Geological profile map processing method based on cross-hole CT (Computed Tomography) and drilling information multi-dimensional coordination

Abstract

The invention discloses a geological profile map processing method based on cross-hole CT and drilling information multi-dimensional cooperation, and relates to the technical field of intelligent processing of engineering geological image achievements. Cross-hole CT data can provide continuous distribution information of rock-soil bodies among drill holes, and lithology distribution can be depicted more finely in combination with drilling data; the advantage that cross-hole CT can detect and recognize inter-hole rock-soil body structures is combined with the advantage that drilling can obtain different-depth rock core distribution, the characteristics of the inter-hole rock-soil body structures obtained through the cross-hole CT are combined with artificial intelligence, intelligent correction of engineering geology profile maps with multi-dimensional cooperation of the cross-hole CT and drilling information is achieved, and the engineering geology profile maps with the multi-dimensional cooperation of the cross-hole CT and the drilling information are obtained. The corrected engineering geological profile map is finer and closer to the real situation, and has better reference value for engineering design and engineering decision; the risk of emergencies in the construction process can be reduced through accurate geological information, the requirement for the number of drilled holes can be reduced, and therefore the exploration cost is reduced, and the exploration time is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent processing of engineering geological image results, and in particular to a method for processing geological section drawings based on multi-dimensional collaboration of cross-hole CT and borehole information. Background Art

[0002] A borehole refers to a cylindrical hole formed by using special drilling equipment to drill into the underground rock and soil mass on the ground according to the predetermined position, direction, angle and depth requirements. This method can obtain the cores of rock and soil masses at different depths underground, which is the basis for drawing traditional engineering geological section drawings. Based on the distribution and related properties of cores at different depths, and on this basis, reasonably infer and extend the area between boreholes, and finally draw an engineering geological section drawing.

[0003] Currently, the method for drawing engineering geological section drawings is to infer and extend the area between boreholes based on the distribution of borehole cores. This means that the structure of the rock and soil mass between boreholes is inferred rather than the real rock and soil mass structure. Especially in the limestone formation, the development of karst caves varies in shape, and there may be a large difference between the inferred rock and soil mass structure and the real rock and soil mass structure.

[0004] The cross-hole CT method is a geophysical exploration method that uses the differences in the propagation characteristics of elastic waves or electromagnetic waves in the medium to image and analyze the geological body between two boreholes. This method can obtain the absorption characteristics of the rock and soil mass medium between boreholes for elastic waves or electromagnetic waves, and then infer the structural characteristics of the rock and soil mass between boreholes. However, this method cannot identify the rock properties and has certain errors.

[0005] Therefore, there is an urgent need to design an intelligent correction method for engineering geological section drawings that can realize multi-dimensional collaboration of cross-hole CT and borehole information to solve the technical problem of insufficient fineness of engineering geological section drawings. Summary of the Invention

[0006] In order to solve the technical problem of intelligent correction of engineering geological section drawings with multi-dimensional collaboration of cross-hole CT and borehole information, the present invention provides a method for processing geological section drawings based on multi-dimensional collaboration of cross-hole CT and borehole information. The following technical solutions are adopted:

[0007] A method for processing geological section drawings based on multi-dimensional collaboration of cross-hole CT and borehole information includes the following steps:

[0008] Step 1, obtaining the graphic results of implementing the cross-hole CT method between boreholes;

[0009] Step 2, matrixing the graphic results, extracting one or more columns of data according to the set horizontal length, and numbering them according to the horizontal length;

[0010] Step 3, obtaining the core histograms of Borehole 1 and Borehole 2;

[0011] Step 4: Map the 0th data of the cross-hole CT data to the depth parameters and lithology parameters of Borehole 1, and map the last data of the cross-hole CT data to the depth parameters and lithology parameters of Borehole 2 to obtain mapping rule data;

[0012] Step 5: Segment the remaining columns of the cross-hole CT data by depth, and take the average value of each segment based on data mutation;

[0013] Step 6: Apply the mapping rule obtained in Step 4 to the remaining columns of the cross-hole CT data to convert the cross-hole CT data into corresponding lithology and depth parameters;

[0014] Step 7: Submit the lithology and depth parameters corresponding to all cross-hole CT column data to the trained artificial intelligence model to generate an engineering geological profile.

[0015] By adopting the above technical solution and combining with the cross-hole CT data, it can more accurately reflect the geotechnical structure between boreholes and reduce the speculation error.

[0016] The cross-hole CT data can provide continuous distribution information of the geotechnical body between boreholes. Combining with the borehole data, the lithology distribution can be described more precisely, such as karst caves, faults, etc.

[0017] The cross-hole CT data can identify special geological bodies that are difficult to obtain through boreholes between boreholes, such as small-scale karst caves, fractures, etc., providing important references for engineering design and construction.

[0018] Combine the advantage of the cross-hole CT in detecting and identifying the geotechnical structure between boreholes with the advantage of boreholes in obtaining the core distribution at different depths. Utilize the characteristics of the geotechnical structure between boreholes obtained by the cross-hole CT and combine with artificial intelligence to realize the intelligent correction of the engineering geological profile with multi-dimensional collaboration of cross-hole CT and borehole information. The corrected engineering geological profile is more refined and closer to the actual situation, having better reference value for engineering design and engineering decision-making;

[0019] Accurate geological information can reduce the risk of unexpected situations during construction, such as encountering karst caves, faults, etc., thus ensuring construction safety. It can reduce the demand for the number of boreholes, thereby reducing exploration costs and time.

[0020] Optionally, in Step 1, the graphic results of implementing the cross-hole CT method between boreholes are obtained by inversion through existing commercial software, and the graphic results are the contour map of the apparent elastic wave velocity distribution.

[0021] By adopting the above technical solution, by using the cross-hole CT method to obtain the contour map of the apparent elastic wave velocity distribution and combining with the borehole information, this technical solution can significantly improve the accuracy of the engineering geological profile, reduce engineering risks, improve engineering efficiency, save engineering costs, and promote the development of geological exploration technology, having important application value. Geophysical inversion is a core technology that reversely deduces the spatial distribution of physical property parameters of underground media from surface-observed geophysical field data (including but not limited to gravity field, electromagnetic field, seismic wave field) through a systematic mathematical modeling method. Its essence is to solve the constrained optimization problem of a non-linear ill-conditioned equation system with multiple solutions. There are many cross-hole CT inversion algorithms involved in this article, such as the algebraic reconstruction method (ART), the simultaneous iterative reconstruction technique (SIRT), the conjugate gradient method (CGT), the back-projection method (BPT), etc.

[0022] Optionally, the specific method of data matrixization in step 2 is as follows:

[0023] 1. The case where the original grid data is not lost. The CT result map is often generated by interpolating the original XYZ data into grid data and then drawing the contour result map. If the grid data can be obtained, then this grid data is the data matrix of the result, containing evenly distributed coordinate points and data information within the profile range.

[0024] 2. Importing picture data and generating a data matrix

[0025] If the original grid data is lost and only picture files are available, the picture can be matrixized according to the following steps:

[0026] (1) Scale conversion: If the image contains a scale (such as marked "1:5000"), determine the conversion relationship between pixels and the actual distance according to the scale (for example: 1 pixel = 0.05 meters).

[0027] (2) Automatic edge detection for extracting key geological interfaces: Identify continuous curves such as stratigraphic boundaries and fault lines through image algorithms. For fuzzy boundaries, adopt adaptive threshold segmentation technology to distinguish different physical property regions, facilitating the acquisition of physical property information at key points in step 4 spatial sampling.

[0028] (3) Generating a coordinate system mapping table to establish a two-dimensional coordinate system: Assume that the upper left corner of the image is the origin (0, 0), the horizontal direction is the X-axis (distance), and the vertical direction is the Y-axis (depth / elevation).

[0029] (4)Uniform spatial point sampling: First, perform grid coverage and generate uniform grid points on the section at a specified interval (e.g., every 0.1 m horizontally and every 0.1 m vertically). A dynamic encryption strategy can be selected to automatically increase the sampling density in areas with drastic lithological changes (such as near faults). Then, perform coordinate back-calculation and annotation, and assign real coordinates (X, Y) to each grid point according to the preset grid interval and coordinate mapping relationship. For example: A certain point is located at the 120th pixel (horizontally) × 300th pixel (vertically) in the image → converted to real coordinates (6 m, 15 m).

[0030] (5)If the image contains a color scale (such as color corresponding to wave velocity value), extract the wave velocity data information through the color value - parameter look-up table, and finally generate a CSV / Excel file containing three columns of data: X coordinate, Y coordinate, and attribute value (such as wave velocity) to obtain a data matrix.

[0031] Optionally, the specific method for obtaining the mapping rule data in step 4 includes the following steps:

[0032] Step 41: According to the depth section of the lithology distribution of the same rock stratum in the borehole columnar diagram, find the inversion result values of the corresponding CT results of the borehole in the depth section and calculate the average value to obtain the mapping between the CT result values of the corresponding lithology of the borehole and the lithology.

[0033] Step 42: Calculate all lithological parameters and their corresponding CT values in the borehole columnar diagram in sequence to obtain the mapping rule of the lithology.

[0034] Step 43: Obtain the mapping rule data of the cross-hole CT data corresponding to two boreholes.

[0035] By adopting the above technical effects, the same rock stratum in the borehole columnar diagram refers to the rock strata with the same lithology and close depths in the borehole columnar diagram. Mainly relying on experience, 30% of the hole spacing can be taken as the threshold. If the depth difference is less than this value, it is considered the same rock stratum.

[0036] Rock strata with the same lithology and close depths in the borehole columnar diagram, such as the karst cave itself and the overlying limestone, sandstone, covering layer (soil layer), etc.;

[0037] The mapping rule is based on the direct correspondence between the borehole core columnar diagram and the CT data, avoiding the subjectivity and uncertainty in artificial lithology judgment and improving the accuracy of lithology identification. The mapping rule can quickly convert CT data into lithology information and improve the efficiency of lithology identification. The mapping rule can provide more refined lithology distribution information, such as lithology changes at different depths and positions, so as to more accurately reflect the true situation of the rock and soil mass between boreholes.

[0038] Mapping rules can help identify and characterize complex geological structures, such as faults, fractures, karst caves, etc., providing more comprehensive geological information for engineering design and construction.

[0039] Optionally, the depth segmentation method in step 5 is an algorithm with the ability to identify data mutation points, and the algorithm with the ability to identify data mutation points is the moving average method or the first-order difference method.

[0040] By adopting the above technical solution, the extracted column data is segmented in depth. The method is to use an algorithm with the ability to identify data mutation points, and the algorithm with the ability to identify data mutation points is the moving average method or the first-order difference method. This method can identify the numerical mutation characteristics to segment the column data, and then calculate the average value of each segment after segmentation to inversely obtain the lithology parameters in combination with the mapping relationship.

[0041] Optionally, the artificial intelligence model in step 7 is based on a graph neural network architecture.

[0042] By adopting the above technical solution, the artificial intelligence model architecture is designed as follows:

[0043] Input layer: 1.1 Data source, structured tabular data: location (X), depth range (Zmin, Zmax), physical property parameters (wave velocity, electromagnetic wave absorption coefficient, etc.), lithology label. It contains multiple data points, and each location X corresponds to a data point.

[0044] 1.2 Preprocessing, depth interval parsing: Convert the depth range in string format (such as "10 - 20") into a numerical pair (Z_top, Z_bottom). Feature standardization: Normalize the physical property parameters (such as Min - Max or Z - Score). Spatial encoding: Map the position coordinates (X, 0) to nodes in a grid or graph structure.

[0045] Core model layer:

[0046] 2.1 Lithology boundary prediction module:

[0047] 1) Model selection:

[0048] Graph neural network (GNN): Model the spatial dependence relationship between data points (such as the lithology continuity of adjacent survey lines). The neighborhood aggregation formula is as follows:

[0049]

[0050] vi: The feature vector of data point i, which includes the position (Xi), depth interval (Ztop,i, Zbottom,i), and physical property parameters. N(i): The set of spatial neighbors of data point i. hi: The hidden representation of data point i after GNN aggregation, which is the fusion result of spatial features and physical property parameters. It converts the discrete data point data into continuous geological structure information and provides key input for subsequent lithology classification and profile generation. This process simulates the thinking mode of geologists "inferring strata by combining surrounding boreholes" but is automated through data-driven methods.

[0051] Spatial dependence modeling: By aggregating the information of neighbor nodes {vj}, the spatial correlation between data points is captured. For example, the lithology between adjacent data points on the same survey line is usually continuous (e.g., a gradual change from sandstone to shale). The aggregation method is as follows:

[0052] Mean pooling:

[0053] Attention mechanism:

[0054] W: Trainable weight matrix.

[0055] If the physical property of data point i is extremely high, but all its surrounding neighbors are sandstone, GNN will reduce the probability of misclassifying this point as limestone.

[0056] Bidirectional LSTM / Transformer: Captures the vertical variation pattern of lithology along the depth direction (Z-axis).

[0057]

[0058] Zi: The depth value of data point i (e.g., taking Ztop,i or the midpoint depth). sz: The depth sequence hidden state output by LSTM, which encodes the lithology change pattern along the depth direction. Vertical sequence modeling: LSTM processes data points in depth order (from top to bottom or from bottom to top) to capture the vertical lithology pattern.

[0059] 2) Input: The feature vector [X, Z_top, Z_bottom, physical property parameters] of each data point.

[0060] Output: The probability distribution of lithology labels P(L|X, Z).

[0061] Probabilistic classification: Maps the hidden state sz output by LSTM to the probability distribution of lithology labels. Multi-class support: Softmax ensures that the sum of all lithology probabilities is 1.

[0062] ;

[0063] W: Trainable weight matrix. b: Bias vector. P(Li = l): Probability that data point i belongs to lithology l.

[0064] Through the above process, the model can reconstruct a continuous and reasonable geological profile from discrete data point data, significantly outperforming traditional threshold segmentation or interpolation methods.

[0065] 2.2 Lithology Region Continuity Modeling Module:

[0066] Objective: Connect discrete prediction results into continuous lithology regions and solve fault and noise problems.

[0067] Method: Conditional Random Field (CRF): Constrain the lithology consistency of adjacent nodes through an energy function. Morphological post-processing: Use dilation / erosion operations to smooth the boundaries.

[0068] Energy function:

[0069]

[0070] Where: Is the single-point lithology probability (from GNN-LSTM). Penalize lithology mutations at adjacent points (i, j).

[0071] Profile generation layer:

[0072] Complete polygon construction and filling. First, perform top and bottom boundary extraction: For each lithology label, extract continuous top boundaries (Z_top) and bottom boundaries (Z_bottom) along the survey line position (X). Second, perform closed area generation: Connect the top and bottom boundary points to form a polygon vertex sequence. Then, perform anti-aliasing: Use Bezier curves to smooth the boundaries. Finally, generate a unique filling color or filling pattern based on the lithology hash value (e.g., sandstone → yellow, shale → gray)

[0073] Training and optimization strategy:

[0074] Loss function:

[0075]

[0076] Classification loss: Cross-Entropy loss is used to optimize lithology prediction:

[0077] Boundary smoothing loss: The CRF energy term constrains spatial continuity

[0078] Distributed training: Use Horovod or PyTorch Distributed to accelerate the training of large-scale survey data.

[0079] Advantages:

[0080] Multi-modal fusion, which simultaneously utilizes physical property parameters and spatial topological relationships to improve the prediction accuracy of complex geological structures.

[0081] End-to-end automation, where the entire process from raw tabular data to a visualized profile requires no manual intervention.

[0082] Dynamic adaptability, which is compatible with heterogeneous geological phenomena such as faults and pinch-outs through CRF and morphological processing.

[0083] A storage medium for storing a processing program designed using a geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information.

[0084] A computer that runs a processing program designed using a geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information and outputs an intelligent corrected engineering geological profile.

[0085] In summary, the present invention includes at least one of the following beneficial technical effects:

[0086] The present invention can provide a geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information. Cross-hole CT data can provide continuous distribution information of the rock and soil mass between boreholes. Combining with borehole data, the lithology distribution can be described more precisely.

[0087] Cross-hole CT data can identify special geological bodies that are difficult to obtain through boreholes between boreholes.

[0088] Combining the advantage of cross-hole CT in detecting and identifying the rock and soil mass structure between holes with the advantage of boreholes in obtaining the core distribution at different depths, using the characteristics of the rock and soil mass structure between holes obtained by cross-hole CT and combining with artificial intelligence, to achieve intelligent correction of the engineering geological profile with multi-dimensional collaboration of cross-hole CT and borehole information. The corrected engineering geological profile is more precise and closer to the actual situation, having better reference value for engineering design and engineering decision-making.

[0089] Accurate geological information can reduce the risk of unexpected situations during construction, such as encountering karst caves, faults, etc., thus ensuring construction safety. It can reduce the demand for the number of boreholes, thereby reducing exploration costs and time. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 is a schematic flowchart of the geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information of the present invention;

[0091] Figure 2(a) is a borehole columnar diagram of borehole ZK01 in a specific embodiment of the present invention;

[0092] Figure 2(b) is a borehole columnar diagram of borehole ZK02 in a specific embodiment of the present invention;

[0093] Figure 3 It is a schematic diagram of a traditional engineering geological profile in a specific embodiment of the present invention;

[0094] Figure 4 It is a schematic diagram of the result of cross-hole seismic wave CT tomography in a specific embodiment of the present invention;

[0095] Figure 5 It is a schematic diagram of the numerical matrix of the cross-hole CT result in a specific embodiment of the present invention;

[0096] Figure 6 It is a schematic diagram of the corrected engineering geological profile in a specific embodiment of the present invention. Specific embodiments

[0097] The present invention will be further described in detail below with reference to the accompanying drawings.

[0098] The embodiment of the present invention discloses a method for processing geological section diagrams based on multi-dimensional collaboration of cross-hole CT and borehole information.

[0099] Referring to Figures 1-6 , Embodiment 1, a method for processing geological section diagrams based on multi-dimensional collaboration of cross-hole CT and borehole information, includes the following steps:

[0100] Step 1, obtaining the graphic results of implementing the cross-hole CT method between boreholes;

[0101] Step 2, matrixing the graphic results into data, extracting one or more columns of data according to a set horizontal length, and numbering them according to the horizontal length;

[0102] Step 3, obtaining the core columnar diagrams of Borehole 1 and Borehole 2;

[0103] Step 4, mapping the 0th data of the cross-hole CT data to the depth parameters and lithology parameters of Borehole 1, and mapping the last data of the cross-hole CT data to the depth parameters and lithology parameters of Borehole 2 to obtain mapping rule data;

[0104] Step 5, performing depth segmentation on the remaining columns of the cross-hole CT data, taking data mutation as the principle and calculating the average value of each segment;

[0105] Step 6, applying the upper mapping rule obtained in Step 4 to the remaining columns of the cross-hole CT data, and converting the cross-hole CT data into corresponding lithology and depth parameters;

[0106] Step 7, submitting the lithology and depth parameters corresponding to all cross-hole CT column data to a trained artificial intelligence model to generate an engineering geological profile diagram.

[0107] Combined with cross-hole CT data, it can more accurately reflect the rock and soil mass structure between boreholes and reduce the speculation error.

[0108] Cross-hole CT data can provide continuous distribution information of the rock and soil mass between boreholes. Combining with borehole data, the lithology distribution can be characterized more precisely, such as karst caves, faults, etc.

[0109] Cross-hole CT data can identify special geological bodies that are difficult to obtain through boreholes between boreholes, such as small-scale karst caves, fractures, etc., providing important references for engineering design and construction.

[0110] Combining the advantage of cross-hole CT in detecting and identifying the structure of the rock and soil mass between boreholes with the advantage of boreholes in obtaining the core distribution at different depths, using the characteristics of the structure of the rock and soil mass between boreholes obtained by cross-hole CT, and combining with artificial intelligence, realizing the intelligent correction of the engineering geological profile with multi-dimensional collaboration of cross-hole CT and borehole information. The corrected engineering geological profile is more precise and closer to the actual situation, having better reference value for engineering design and engineering decision-making;

[0111] Accurate geological information can reduce the risk of unexpected situations during construction, such as encountering karst caves, faults, etc., thus ensuring construction safety. It can reduce the demand for the number of boreholes, thereby reducing exploration costs and time.

[0112] In Example 2, in step 1, the graphic results of implementing the cross-hole CT method between boreholes are obtained by inversion through existing commercial software. The graphic results are the contour map of the apparent elastic wave velocity distribution.

[0113] By using the cross-hole CT method to obtain the contour map of the apparent elastic wave velocity distribution and combining with borehole information, this technical solution can significantly improve the accuracy of the engineering geological profile, reduce engineering risks, improve engineering efficiency, save engineering costs, and promote the development of geological exploration technology, having important application value.

[0114] Geophysical inversion is a core technology that reversely deduces the spatial distribution of physical property parameters of underground media through a systematic mathematical modeling method from surface-observed geophysical field data (including but not limited to gravitational field, electromagnetic field, seismic wave field). Its essence is to solve the constrained optimization problem of a non-linear ill-posed equation system with multiple solutions. There are many cross-hole CT inversion algorithms involved in this article, such as the algebraic reconstruction method (ART), the simultaneous iterative reconstruction technique (SIRT), the conjugate gradient method (CGT), the back-projection method (BPT), etc.

[0115] Geophysical forward modeling refers to simulating the geophysical field responses (such as gravity anomalies, seismic wave fields, electromagnetic signals, etc.) observed on the surface or in wells based on known underground physical parameters (such as density, velocity, resistivity, etc.) through physical laws and mathematical models. It is one of the core tools for geophysical exploration and research, providing a theoretical basis for data interpretation, inversion, and geological modeling.

[0116] Taking the conjugate gradient method (CGT) as an example, the specific implementation steps are as follows:

[0117] Data preparation: Collect the observation data of cross-hole electromagnetic waves, including the positions of the emission points and receiving points, and information such as the travel time and amplitude of the electromagnetic waves.

[0118] Initial model setting: Set the initial parameter distribution of the underground medium, such as a uniform model or a model based on prior information.

[0119] Forward simulation: Use the current model to calculate the predicted data, such as the travel time or attenuation of each ray.

[0120] Calculate the residual: The difference Δd between the observation data and the predicted data.

[0121] Construct the Jacobian matrix J: Calculate the sensitivity of each model parameter to the data of each ray.

[0122] Construct the inversion equation, consider the regularization term, and form a linear system.

[0123] Apply the conjugate gradient method to solve for the model update Δm.

[0124] Update the model: m = m + Δm.

[0125] Judge convergence: Check whether the residual is small enough, or whether the number of iterations reaches the preset value. If not converged, return to step 3.

[0126] Output the final inversion result.

[0127] The above steps are implemented by mature commercial software. For example, the electromagnetic wave CT multi-pair hole joint inversion software V2.1; the electromagnetic wave CT inversion software V2.6.

[0128] Example 3, the specific method of data matrixization in step 2 is:

[0129] 1. The case where the original grid data is not lost. The CT result map is often generated by interpolating the original XYZ data into grid data, and then drawing the contour result map. If the grid data can be obtained, then this grid data is the data matrix of the result, containing evenly distributed coordinate points and data information within the profile range.

[0130] 2. Import the picture data and generate the data matrix

[0131] If the original grid data is lost and only the picture file remains, the picture can be matrixized according to the following steps:

[0132] (1) Perform scale conversion: If the image contains a scale (such as marked "1:5000"), determine the conversion relationship between pixels and the actual distance according to the scale (for example: 1 pixel = 0.05 meters).

[0133] (2)Automated edge detection for extracting key geological interfaces: Identify continuous curves such as formation boundaries and fault lines through image algorithms. For fuzzy boundaries, adopt adaptive threshold segmentation technology to distinguish different physical property regions, facilitating the acquisition of physical property information at key points during the spatial sampling in Step 4.

[0134] (3)Generate a coordinate system mapping table and establish a two-dimensional coordinate system: Assume that the upper left corner of the image is the origin (0, 0), the horizontal direction is the X-axis (distance), and the vertical direction is the Y-axis (depth / elevation).

[0135] (4)Uniform spatial point sampling; First, perform grid coverage, generating uniform grid points on the section at a specified spacing (e.g., every 0.1 meter horizontally and every 0.1 meter vertically). A dynamic encryption strategy can be selected to automatically increase the sampling density in areas with drastic lithological changes (such as near faults). Then, perform coordinate back-calculation and annotation. According to the preset grid spacing and coordinate mapping relationship, assign real coordinates (X, Y) to each grid point. For example: A certain point is located at the 120th pixel (horizontally) × 300th pixel (vertically) in the image → converted to real coordinates (6 meters, 15 meters).

[0136] (5)If the image contains a color scale (such as color corresponding to wave velocity values), extract wave velocity data information through a color value - parameter comparison table, and finally generate a CSV / Excel file containing three columns of data: X coordinate, Y coordinate, and attribute value (such as wave velocity) to obtain a data matrix.

[0137] In Example 4, the specific method for obtaining the mapping rule data in Step 4 includes the following steps:

[0138] Step 41, According to the depth section of the lithology distribution of the same rock layer in the borehole histogram, find the inversion result values of the corresponding CT results of the borehole in the depth section and calculate the average value to obtain the mapping between the CT result values of the corresponding lithology of the borehole and the lithology;

[0139] Step 42, Calculate all lithology parameters and their corresponding CT values in the borehole histogram in sequence to obtain the mapping rule of the lithology;

[0140] Step 43, Obtain the mapping rule data of the cross-hole CT data corresponding to two boreholes.

[0141] The same rock layer in the borehole histogram refers to the rock layers with the same lithology and close depths in the borehole histogram. Mainly relying on experience, 30% of the hole spacing can be taken as the threshold. If the depth difference is less than this value, it is considered the same rock layer.

[0142] The mapping rule is based on the direct correspondence between the borehole core columnar diagram and CT data, avoiding the subjectivity and uncertainty in lithology judgment by humans and improving the accuracy of lithology identification. The mapping rule can quickly convert CT data into lithology information, improving the efficiency of lithology identification. The mapping rule can provide more detailed lithology distribution information, such as lithology changes at different depths and positions, thus more accurately reflecting the true situation of the geotechnical body between boreholes.

[0143] The mapping rule can help identify and characterize complex geological structures, such as faults, fractures, karst caves, etc., providing more comprehensive geological information for engineering design and construction.

[0144] In Example 5, the depth segmentation method in step 5 is to adopt an algorithm with the ability to identify data mutation points, and the algorithm with the ability to identify data mutation points is the moving average method or the first-order difference method.

[0145] The extracted column data is segmented by depth. The method is to adopt an algorithm with the ability to identify data mutation points. The algorithm with the ability to identify data mutation points is the moving average method or the first-order difference method. This method can identify the numerical mutation characteristics to segment the column data, and then calculate the average value of each segment after segmentation to inversely obtain the lithology parameters in combination with the mapping relationship.

[0146] In Example 6, the artificial intelligence model in step 7 is based on the graph neural network architecture.

[0147] The architecture design of the artificial intelligence model is as follows:

[0148] Input layer: 1.1 Data source, structured tabular data: location (X), depth range (Zmin, Zmax), physical property parameters (wave velocity, electromagnetic wave absorption coefficient, etc.), lithology label. It contains multiple data points, and each location X corresponds to a data point.

[0149] 1.2 Preprocessing, depth interval parsing: Convert the depth range in string format (such as "10 - 20") into a numerical pair (Z_top, Z_bottom). Feature standardization: Normalize the physical property parameters (such as Min - Max or Z - Score). Spatial encoding: Map the position coordinates (X, 0) to nodes in a grid or graph structure.

[0150] Core model layer:

[0151] 2.1 Lithology boundary prediction module:

[0152] 1) Model selection:

[0153] Graph neural network (GNN): Model the spatial dependence relationship between data points (such as the lithology continuity of adjacent survey lines). The neighborhood aggregation formula is as follows:

[0154]

[0155] vi: The feature vector of data point i, which includes the location (Xi), the depth interval (Ztop,i, Zbottom,i), and the physical property parameters. N(i): The set of spatial neighbors of data point i. hi: The hidden representation of data point i after GNN aggregation, which is the fusion result of spatial features and physical property parameters. It converts the discrete data point data into continuous geological structure information and provides key input for subsequent lithology classification and profile generation. This process simulates the thinking mode of geologists "inferring strata by combining surrounding boreholes" but realizes automation through data driving.

[0156] Spatial dependence modeling: Capture the spatial correlation between data points by aggregating the information of neighbor nodes {vj}. For example, the lithology between adjacent data points on the same survey line is usually continuous (such as the gradual change from sandstone to shale). The aggregation method is as follows:

[0157] Mean pooling:

[0158] Attention mechanism:

[0159] W: Trainable weight matrix.

[0160] If the physical property of data point i is extremely high, but all its surrounding neighbors are sandstone, GNN will reduce the probability of misclassifying this point as limestone.

[0161] Bidirectional LSTM / Transformer: Capture the vertical lithology change pattern along the depth direction (Z-axis).

[0162]

[0163] Zi: The depth value of data point i (such as taking Ztop,i or the midpoint depth). sz: The depth sequence hidden state output by LSTM, which encodes the lithology change pattern along the depth direction. Vertical sequence modeling: LSTM processes data points in depth order (from top to bottom or from bottom to top) to capture the vertical lithology pattern.

[0164] 2) Input: The feature vector [X, Z_top, Z_bottom, physical property parameters] of each data point.

[0165] Output: The probability distribution P(L|X, Z) of lithology labels.

[0166] Probabilistic classification: Map the hidden state sz output by LSTM to the probability distribution of lithology labels. Multi-class support: Softmax ensures that the sum of all lithology probabilities is 1.

[0167] ;

[0168] W: Trainable weight matrix. b: Bias vector. P(Li = l): Probability that data point i belongs to lithology l.

[0169] Through the above process, the model can reconstruct a continuous and reasonable geological profile from discrete data points, significantly outperforming traditional threshold segmentation or interpolation methods.

[0170] 2.2 Lithology Region Continuity Modeling Module:

[0171] Objective: Connect discrete prediction results into continuous lithology regions and solve fault and noise problems.

[0172] Method: Conditional Random Field (CRF): Constrain the lithology consistency of adjacent nodes through an energy function. Morphological post-processing: Use dilation / erosion operations to smooth the boundaries.

[0173] Energy function:

[0174]

[0175] Where: Is the single-point lithology probability (from GNN-LSTM). Punish the lithology mutation of adjacent points (i, j).

[0176] Profile generation layer:

[0177] Complete polygon construction and filling. First, extract the top and bottom boundaries: For each lithology label, extract the continuous top boundary (Z_top) and bottom boundary (Z_bottom) along the survey line position (X). Second, generate closed regions: Connect the top and bottom boundary points to form a polygon vertex sequence. Then, perform anti-aliasing processing: Use Bezier curves to smooth the boundaries. Finally, generate a unique filling color or filling pattern based on the lithology hash value (e.g., sandstone → yellow, shale → gray)

[0178] Training and optimization strategy:

[0179] Loss function:

[0180]

[0181] Classification loss: Cross-Entropy loss optimizes lithology prediction:

[0182] Boundary smoothing loss: CRF energy term constrains spatial continuity

[0183] Distributed training: Use Horovod or PyTorchDistributed to accelerate large-scale survey data training.

[0184] Advantages:

[0185] Multi-modal fusion, which simultaneously utilizes physical property parameters and spatial topological relationships to improve the prediction accuracy of complex geological structures.

[0186] End-to-end automation, where the entire process from raw tabular data to visual profiles requires no manual intervention.

[0187] Dynamic adaptability, which is compatible with heterogeneous geological phenomena such as faults and pinch-outs through CRF and morphological processing.

[0188] Example 7, A storage medium for storing a processing program designed using the geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information.

[0189] Example 8, A computer that runs a processing program designed using the geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information and outputs an intelligent corrected engineering geological profile.

[0190] The following uses specific examples to illustrate the implementation principle of the geological profile processing method based on the multi-dimensional collaboration of cross-hole CT and borehole information of the present invention:

[0191] There are two limestone area boreholes ZK01 and ZK02. The columnar diagrams are shown in Figures 2(a) and 2(b). The distance between the two holes is 5.4 m. Cross-hole seismic wave CT method was implemented between the holes. The detection target position was from elevation -20 m to the bottom of the boreholes. The contour results of the apparent elastic wave velocity distribution of the rock and soil mass structure between the boreholes were obtained by inversion. The graphic results are shown in Figure 4 .

[0192] Without combining the cross-hole CT results, the schematic diagram of the traditional engineering geological profile under the conditions of this pair of boreholes is shown in Figure 3 :

[0193] The cross-hole elastic wave CT results are numerically matrixed, as shown in Figure 5, the horizontal length with a fixed value is taken as 1m. The horizontal length of 0m corresponds to borehole ZK01, and the horizontal length of 5.4m corresponds to borehole ZK02. As shown in the columnar section of ZK01, the lithology from -20m to -25.51m is semi-rock and semi-soil strongly weathered mudstone, and the average value of the data from -20m to 25.51m corresponding to the elastic wave CT results is 2522m / s. The lithology from -25.51m to -29.01m is extremely fractured moderately weathered limestone, and the average value of the data from -25.51m to -29.01m corresponding to the elastic wave CT results is 3275m / s. The lithology from -29.01m to -31.51m is a fully filled karst cave, and the filling is silty clay with a small amount of karst erosion fragments. The average value of the data from -29.01m to -31.51m corresponding to the elastic wave CT results is 2728m / s. The lithology from -31.51m to -37.51m is moderately fractured moderately weathered limestone, and the average value of the data from -31.51m to -37.51m corresponding to the elastic wave CT results is 3320m / s. As shown in the columnar section of ZK02, the lithology from -20m to -26.71m is semi-rock and semi-soil strongly weathered mudstone, and the average value of the data from -20m to -26.71m corresponding to the elastic wave CT results is 2586m / s. The lithology from -26.71m to -27.81m is extremely fractured moderately weathered limestone, and the average value of the data from -26.71m to -27.81m corresponding to the elastic wave CT results is 3501m / s. The lithology from -27.81m to -37.21m is beaded karst caves, and the filling is silty clay and some gravel. The average value of the data from -27.81m to -37.21m corresponding to the elastic wave CT results is 2858m / s. The lithology from -37.21m to -43.21m is moderately fractured moderately weathered limestone, and the average value of the data from -37.21m to -43.21m corresponding to the elastic wave CT results is 3386m / s. The summary mapping rules are shown in Table 1. Since the distance between the boreholes is relatively close, it can be considered that the mapping rules between the lithology and the CT data are similar. The depth and apparent velocity data of the corresponding elastic wave CT are extracted at a unit length of 1m, with a total of 5 groups of data for horizontal lengths of 1m, 2m, 3m, 4m, and 5m. The moving average method is used to identify the apparent velocity mutation points of these 5 groups of data, segmented according to the mutation point depth, and the average value of the apparent velocity of the elastic wave CT results corresponding to the corresponding depth segments is extracted. Then, lithology mapping is carried out according to the parameter mapping rules between the borehole lithology and the CT apparent velocity. The detailed parameters are shown in Table 2.

[0194] Table 1 Mapping Table of Lithology and Cross-hole Elastic Wave CT Data

[0195]

[0196] Table 2 Lithology Parameter Table for Mapping Cross-hole Elastic Wave CT Data

[0197]

[0198] Thus, the lithological information of the cross-section between boreholes with a spacing of 1 m is obtained. Submitting this information to the trained artificial intelligence model enables the drawing of a more refined schematic diagram of the engineering geological section. The resulting diagram is as shown in Figure 6 the following figure.

[0199] There are obvious differences between the revised engineering geological section diagram and the traditional one in terms of the karst cave representation between boreholes. The revised engineering geological section diagram is more refined and closer to the actual situation, providing better reference value for engineering design and engineering decision-making.

[0200] The above are all preferred embodiments of the present invention, and do not limit the protection scope of the present invention. Therefore, any equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for processing geological section maps based on multi-dimensional collaboration of cross-hole CT and borehole information, characterized in that It includes the following steps: Step 1: Obtain the graphic results of the cross-hole CT method implemented between drill holes. Step 2: Matrixize the graphic results, extract one or more columns of data according to the set horizontal length, and number them according to the horizontal length. Step 3: Obtain the core columnar diagrams of borehole 1 and borehole 2. Step 4: Map the 0th data of the cross-hole CT data to the depth parameters and lithology parameters of borehole 1, and map the last data of the cross-hole CT data to the depth parameters and lithology parameters of borehole 2 to obtain mapping rule data. Step 5: Segment the depth of the remaining columns of the cross-hole CT data, and take the data mutation as the principle and calculate the average value of each segment. Step 6: Apply the upper mapping rule obtained in Step 4 to the remaining columns of the cross-hole CT data, and convert the cross-hole CT data into corresponding lithology and depth parameters. Step 7: Submit the lithology and depth parameters corresponding to all cross-hole CT column data to the trained artificial intelligence model to generate an engineering geological profile.

2. The method for processing geological section maps based on multi-dimensional collaboration of cross-hole CT and borehole information according to claim 1, wherein: In Step 1, the graphic results of the cross-hole CT method implemented between drill holes are obtained by inversion using existing commercial software, and the graphic results are contour maps of elastic wave velocity distribution.

3. The method for processing a geological section diagram based on multi-dimensional collaboration of cross-hole CT and borehole information according to claim 1, wherein, The specific method for obtaining the mapping rule data in Step 4 includes the following steps: Step 41: According to the depth section of the lithology distribution of the same rock layer in the borehole columnar diagram, find the inversion result values of the corresponding CT results of the borehole in the depth section and calculate the average value to obtain the mapping of the CT result value and lithology of the corresponding lithology of the borehole. Step 42: Calculate all lithology parameters and their corresponding CT values in the borehole columnar diagram in sequence to obtain the mapping rule of the lithology. Step 43: Obtain the mapping rule data of the cross-hole CT data corresponding to the two boreholes.

4. The geological section processing method based on multi-dimensional collaboration of cross-hole CT and borehole information according to claim 1, characterized in that The depth segmentation method in Step 5 is to use an algorithm with the ability to identify data mutation points. The algorithm with the ability to identify data mutation points is the moving average method or the first-order difference method.

5. The geological section processing method based on multi-dimensional collaboration of cross-hole CT and borehole information according to claim 1, wherein The artificial intelligence model in Step 7 is based on a graph neural network architecture.

6. A storage medium, characterized in that, For storing the processing program designed by the geological profile processing method based on multi-dimensional collaboration of cross-hole CT and borehole information described in any one of claims 1-5.

7. A computer, characterized in that, The computer runs the processing program designed by the geological profile processing method based on multi-dimensional collaboration of cross-hole CT and borehole information described in any one of claims 1-5, and outputs an intelligent corrected engineering geological profile.

Images