata thickness and the distance between the rock strata and the research area. The formula for determining the concealment requirement of non-coal seam related rock strata is as follows:

[0122] In (T) R ≠k c )∧(T P >T max When H nc (P) = 1;

[0123] If (T) is not satisfied R ≠k c )∧(T P >T max )hour,

[0124] In the formula: T R k is the lithological analysis value. c The numerical representative value of the coal seam is obtained through preset calculation, T P T represents the thickness of the rock strata. max H represents the maximum thickness of the rock strata. nc (P) represents the hidden demand analysis value of the relevant rock strata in non-coal seam research, and D... P,max Let D be the maximum depth of the coal seam at point P. pro R represents the distance between the rock strata and the study area. min α2 is the preset minimum distance between the rock strata and the study area, σ2 is the dimensional balance coefficient, σ2 is the influence regulation factor of the distance between the rock strata and the study area, and sigmoid(·) is the sigma function.

[0125] H nc If (P) is greater than the preset threshold, the area is considered a non-coal seam study area and is removed by the region segmentation algorithm.

[0126] (T R ≠k c )∧(T P >T maxWhen the rock type is not a coal seam and the rock layer thickness exceeds the maximum threshold, the rock layer is judged to be a non-coal seam research-related rock layer, and the hiding requirement value is set to 1, indicating that the rock layer needs to be hidden. The hiding requirement value is adjusted by using the sigmoid function and the Gaussian function. The sigmoid function corrects the hiding requirement based on the rock layer depth and the maximum depth value, ensuring that the hiding requirement is enhanced only when the rock layer depth is greater than its maximum depth. At the same time, the Gaussian function considers the distance between the rock layer and the research area. Rock layers that are closer to the research area are weighted according to the square difference of the distance, ensuring that rock layers that are farther away are not misjudged as research-related rock layers.

[0127] Example 2

[0128] To further illustrate the processing procedure, the following example is provided:

[0129] The standardized and normalized data information of the acquired data points is shown in Table 1 below:

[0130] Table 1: Data Details Table

[0131]

[0132]

[0133] The default parameter is set to: k c =0.50, T max =0.70, α2=0.20, R min =0.30;

[0134] For point P6:

[0135] Inspection conditions: (T) R ≠k c )∧(T P >T max P6 point T R =0.30 and k c =0.50 are not equal, and T P =0.80 is greater than T max =0.70, the condition is met, H is directly defined. nc (P) = 1;

[0136] For point P7:

[0137] Inspection conditions: (T) R ≠k c )∧(T P >T max P7 point T R =0.50 and k c =0.50 is equal, the condition is not met, proceed to the next branch, and use the sigmoid and Gaussian functions to calculate.

[0138] For point P8:

[0139] Inspection conditions: (T) R ≠k c )∧(T P >T max P8 point T R =0.4 and k c =0.50 are not equal, but T P =0.60 is greater than T max =0.70, the condition is not fully met, proceed to the next branch, and use the sigmoid and Gaussian functions to calculate.

[0140] Perform hidden detection and region segmentation:

[0141] With a preset hiding threshold of 0.5, when H nc If (P) is greater than 0.5, the area is determined to be a non-coal seam research area. This area is hidden using a region segmentation algorithm. The determination of each point is as follows:

[0142] For point P6: H nc (P) = 1 > 0.5, which is considered a non-coal seam related rock stratum and needs to be hidden;

[0143] For point P7: H nc (P) = 0.307 < 0.5, which is considered a region relevant to coal seam research, indicating;

[0144] For point P8: H nc (P) = 0.0166 < 0.5, which is considered to be the region related to coal seam research, as shown.

[0145] It should be noted that the formula for determining hidden demand in already mined areas is based on the depth of the mined area, the mining progress, and its relative position to the current research area. The formula for determining hidden demand in already mined areas is as follows:

[0146]

[0147] In the formula: H Mining (P) represents the hidden demand assessment value for the already mined area at point P, and L... expl (P) represents the mining progress of the coal seam where point P is located, and L expl,min The minimum threshold for indicating coal seam mining progress, α3 is the dimensional balance coefficient, and D expl(P) represents the distance between the mined area of ​​the coal seam where point P is located and the study area, β2 represents the correction coefficient for the apparent influence of the distance between the mined area and the study area, and σ3 represents the control parameter for the apparent influence of the distance between the mined area and the study area.

[0148] H Mining If (P) is greater than a preset threshold, the area is considered a mining area and is masked using masking technology.

[0149] It is an S-shaped curve used to assess whether an area needs to be hidden based on the mining progress. As the mining progress increases, the value of the hiding requirement gradually increases. When the mining progress is higher than the minimum threshold, the value tends to 1, indicating that the area is considered to have been mined and hidden. This is a Gaussian distribution function used to determine whether to hide areas based on distance. Areas farther from the study area have less impact, while closer, already mined areas require a stronger hiding value. The formula effectively combines mining progress and the relative position of the mined area to the study area to ensure that only data relevant to the study is retained. By controlling thresholds and correction coefficients, it flexibly hides mined areas, avoiding unnecessary influence of mined areas on ore body analysis and modeling results, thus improving the accuracy and effectiveness of mineral resource assessment.

[0150] Example 3

[0151] To further illustrate the specific function of the analytical formula proposed in this invention, example data from P9, P10, and P11, after format standardization, local standardization, and global normalization, are provided in Table 2.

[0152] Table 2. Detailed data of example points P9, P10, and P11

[0153]

[0154] Meanwhile, the preset parameters are: α3 = 10, L expl,min =0.4, β2=5, σ3=0.2;

[0155] The distribution calculation formula has two parts, the first being... Analysis and calculation of three points:

[0156] (1) P9:

[0157] (2) P10:

[0158] (3) P11:

[0159] Next, calculate the corresponding values ​​of the three points. Partially, specifically:

[0160] (1) P9:

[0161] (2) P10:

[0162] (3) P11:

[0163] Then obtain H Mining (P) is as follows:

[0164] (1) P9: H Mining (P) = 0.931 * 4.25 * 10 -18 =3.96*10 -18 (2) P10: H Mining (P) = 0.9986 * 4.54 * 10 -5 =4.54*10 -5 ;

[0165] (3) P11:H Mining (P)=0.9995*0.535=0.5347;

[0166] Assuming the preset threshold is 0.5, H Mining If (P) is greater than 0.5, then the area where this point is located is considered to be a mined area and needs to be masked using masking technology.

[0167] P9:H Mining (P) = 3.96 * 10 -18 If the value is below 0.5, it will not be hidden;

[0168] P10: H Mining (P) = 4.54 * 10 -5 If the value is below 0.5, it will not be hidden;

[0169] P11: H Mining (P) = 0.5347, greater than 0.5, hidden.

[0170] It should be noted that the formula for determining the hidden demand of inactive faults is based on the fault's activity, location, and magnitude of influence to determine whether these faults are hidden. The formula for determining the hidden demand of inactive faults is as follows:

[0171]

[0172] In the formula: H fault (P) represents the hidden demand assessment value for inactive faults, F fa(P) represents the fault activity analysis value, obtained through seismic activity and slippage analysis; α4 is the dimensional equilibrium coefficient; f fa,min D is the minimum value of the fault activity analysis. TCA (P) represents the distance between the fault region where point P is located and the study area, and D TCA,max β3 represents the maximum permissible display distance between the fault region and the study area, and β3 is the adjustment factor for the influence of the fault region distance.

[0173] H fault If (P) is greater than the preset threshold, the area is considered an inactive fault. The inactive fault is then rendered with a transparency effect using segmented rendering technology.

[0174] The formula for determining the need to hide inactive faults is used to judge whether these faults need to be hidden during the modeling process based on their activity, location, and impact on mineral resource distribution. This formula comprehensively evaluates factors such as fault activity and the extent of their influence to determine whether to hide faults, thereby improving the accuracy of mineral resource assessment and avoiding interference from irrelevant faults in the model results. The formula accurately determines whether inactive faults should be hidden in geological modeling by comprehensively evaluating fault activity, distance from the study area, and their impact on mineral resources. This formula uses fault activity analysis values, location distance, and influence factors, and employs advanced mathematical methods such as Gaussian distribution and the sigmoid function to dynamically adjust the degree of influence of fault display. Specifically, by combining fault activity with location relationships, the formula automatically adjusts the visibility of inactive faults in the 3D model, effectively avoiding irrelevant faults from misleading mineral resource distribution and assessment, reducing noise interference in the model, and improving modeling accuracy. Meanwhile, this formula can precisely control the display range of faults, ensuring that only faults that have a substantial impact on mineral resource assessment are retained. Ultimately, through refined fault concealment operations, the efficiency and accuracy of the mineral resource assessment process are improved, providing a scientific basis for actual geological exploration and mineral resource development. Especially under complex geological conditions, by removing unnecessary inactive faults, resource distribution prediction is optimized.

[0175] Example 4

[0176] To further illustrate the analysis process of the above formula, the analysis process is explained in detail using data from P12, P13, and P14 after standardization and normalization. The point data for P12, P13, and P14 are shown in Table 3.

[0177] Table 3. Detailed data of points P12, P13, and P14

[0178]

[0179] Meanwhile, the preset parameters are: α4 = 5, f fa,min =0.3, β3=4, D TCA,max =1;

[0180] First, calculate the formulas respectively. part:

[0181] (1) P12:

[0182] (2) P13:

[0183] (3) P14:

[0184] Next calculation part:

[0185] (1) P12:

[0186] (2) P13:

[0187] (3) P14:

[0188] Again, H on pages 12, 13, and 14. fault (P) are 0.6711*0.931=0.624, 0.0681*0.55=0.0375, and 0.1712*0.802=0.1373 respectively;

[0189] Finally, a judgment and subsequent processing are performed: when the preset hiding threshold is 0.5, H fault If (P) is greater than 0.5, the area is considered an inactive fault and is treated with a transparent effect using segmented rendering technology.

[0190] For P12: H fault (P) = 0.624 > 0.5, indicating that this area is an inactive fault and should be hidden using transparent processing;

[0191] For P13: H fault (P) = 0.0375 < 0.5, indicating that this area is not an inactive fault;

[0192] For P14: H fault (P) = 0.1373 < 0.5, indicating that the area is not an inactive fault.

[0193] Actual effect:

[0194] By applying the technical methods specifically described in Examples 1 to 4 above, the accuracy of the generated three-dimensional model in assessing coal seams is improved by more than 20%, the data processing efficiency is improved by 30%, and subsequent mineral resource prediction and development planning receive more precise guidance.

[0195] It should be noted that in step five, model calibration and optimization includes the following steps:

[0196] Step 51, Correcting the lithological layers and fault surfaces in the model: Based on the geological profile data, adjust the three-dimensional lithological layers and fault surfaces. Adjust the boundaries, thickness, and spatial distribution of the lithological layers in the three-dimensional model, especially for the contact surfaces between different lithologies. Use interpolation to fill in the data gaps to ensure the continuity of the lithological layers throughout the region. Adjust each lithological layer in the three-dimensional model according to the changes in lithology in the actual profile, especially for abrupt changes, to ensure that the lithological layers conform to the geological structural constraints in three-dimensional space. Use the location information of fault lines in the geological profile data and geological maps to obtain the strike, dip angle, and displacement of faults. Adjust the fault surfaces in the three-dimensional model according to the location and properties of the actual fault surfaces to ensure that they conform to the actual structural characteristics. For known active faults, emphasize the influence of their displacement. Use geometric constraints (such as plane equations) to correct the fault surfaces to ensure that the shape of the fault surfaces is consistent with the actual profile data.

[0197] Step 52, Correcting rock property data based on geophysical inversion results: The seismic velocity model obtained based on reflected wave data is used to estimate underground lithology and porosity. The density distribution of underground materials is inverted through geophysical measurement data. The underground electrical parameters obtained from electromagnetic wave detection results are used. The seismic velocity model and geophysical inversion results are used to compare the inverted data with the original property data through regression analysis to obtain an optimized rock property model.

[0198] Step 53, Verify the corrected 3D geological model: Compare the corrected 3D geological model with the actual geological profile, strictly check the accuracy of lithological distribution and fault location, calculate the error of the model before and after correction, including but not limited to lithological distribution error, fault location error, and physical property data error, and use standard error for evaluation.

[0199] Step five involves precisely adjusting the lithological layers and fault surfaces in the 3D model by combining actual geological profile data and geological map information, ensuring that the model's geological structure is highly consistent with the actual situation. Interpolation methods are used to fill data gaps, adjust lithological layer boundaries and thicknesses, and correct the strike, dip, and displacement of fault surfaces, optimizing the model's continuity and accuracy. Particular emphasis is placed on handling abrupt change zones and active faults, ensuring the model accurately reflects geological changes. This process improves the model's accuracy and reliability, effectively avoiding the impact of model errors on mineral resource assessment, and providing a more scientific and reliable foundation for further mining exploration and development planning.

[0200] In summary, this invention integrates 3D implicit modeling, machine learning optimization, and geological constraint correction. By combining geological information databases and geophysical data, it utilizes machine learning methods to extract geological structural constraints, constructs a 3D structural model, and achieves lithology and fault surface modeling through 2D geoscientific surface models and fault boundary constraints. An unsupervised clustering algorithm is used for preliminary classification of geological bodies, and a supervised classification algorithm is combined with borehole data to further optimize the model. This overcomes the challenge of lithological class imbalance and improves the accuracy and efficiency of modeling. During the modeling process, the algorithm extracts lithological distribution, ore body characteristics, and fault information from known borehole data and geological profiles, ensuring that the model structure matches the actual geological conditions. Geophysical inversion and statistical laws of rock properties, combined with human-computer interactive interpretation, effectively correct deviations in the lithological model. By establishing 3D lithological layers and fault surfaces, non-target coal seams, non-coal seams, mined areas, and inactive faults are hidden based on preset formulas. The system automatically judges and processes the data that needs to be hidden, reducing the complexity of manual intervention and data processing. It clearly displays the underground structure and resource distribution of the mining area, providing a reliable basis for quantitative prediction of mineral resources.

[0201] like Figures 6 to 8 The image shown is a partial screenshot of various interfaces of the platform on which the method proposed in this invention is applied. Figure 6 The image shows a screenshot of the algorithm control interface. The left side displays the parameter setting box for the algorithm applied by the method. Figure 6 The right side displays the model training progress and classification results. Figure 7 The text describes how data sources and preprocessing are presented in the data processing, displaying information about data source acquisition and line graphs previewing the data. This allows staff to easily view the data. Figure 8 The system further expands upon the application of this method by showcasing lithology, geological age, strata distribution, and fault analysis, facilitating a more detailed understanding of the specific conditions of each geological profile and aiding in the analysis of the geological conditions of the study area.

[0202] exist Figure 9The document showcases the interface for model control on a platform using a 3D geological profile modeling method proposed in this invention. Specifically, it demonstrates controls for model rotation, scaling, and view movement, enabling 3D rotation and view movement of the model, scaling it to an appropriate size, and displaying front, side, and top views to present richer geological profile information in a 2D format. The stratigraphic display allows for adjustment of stratigraphic transparency and displays stratigraphic labels. Layer display control is implemented through step 13 of the method proposed in this invention, and the document uses four formulas to display thresholds for non-target coal seams, irrelevant rock strata, mined areas, and inactive faults. This facilitates the creation of 3D lithological layers and fault surfaces, and the automatic identification and processing of data requiring concealment based on preset formulas. The bottom of the 3D geological profile model also displays specific stratigraphic information, including stratigraphic depth, dip angle, and thickness. The specific 3D geological profile model is obtained by modeling based on actual acquired data. Figure 9 This only shows the interface for controlling layers using the values ​​calculated from the four formulas involved in the three-dimensional modeling method for geological profiles proposed according to this invention.

[0203] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0204] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A three-dimensional modeling method for geological profiles, characterized in that, Includes the following steps: Step 1, Data Acquisition and Preprocessing: Combining geological information data source databases and geophysical data, machine learning methods are used to extract geological structural constraints, and human-computer interactive interpretation is performed by combining two-dimensional geoscientific information and geophysical information gradient zones. Step 2, Geological Model Constraints: Using geological structure data from a two-dimensional geoscience database, a three-dimensional geological model is constructed using linear model constraints. This model models the lithology, combines stratigraphic data and physical properties, and employs a surface model. Planar geological contact surfaces are established using lithological data, and fracture surfaces are constrained using fracture boundaries on geological profiles. Fractured sections that do not affect the structure are removed during the visualization stage. Step 3, Implicit Modeling and Machine Learning Optimization: Unsupervised clustering algorithm is used to initially classify geological bodies into hierarchical categories, automatically monitor the spatial distribution of lithology categories, and supervised classification algorithm is used in combination with labeled borehole data to further improve the classification model; Step 4, 3D Modeling and Graphics Rendering Optimization: Voxel technology is used to create volumetric rendering of underground lithology. Volumetric data represents the transitions and contact relationships between different underground rock layers. Color mapping and transparency settings are used to visualize different lithological layers. An algorithm based on viewpoint selection and geometric visibility is used to remove identified parts that need to be hidden during the 3D image generation process. The requirements for hiding include the hiding of non-target coal seams, the hiding of rock layers related to non-coal seams, the hiding of mined areas, and the hiding of inactive faults. Numerical judgments are made using the formulas for judging the hiding of non-target coal seams, the hiding of rock layers related to non-coal seams, the hiding of mined areas, and the hiding of inactive faults, respectively. The formula for judging the hiding of non-target coal seams is based on the quality and depth of the coal seam. A Gaussian distribution is used to measure the depth difference of the coal seam, and a weighted activation function is used to process the difference between the coal seam depth and the maximum depth to analyze the mineability of the coal seam. The formula for determining the hidden demand of non-coal seam related rock strata is based on the type, thickness, and location of the rock strata to determine whether they are related to non-coal seam research. It uses a threshold to determine whether they are unrelated to coal seam research and further dynamically adjusts the hidden demand judgment based on the rock strata thickness and distance from the research area. The formula for determining the hidden demand of mined areas is based on the depth of the mined area, the mining progress, and the relative position of the mined area to the current research area. The formula for determining the hidden demand of inactive faults is based on the activity, location, and magnitude of the fault to determine whether these faults should be hidden. Step 5, Model calibration and optimization: Adjust the three-dimensional lithological layers and fault surfaces of the model according to the actual geological profile data to ensure that they conform to the geological conditions. Based on the geophysical inversion results, calibrate the rock physical property data in the model. Step Six, Dynamic Interaction and Analysis: Using virtual reality technology, dynamic interaction is performed on the three-dimensional geological model. Through user-interactive control, geological profiles at different depths can be viewed in real time, and the distribution information of ore bodies can be analyzed.

2. The three-dimensional modeling method for geological profiles according to claim 1, characterized in that, In step four, the removal of the identified parts that need to be hidden during the 3D image generation process is achieved in the following way: Step 11, Calculate the normal vector: Obtain the unit vector of the obtained vector by taking the cross product of the two vectors formed by the key points of the stratigraphic profile on each geological profile and its two nearest neighboring points; Step 12, generate shadows: Based on the above key points, determine whether there is a shadow by the angle between the light source and the normal vector of the key point. If the angle is 0, there is no shadow. If the angle is not zero, there is a shadow. The maximum value between the dot product of the normal vector of the key point and the unit vector from the light source to the key point and 0 is used as the illumination intensity of the key point. Step 13, Remove hidden parts: For the above key points, if the distance from the key point to the surrounding set area is less than a preset threshold, the set area is regarded as the geological area that needs to be hidden, and the geological area that needs to be hidden is removed. Step 14, Generate Rendering: Combining the shadow intensity with the information that needs to be hidden in Step 13, generate a visualized geological profile rendering.

3. The three-dimensional modeling method for geological profiles according to claim 2, characterized in that, The formula for judging the hidden demand of relevant rock strata in non-coal seam research is as follows: exist hour, ; In dissatisfaction hour, ; In the formula: These are lithological analysis values. The numerical values ​​are representative of the coal seam and are obtained through preset methods. For the thickness of the rock strata, This represents the maximum thickness of the rock strata. This is the value of hidden demand analysis for related rock strata in non-coal seam studies. Let P be the coal seam quality at point P. Let P be the maximum depth of the coal seam at point P. The distance between the rock strata and the study area. This is the preset minimum distance between the rock strata and the study area. The influencing factor on the coal seam depth at point P is... The distance between the rock strata and the study area is a regulating factor. It is the sigma function; If the value exceeds a preset threshold, the geological area is considered a non-coal seam study area, and the non-coal seam study area is removed using a region segmentation algorithm.

4. The three-dimensional modeling method for geological profiles according to claim 2, characterized in that, The formula for determining hidden demand in already exploited areas is: ； In the formula: The hidden demand assessment value for the already mined area at point P is given. This represents the mining progress of the coal seam where point P is located. This represents the minimum threshold for displaying coal seam mining progress. This is a correction factor for the coal seam mining progress. This represents the distance between the mined area of ​​the coal seam where point P is located and the study area. This is a correction factor to show the impact of the distance between the mined area and the study area. The parameter representing the influence of the distance between the mined area and the study area is used to control the impact. If the value exceeds a preset threshold, the geological area is considered a mining area, and a masking technique is used to cover the mining area.

5. A three-dimensional modeling method for geological profiles according to claim 1, characterized in that, The data acquisition and preprocessing process in step one includes: Step 21, Data Acquisition: Obtain geological profile data, geological maps and geological structure maps, groundwater and soil information, and production exploration data from geological information data. Geological profile data includes borehole data, stratigraphic thickness, lithological information, and mineral composition. Geological maps and geological structure maps include fault location, folds, and lithological distribution data. Groundwater and soil information includes hydrogeology and groundwater flow. Production exploration data includes mining information and ore body analysis information. Obtain gravity data, magnetic data, electromagnetic data, and seismic data from geophysical databases. Among them, seismic data includes lithology and fault analysis information analyzed through seismic wave velocity. Step 22, Data Storage and Integration: Establish a unified data repository, use a geographic information system to store and manage spatial data, integrate data from different sources, ensure spatial consistency between data, and convert them into a unified format.

6. The three-dimensional modeling method for geological profiles according to claim 1, characterized in that, The specific implementation process for step two, which involves constraining the geological model, is as follows: Step 31, Feature Extraction and Structural Constraints: Extract the stratum thickness, lithological information, location, strike and dip angle of geological faults, mineral composition and spatial distribution based on geological maps and profiles. Extract data related to gravity anomalies and underground density distribution, magnetic anomalies and rock magnetism, seismic wave velocity and lithology, geological structure, and electrical conductivity based on gravity, magnetic, and seismic wave data. Combine geological profiles and geophysical data, constrain the spatial distribution of different lithological regions based on the physical characteristics of different rock layers and seismic wave and gravity anomaly data. Infer the geometric shape of faults based on their strike and dip angle and seismic reflection data. Construct the geometry of the stratigraphic contact surface based on mineral composition, lithology, and fault information using seismic wave reflection data. Step 32, Human-computer interactive explanation: Use Web GIS to provide geological experts with an intuitive data viewing and editing interface, supporting experts to annotate, edit and confirm unclear areas of data, optimize the machine learning model based on expert feedback, and update the data source and model of the machine learning model based on the new annotations added during the human-computer interaction process; Step 33, Data Integration and Output: Merge geological and geophysical data, integrate all constraints based on machine learning models and expert feedback, and form the final input data.

7. A three-dimensional modeling method for geological profiles according to claim 1, characterized in that, In step three, the execution flow for implicit modeling and machine learning optimization includes: Step 41, Unsupervised algorithm for preliminary hierarchical classification: The K-means clustering algorithm is used to divide the data points into K clusters, where the center of each cluster is the average value of the data points. Geological profile data, lithological data, and geophysical data are input into the clustering algorithm. The clustering algorithm automatically discovers the hierarchical structure of the geological bodies and outputs the cluster identifier of each geological body point, thus performing preliminary hierarchical classification. Step 42: Further optimize the model using a supervised classification algorithm: Combine the labeled borehole data, select features that affect lithology classification, train a random forest model using the labeled borehole data to enable the model to learn the lithology classification pattern, evaluate the model performance using cross-validation, adjust the model's hyperparameters based on accuracy and recall, optimize the classification results, and output the classification results of the geological body based on the probability values ​​output by the classification model. Step 43, combine unsupervised and supervised classification results: combine the preliminary results of unsupervised classification with the classification results of supervised learning, and use a weighted average method to obtain the classification result for each geological body.

Images