Method and system for improving precision of transparent geologic model based on tunneling navigation data elements

Abstract

The invention belongs to the technical field of intelligent mines, and discloses a method and system for improving the precision of a transparent geologic model based on tunneling navigation data elements. According to the method, multi-source data acquisition of tunneling navigation data and advanced detection geological data is carried out, the acquired multi-source data is preprocessed, data formats of different sources are converted into a unified database format, and data storage is carried out; multi-source data fusion is completed; modeling a three-dimensional dynamic geologic model based on a triangulated irregular network-generalized triangular prism model; and iteratively optimizing the obtained three-dimensional dynamic geologic model to obtain a three-dimensional dynamic transparent geologic model, and outputting and visualizing the three-dimensional dynamic transparent geologic model. Through multi-source data fusion and dynamic modeling, the spatial precision and the real-time response capability of the model are improved, and more reliable geological information support is provided for rapid tunneling of a coal mine.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent mines, and in particular relates to a method and system for improving the accuracy of a transparent geological model based on excavation navigation data elements. Background Art

[0002] In the field of smart mines, transparent geological models are one of the core technologies for realizing intelligent and precise mining of coal mines. Their accuracy directly affects tunneling path planning, safety risk assessment and resource mining efficiency. It can intuitively present the undulating morphology of coal seams, providing key support for coal mine tunneling, resource planning and safety management. At present, most geological models rely on static geological exploration data (such as realistic data, drilling data, and logging data). These data have the disadvantages of long collection cycles, low spatial resolution, and poor real-time performance, which makes it difficult for geological models to accurately reflect the actual geological conditions. The application of tunneling navigation technology provides real-time and refined data support for geological modeling. Navigation equipment (such as lidar, inertial navigation system, geophysical exploration during tunneling, etc.) can obtain multi-dimensional data including tunneling paths, coal (rock) layer properties, location information, etc.

[0003] However, the existing technology lacks effective data fusion and real-time modeling methods, resulting in the failure of tunneling data to fully play its role. Summary of the invention

[0004] In order to overcome the problems existing in the related technology, the disclosed embodiment of the present invention provides a method and system for improving the accuracy of a transparent geological model based on tunneling navigation data elements, specifically relates to a method for improving the accuracy of a transparent geological model based on tunneling navigation data elements, which is suitable for scenarios such as coal mine tunneling.

[0005] The technical solution is as follows: A method for improving the accuracy of a transparent geological model based on excavation navigation data elements, comprising: S1, collect multi-source data of tunneling navigation data and advanced geological data, where tunneling navigation data includes spatial location information, hidden disaster-causing parameters, dynamic tunneling data, and collection frequency, and advanced geological data includes drilling data, geophysical data, and historical geological data; S2, pre-processing the collected multi-source data, converting the data formats from different sources into a unified database format, and storing the data in the database; S3, calling the pre-processed multi-source data, using data vectorization, multi-source data weight allocation, multi-source data vector space fusion and vectorized fusion multi-source data to return to the weight allocation model space to complete multi-source data fusion; S4, based on the multi-source data fusion results, three-dimensional dynamic geological model building is carried out using the irregular triangulated network-generalized triangular prism model; S5, iteratively optimizing the obtained three-dimensional dynamic geological model, obtaining a three-dimensional dynamic transparent geological model for output and visualization.

[0006] In step S2, the collected multi-source data is preprocessed, including: data cleaning and denoising; the collected excavation navigation data is cleaned and denoised, and outliers and noise data are removed. Transformation is used to attenuate the underground environmental noise. The expression is: ; In the formula, is the attenuation value of the underground environmental noise, It is the navigation signal for tunneling at all times. is the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Renault denoising value; For output data After noise interference attenuation, use The inverse transform returns the noise-attenuated data.

[0007] In step S2, the data formats from different sources are converted into a unified database format and the data is stored in the database; including: Data coordinate alignment: coordinate conversion and alignment of tunneling navigation data and traditional geological data, so that the data are expressed in a unified geological coordinate system, and spatial registration is performed using control points; Data format unification: convert point cloud, raster, and vector data formats from different sources into a unified database format; The data storage includes: using data extraction, conversion and loading ETL method, according to a unified data conversion rule set, to achieve the structural conversion of the business system database to the geological database; the data extraction, conversion and loading ETL method adopts two methods: active data push and passive data extraction; using the database connection tool PostgreSQL to establish a connection with the database, constructing corresponding data insertion statements according to the database table structure and data format, inserting the data into the database, and verifying the data integrity and accuracy after storage.

[0008] In step S3, data vectorization includes: vectorizing the acquired multi-source data and mapping them to the same vector space; extracting image features through texture analysis and edge detection for image data and then vectorizing them; extracting features through Mel frequency cepstral coefficients MFCC and spectrum features for sound data and then vectorizing them; The weight allocation of multi-source data includes: using the weight allocation model to fuse the multi-source data of tunneling navigation data and advanced geological data, and the expression is: ; In the formula, For the The calculated weights of the input elements, is the maximum weight distribution function, is a linear calculation function, For the The preprocessed value of the input elements.

[0009] In step S3, vector space fusion of multi-source data includes: After the weights are redistributed, data fusion is performed and the expression is: ; In the formula, The data output after fusion is calculated by adopting new weights; The vectorized fusion of multi-source data returns to the weight distribution model space, including: returning the vectorized fused data to the weight distribution model space, the expression is: ; In the formula, For elements converted to model space, is the conversion factor, is the equalization coefficient.

[0010] In step S4, a three-dimensional dynamic geological model is built using an irregular triangulated network-generalized triangular prism model, including: Step 1: Encrypt the tunneling cloud data using a smooth discrete interpolation algorithm; Step 2: For the point cloud data with the same number, the irregular triangulated network model of each geological interface is generated by point-by-point insertion method to describe the changes and characteristics of the interface; Step 3: Based on the irregular triangulated network model, the stratigraphic information is searched in a top-down order; the initial generalized triangular prism model is generated by comparing and connecting the stratigraphic information between each group of three point cloud data; Step 4: For areas with complex geological conditions, the initial generalized triangular prism model is expanded into a smaller generalized triangular prism model, or degenerated into a tetrahedron model and a pyramid model, so that the initial generalized triangular prism model or tetrahedron model and pyramid model are consistent with the actual geological phenomena; Step 5: By matching and fusing the triangular face elements in the irregular triangulated network model with the geological volume elements in the generalized triangular prism model, a hybrid space model of the irregular triangulated network and the generalized triangular prism model is constructed; Step 6: Verify and optimize the constructed irregular triangulated network-generalized triangular prism model hybrid model; Step 7: Use the irregular triangulated network-generalized triangular prism model hybrid model to perform three-dimensional geological geometry modeling, and perform data analysis and mining at the same time to obtain useful information from the irregular triangulated network-generalized triangular prism model hybrid model.

[0011] In step 5, by matching and fusing the triangular face elements in the irregular triangulated network model with the geological volume elements in the generalized triangular prism model, an irregular triangulated network-generalized triangular prism model hybrid space model is constructed, including: (1) Input the triangular face elements in the irregular triangulated network model and sort them into the triangulated network domain. Let the triangular face elements be , Indicates the coordinates of the center point; (2) Select the first triangulated channel set in the data arrangement , Indicates the sampling point number of a single channel record. Indicates the survey line number in the triangulated channel concentration, and selects the adjacent excavation data according to the center point, and calculates the refractive coefficient sequence of the underground medium of the center point excavation at different offset distances based on the triangulated network diffraction equation ; (3) Establishing a geological volume element model in a generalized triangular prism model whose main frequency information is consistent with the triangulated network channel set : (I) Perform spectrum analysis on the zero-offset tunneling channel of the triangulated network channel set to obtain the main frequency of the triangulated network diffraction wave of the channel set ; (II) In the fusion simulation, the zero-phase Ricker wavelet is used as the geological body diffraction wavelet in the triangular prism model, so the main frequency of the triangular network diffraction wave Substituting into the Ricker wavelet formula, the expression is: ; In the formula, for The geological volume element model in the generalized triangular prism model of time, The main frequency of the diffraction wave containing the triangular network The Ricker wavelet, For time; About Time Discrete the model to obtain the geological volume element model in the generalized triangular prism model ; (4) Based on the geological volume element model obtained in the generalized triangular prism model and refractive index To do fusion, the expression is: ; In the formula, is the geological channel set, for After time discretization The geological volume element model in the generalized triangular prism model from time to initial value; Fusion simulation of geological channel set in generalized triangular prism model ; (5) The geological channel set in the simulated generalized triangular prism model With the input triangulation channel set For fusion correlation analysis, the expression is: ; In the formula, is the depth position correlation value of the deformation interface, For the time; When calculating the fusion correlation, the correlation is calculated according to the depth position of the deformation interface. Computational The range is truncated to obtain the correlation of each event axis at different offsets; (6) Set the threshold variable Q to 0.2 and calculate the correlation of each in-phase axis. The tunneling channel numbers below the threshold are marked and defined as distortion points; from this point, the event axis is severely distorted, so the data after each event axis distortion point is removed to achieve dynamic correction stretching distortion removal processing for the triangulated network channel set and the geological body channel set in the generalized triangular prism model; (7) Change the center point and repeat steps (2) to (6) until distortion removal is achieved for all triangulated channel sets and geological body channel sets in the generalized triangular prism model, and finally obtain an irregular triangulated network-generalized triangular prism model hybrid space model.

[0012] In step (2), the refractive index sequence of the underground medium at different offset distances at the center point is calculated based on the triangulated diffraction equation. ,include: (a) Determine the depth position of the underground deformation interface at the center point based on the excavation geological layer data and depth gauge in the excavation data , Display interface Depth (b) Convert the density curve and time difference curve in the tunneling data to obtain the density curve , P-wave velocity curve and shear wave velocity curve, and according to the depth position of the deformation interface Calculate the deformation parameters of the media on both sides of each interface and obtain the center point deformation parameter sequence ; (c) Substitute the deformation parameter sequence into the triangulated diffraction equation to calculate the refractive index of each deformation interface , represents the incident angle, at this time It reflects the relationship between the refractive index and the angle of incidence; (d) According to the line step length in the triangulation network Convert the refraction angle to Reflecting the change relationship with the offset distance, the conversion formula is: ; In the formula, is the transformation value of the incident angle with the offset distance, is the interval distance; Thus, the refractive index sequence of the central point underground medium at different offset distances is obtained. , the triangulated diffraction equation is as follows: ; In the formula, are the first incident angle, the second incident angle, are the first longitudinal wave velocity, the second longitudinal wave velocity, are the first refraction angle, the second refraction angle, are the first shear wave density and the second shear wave density, are the first shear wave velocity and the second shear wave velocity, They are respectively the refractive index of the deformation interface between longitudinal wave and longitudinal wave, and the refractive index of the deformation interface between longitudinal wave and transverse wave.

[0013] In step S5, the three-dimensional dynamic geological model is iteratively optimized, including: During the excavation process, the model update is triggered every 10m or when encountering faults, collapse columns, and unfavorable geological bodies in the scour zone. The iterative formula is expressed as: ; In the formula, For the next position The model status; is a function that represents how the model is updated according to the current position and geological conditions during the excavation process; For the current position The model status; is the interval distance; To trigger updates when encountering bad geological bodies; The acquisition of the three-dimensional dynamic transparent geological model includes: model cutting and slicing, data mapping and conversion, output generation and optimization; Among them, model cutting and slicing: cutting or slicing the three-dimensional dynamic geological model to extract data from the excavation area; Data mapping and conversion: Map the 3D dynamic geological model data into the data structure of the target output format, and perform conversion and processing, including color mapping, texture mapping and lighting model application; Output generation and optimization: Generate the corresponding output file, i.e., 3D dynamic transparent geological model, according to the selected output format, and perform optimization processing such as compression and format adjustment; The visualization includes: extracting isosurfaces through MC algorithm and performing volume rendering through ray casting and texture mapping algorithm, which are used to render dynamic transparent geological models; and using cluster analysis to identify and extract similar geological features, including low-resistance anomalies in front of the tunneling.

[0014] Another object of the present invention is to provide a system for improving the accuracy of a transparent geological model based on tunneling navigation data elements, the system implementing the method for improving the accuracy of a transparent geological model based on tunneling navigation data elements, the system comprising: The multi-source data acquisition module is used for multi-source data acquisition of tunneling navigation data and advanced geological data. The tunneling navigation data includes spatial location information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency. The advanced geological data includes drilling data, geophysical data, and historical geological data. The data preprocessing and warehousing module is used to preprocess the collected multi-source data, convert the data formats from different sources into a unified database format, and store the data in the database; A multi-source data fusion module is used to call the pre-processed multi-source data, and complete the multi-source data fusion by using data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and vectorized fusion of multi-source data to return the weight allocation model space; The 3D dynamic geological modeling module is based on the multi-source data fusion results and uses the irregular triangulated network-generalized triangular prism model to perform 3D dynamic geological modeling; The three-dimensional dynamic transparent geological model acquisition and visualization module is used to iteratively optimize the obtained three-dimensional dynamic geological model and obtain the three-dimensional dynamic transparent geological model for output and visualization.

[0015] In combination with all the above-mentioned technical solutions, the beneficial effects of the present invention are as follows: the present invention improves the spatial accuracy and real-time response capability of the model through multi-source data fusion and dynamic modeling, and provides more reliable geological information support for rapid excavation of coal mines.

[0016] Real-time dynamic update: Utilize the real-time data generated during the excavation process to dynamically adjust the geological model and improve the adaptability to complex geological conditions.

[0017] Multi-source data fusion: Combine traditional geological data with real-time navigation data to achieve data complementarity and enhance the reliability and accuracy of the model.

[0018] High-resolution local modeling: Detailed modeling of key areas ahead of excavation to provide reliable and effective geological support for precise excavation.

[0019] Intelligent decision support: Through dynamic models and 3D visualization technology, it provides a scientific basis for tunneling path optimization and risk warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure; Figure 1 It is a flow chart of a method for improving the accuracy of a transparent geological model based on tunneling navigation data elements provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of a system for improving the accuracy of a transparent geological model based on excavation navigation data elements provided by an embodiment of the present invention; In the figure: 1. Multi-source data acquisition module; 2. Data preprocessing and storage module; 3. Multi-source data fusion module; 4. 3D dynamic geological model building module; 5. 3D dynamic transparent geological model acquisition and visualization module. DETAILED DESCRIPTION

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0022] Embodiment 1: The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements provided by the embodiment of the present invention includes: S1, collect multi-source data of tunneling navigation data and advanced geological data, where tunneling navigation data includes spatial location information, hidden disaster-causing parameters, dynamic tunneling data, and collection frequency, and advanced geological data includes drilling data, geophysical data, and historical geological data; S2, pre-processing the collected multi-source data, converting the data formats from different sources into a unified database format, and storing the data in the database; S3, calling the pre-processed multi-source data, using data vectorization, multi-source data weight allocation, multi-source data vector space fusion and vectorized fusion multi-source data to return to the weight allocation model space to complete multi-source data fusion; S4, based on the multi-source data fusion results, three-dimensional dynamic geological model building is carried out using the irregular triangulated network-generalized triangular prism model; S5, iteratively optimizing the obtained three-dimensional dynamic geological model, obtaining a three-dimensional dynamic transparent geological model for output and visualization.

[0023] Example 2, as another implementation of the present invention, an embodiment of the present invention provides a method for improving the accuracy of a transparent geological model based on tunneling navigation data elements; Step 1, data collection.

[0024] Before excavation, the front profile of the excavation is anticipated through drilling data and geophysical data (such as 3D seismic data). During excavation, the geological survey data of the tunnel is recorded, and the relevant parameters of the excavation equipment (location information, inclination, speed, lithology, etc. of the excavation path; tunnel morphology point cloud data from LiDAR; and the ups and downs and structural data of the coal seam ahead from geophysical exploration) are collected as the basic data source for model construction.

[0025] Exemplarily, data collection includes excavation navigation data and advance detection geological data; (1) Excavation navigation data.

[0026] The types of equipment used include laser radar, inertial navigation system, and seismic monitoring system during excavation. The collected content includes spatial location information, hidden disaster-causing parameters, dynamic excavation data, and collection frequency; among them, spatial location information includes the three-dimensional coordinates, inclination, azimuth, etc. of the excavation path. Hidden disaster-causing parameters include abnormal water accumulation areas and faults. Dynamic excavation data includes excavation speed, excavation distance, and equipment status. The collection frequency includes the equipment recording data at a frequency of 1 second / time to ensure the temporal and spatial continuity of the data.

[0027] (2) Advance detection of geological data.

[0028] Data types include drilling data, geophysical data (such as electrical method, magnetic method, triangulated diffraction wave detection data), and historical geological data. The main contents include drilling columnar diagrams and profiles; drilling columnar diagrams include lithology, coal (rock) layer distribution characteristics, etc. Profiles include the location of hidden anomalies, fault morphology, etc.

[0029] Step 2, data preprocessing, includes: 2.1, data cleaning and denoising; The equipment drift and environmental noise (such as vibration or interference from tunneling equipment) in the tunneling navigation data are corrected, that is, the collected tunneling navigation data is cleaned and denoised, and outliers and noise data are removed to ensure data quality. The denoising uses s transform to attenuate the underground environmental noise. The expression is: ; In the formula, is the attenuation value of the underground environmental noise, for The excavation navigation signal at all times, is the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Renault denoising value; For output data After noise interference attenuation, use The inverse transform returns the noise-attenuated data.

[0030] The tunneling navigation data and traditional geological data are converted and aligned to ensure that the data are represented in a unified geological coordinate system.

[0031] 2.2. Data coordinate alignment: Convert and align the excavation navigation data with the traditional geological data to ensure that the data is represented in a unified geological coordinate system.

[0032] Exemplarily, the local coordinate system of the tunneling navigation data is converted into a global coordinate system (such as the UTM coordinate system) that is consistent with traditional geological data.

[0033] Spatial registration using control points (such as known drill hole locations) ensures that the multi-source data alignment error is less than 0.1 m.

[0034] 2.3. Unify data formats: Convert data formats from different sources (such as point clouds, rasters, and vectors) into a unified database format to facilitate subsequent fusion and analysis.

[0035] Step 3: Data storage; The data extraction, transformation and loading (ETL) method is adopted to realize the structural conversion of the business system database to the geological database according to the unified data conversion rule set. The data extraction, transformation and loading (ETL) mode can adopt two methods: active data push and passive data extraction. Use the database connection tool PostgreSQL to establish a connection with the database, and construct the corresponding data insertion statement (such as INSERT INTO statement) according to the database table structure and data format to insert the data into the database. After entering the database, the data integrity and accuracy are verified, and constraints (Constraints) and triggers (Triggers) are used to verify the database integrity and data accuracy.

[0036] Step 4: Multi-source data fusion, including: Step 4.1 Data vectorization.

[0037] The multi-source data obtained in step 1 are vectorized and mapped to the same vector space. For common image data, the image features can be extracted through texture analysis and edge detection techniques, and then vectorized; for general sound data, the features can be extracted through Mel Frequency Cepstral Coefficients (MFCC) and spectral features, and then vectorized.

[0038] Step 4.2 Multi-source data weight allocation.

[0039] Using machine learning algorithms or weight distribution models, tunneling navigation data is integrated with multi-source data such as geological data. The expression is: ; In the formula, For the The calculated weights of the input elements, is the maximum weight distribution function, is a linear calculation function, For the The preprocessed value of the input elements Analyze the applicability of multi-source data. When the geological environment changes are complex, increase the weight of tunneling navigation data. When the geological conditions are simple, increase the weight of traditional geological data.

[0040] Step 4.3: Vector space fusion of multi-source data.

[0041] After the weights are redistributed, the expression for data fusion is: ; In the formula, The data output after fusion is calculated by adopting new weights; Step 4.4 vectorizes the fused multi-source data and returns it to the model space.

[0042] The vectorized fused data is returned to the machine learning algorithm or weight distribution model space through the following formula to provide fused data support for subsequent geological modeling.

[0043] ; In the formula, For elements converted to model space, is the conversion factor, is the equalization coefficient.

[0044] Step 5: Accurate three-dimensional dynamic geological model building.

[0045] 5.1, Three-dimensional dynamic geological model modeling algorithm.

[0046] The existing coal mine production geological data is used in combination with exploration drilling data, ground three-dimensional seismic detection and other data to build an overall three-dimensional geological model. Then, by dynamically collecting real-time dynamic data information such as while-drilling detection, while-digging detection, tunnel excavation exposure, while-mining detection, working face mining exposure, and while-falling detection generated during coal mine excavation, a transparent geological information database is built. After multi-dimensional data fusion processing, the Kriging interpolation algorithm is used to update the local three-dimensional geological model in front of the excavation in real time, and the local features of the three-dimensional geological model are dynamically corrected to achieve accurate and transparent reconstruction of the overall and local information of the tunnels, working faces, mining areas, etc. in the three-dimensional geological model, and obtain a three-dimensional dynamic geological model. Exemplarily, the three-dimensional dynamic geological model is constructed based on the irregular triangulated network-generalized triangular prism model, and the key steps include: Step 1: Encrypt the tunneling cloud data using a smooth discrete interpolation algorithm.

[0047] Step 2: For point cloud data with the same number, use the point-by-point insertion method to generate irregular triangulated network models of various geological interfaces to describe the changes and characteristics of the interfaces.

[0048] Step 3: Based on the irregular triangulated network model, search for stratigraphic information in a top-down order. Generate an initial generalized triangular prism model by comparing and connecting the stratigraphic information between each set of three point cloud data.

[0049] Step 4: For areas with complex geological conditions, the generalized triangular prism model is expanded into a smaller generalized triangular prism model, or degenerated into a tetrahedron model and a pyramid model to ensure that the model is consistent with the actual geological phenomena.

[0050] Step 5: By matching and integrating the triangular face elements in the irregular triangulated network model with the geological volume elements in the generalized triangular prism model, a hybrid space model of the irregular triangulated network and the generalized triangular prism model is constructed.

[0051] Exemplarily, in step 5, by corresponding and fusing the triangular face elements in the irregular triangulated network model with the geological volume elements in the generalized triangular prism model, the irregular triangulated network-generalized triangular prism model hybrid space model is constructed, including: (5.1) Input the triangular face element in the irregular triangulated network model and sort it into the triangulated network domain. Let the triangular face element be , Indicates the coordinates of the center point; (5.2) Select the first triangulated channel set in the data array , Indicates the sampling point number of a single channel record. Indicates the survey line number in the triangulated channel concentration, and selects the adjacent excavation data according to the center point, and calculates the refractive coefficient sequence of the underground medium of the center point excavation at different offset distances based on the triangulated network diffraction equation ,include: (a) Determine the depth position of the underground deformation interface at the center point based on the excavation geological layer data and depth gauge in the excavation data , Display interface Depth (b) Convert the density curve and time difference curve in the tunneling data to obtain the density curve , P-wave velocity curve and shear wave velocity curve, and according to the depth position of the deformation interface Calculate the deformation parameters of the media on both sides of each interface and obtain the center point deformation parameter sequence ; (c) Substitute the deformation parameter sequence into the triangulated diffraction equation to calculate the refractive index of each deformation interface , represents the incident angle, at this time It reflects the relationship between the refractive index and the angle of incidence; (d) According to the line step length in the triangulation network Convert the refraction angle to Reflecting the change relationship with the offset distance, the conversion formula is: ; In the formula, is the transformation value of the incident angle with the offset distance, is the interval distance; Thus, the refractive index sequence of the central point underground medium at different offset distances is obtained. , the triangulated diffraction equation is as follows: ; In the formula, are the first incident angle, the second incident angle, are the first longitudinal wave velocity, the second longitudinal wave velocity, are the first refraction angle, the second refraction angle, are the first shear wave density and the second shear wave density, are the first shear wave velocity and the second shear wave velocity, are the refractive index of the deformation interface between longitudinal wave and longitudinal wave, and the refractive index of the deformation interface between longitudinal wave and transverse wave respectively; (5.3) Establishing a geological volume element model in a generalized triangular prism model whose main frequency information is consistent with the triangulated network channel set : (I) Perform spectrum analysis on the zero-offset tunneling channel of the triangulated network channel set to obtain the main frequency of the triangulated network diffraction wave of the channel set ; (II) In the fusion simulation, the zero-phase Ricker wavelet is used as the geological body diffraction wavelet in the triangular prism model, so the main frequency of the triangular network diffraction wave Substituting into the Ricker wavelet formula, the expression is: ; In the formula, for The geological volume element model in the generalized triangular prism model of time, The main frequency of the diffraction wave containing the triangular network The Ricker wavelet, For time; About Time Discrete the model to obtain the geological volume element model in the generalized triangular prism model ; (5.4) According to the geological volume element model in the generalized triangular prism model and refractive index To do fusion, the expression is: ; In the formula, is the geological channel set, for After time discretization The geological volume element model in the generalized triangular prism model from time to initial value; Fusion simulation of geological channel set in generalized triangular prism model ; (5.5) The geological channel set in the simulated generalized triangular prism model is With the input triangulation channel set For fusion correlation analysis, the expression is: ; In the formula, is the depth position correlation value of the deformation interface, For the time; When calculating the fusion correlation, the correlation is calculated according to the depth position of the deformation interface. Computational The range is truncated to obtain the correlation of each event axis at different offsets; (5.6) Set the threshold variable Q to 0.2, and calculate the correlation of each phase axis The tunneling channel numbers below the threshold are marked and defined as distortion points; from this point, the event axis is severely distorted, so the data after each event axis distortion point is removed to achieve dynamic correction stretching distortion removal processing for the triangulated network channel set and the geological body channel set in the generalized triangular prism model; (5.7) Change the center point and repeat steps (5.2) to (5.6) until distortion removal is achieved for all triangulated network channel sets and geological body channel sets in the generalized triangular prism model, and finally an irregular triangulated network-generalized triangular prism model hybrid space model is obtained.

[0052] Step 6: Verify and optimize the constructed irregular triangulated network-generalized triangular prism model hybrid model to ensure that the model can accurately and efficiently describe the three-dimensional spatial data.

[0053] Step 7: Use the irregular triangulated network-generalized triangular prism model hybrid model to perform three-dimensional geological geometry modeling, and perform data analysis and mining at the same time to obtain useful information from the model.

[0054] 5.2, Dynamic feedback of abnormal geological bodies.

[0055] Take the fault as an example to illustrate the modeling method of the abnormal body. When modeling the fault, the "holistic method" is adopted to solve the fault data modeling problem. The specific process is as follows: (i) Calculate the fault plane; a fault consists of one or more fault points, each of which includes strike, dip, and drop. When only one fault point is recorded, extension data should also be included. The fault triangle surface can be calculated based on the attributes of the fault initial point; (ii) Calculate the coal-break intersection line; the intersection line obtained by intersecting the fault plane and the coal seam irregular triangulated network is the coal-break intersection line. The drop of the corresponding point on the coal-break intersection line is calculated based on the initial point drop of the fault and the distribution ratio; (iii) Segmentation of coal seam irregular triangulated network: the stratum is divided into the upper and lower wall regions of the fault according to the positive and reverse attributes of the fault, the intersection line is inserted, and the coal seam triangle within the range is segmented into upper and lower wall triangles; (iv) Calculate the coal seam dislocation; calculate the dislocation distance of each corner point of the upper and lower wall triangles in the XOY plane and the elevation change in the Z direction from the fault plane; (v) Calculation of influence domain: Calculate the influence domain of the fault based on the fall value of the fault coal intersection line. The default correction radius for each meter of fall is 10m. Calculate the Z coordinate change of the corner points of the irregular triangulated network triangles in the fault influence domain. (vi) Calculate the displacement of coal seams within the influence domain; calculate the change in the XOY plane position of the corner points of the irregular triangulated network of the strata within the influence domain of the normal fault.

[0056] Step 6: Iterative optimization of the three-dimensional dynamic geological model.

[0057] During the excavation process, the model update is triggered every 10m or when encountering unfavorable geological bodies such as faults, collapse columns, and scour zones, thereby improving the dynamic response capability of the model.

[0058] The basic form of the iteration formula can be expressed as: ; In the formula, For the next position The model status; is a function that represents how the model is updated during excavation according to the current position and geological conditions. Usually, this function can be a numerical model of geological engineering, taking into account the geometric model and properties. is the interval distance, here is the interval between each update; Geological Condition refers to the situation of encountering unfavorable geological bodies (such as faults, collapse columns, etc.), which triggers special update rules; For the current position The model status.

[0059] Another example is that the model is updated when the excavation advance distance reaches 10 meters or the navigation data changes significantly, such as when a fault is exposed. By entering the excavation navigation data, the model within the customized range of the newly added data is updated using the dynamic grid technology to meet the timely update of the mine stratigraphic model. The working face model can use the measurement data, inertial navigation data and excavation seismic data to quickly update the local or overall coal seam model in a dynamic manner to dynamically improve the accuracy of the coal seam model ahead of the mining. The basic resolution is 0.5 meters, and it is increased to 0.1 meters in geologically complex areas to capture more detailed changes in geological features. The latest data is loaded in real time and the model is reconstructed, while the historical model is retained as a reference.

[0060] Error analysis: Compare the model prediction results with the actual excavation data, and according to the actual excavation situation, modify the predicted geological model, and use this as a basis to analyze and predict the geological model of the unexcavated area. The process of model verification analysis and iterative optimization is as follows: complex three-dimensional geological body in a typical research site → original geological data → three-dimensional geological comprehensive database → initial three-dimensional dynamic geological model → accuracy assessment model → model accuracy visualization (three-dimensional spatial distribution model of geological structural uncertainty) → test whether the model accuracy meets the requirements → if not, correct the model error → corrected three-dimensional dynamic geological model → re-test whether the model accuracy meets the requirements, if not, correct the model error; if it meets the requirements, conduct actual engineering experimental verification in the research site → correct the accuracy assessment model → model prediction and actual engineering analysis and application.

[0061] Step 7: Output and visualization of the three-dimensional dynamic transparent geological model.

[0062] Generate dynamic transparent geological models by using data integration and real-time update capabilities, 3D dynamic geological modeling and visualization technology. This includes: real-time display through a 3D visualization platform, which allows decision makers to intuitively understand the geological conditions and optimize the excavation plan. Output dynamic transparent geological models by using 3D dynamic geological model slicing and data mapping technology, and achieve 3D visualization by using volume rendering using MC algorithms and ray casting.

[0063] Exemplarily, the three-dimensional dynamic transparent geological model output includes model cutting and slicing, data mapping and conversion, output generation and optimization; Among them, model cutting and slicing: cutting or slicing the three-dimensional dynamic geological model to extract data from the excavation area.

[0064] Data mapping and conversion: Map the 3D dynamic geological model data into the data structure of the target output format (.obj, .fbx, .shp, .dwg and other general formats), and perform necessary conversion and processing, such as color mapping, texture mapping and lighting model application.

[0065] Output generation and optimization: Generate the corresponding output file, i.e., the three-dimensional dynamic transparent geological model, according to the selected output format, and perform optimization processing such as compression and format adjustment to improve the readability and usability of the output file.

[0066] Visualization techniques: The MarchingCubes (MC) algorithm is used to extract isosurfaces and ray casting and texture mapping algorithms are used for volume rendering to render dynamic transparent geological models. Cluster analysis is also used to identify and extract similar geological features, such as low-resistance anomalies in front of the tunnel.

[0067] Embodiment 3, as Figure 2 As shown, a system for improving the accuracy of a transparent geological model based on excavation navigation data elements comprises: The multi-source data acquisition module 1 is used for multi-source data acquisition of excavation navigation data and advanced geological data, wherein the excavation navigation data includes spatial position information, hidden disaster-causing parameters, dynamic excavation data, and acquisition frequency, and the advanced geological data includes drilling data, geophysical data, and historical geological data; Data preprocessing and warehousing module 2 is used to preprocess the collected multi-source data, convert the data formats from different sources into a unified database format, and store the data in the database; The multi-source data fusion module 3 is used to call the pre-processed multi-source data, and complete the multi-source data fusion by using data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and vectorized fusion of multi-source data to return weight allocation model space; 3D dynamic geological model building module 4, based on the multi-source data fusion results, uses the irregular triangulated network-generalized triangular prism model to build a 3D dynamic geological model; The three-dimensional dynamic transparent geological model acquisition and visualization module 5 is used to iteratively optimize the obtained three-dimensional dynamic geological model, obtain the three-dimensional dynamic transparent geological model for output and visualization.

[0068] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for improving the accuracy of a transparent geological model based on tunneling navigation data elements, characterized in that: The method includes: S1, collect multi-source data of tunneling navigation data and advanced geological data, where tunneling navigation data includes spatial location information, hidden disaster-causing parameters, dynamic tunneling data, and collection frequency, and advanced geological data includes drilling data, geophysical data, and historical geological data; S2, pre-process the collected multi-source data, convert the data formats from different sources into a unified database format, and store the data in the database; S3, calling the pre-processed multi-source data, using data vectorization, multi-source data weight allocation, multi-source data vector space fusion and vectorized fusion multi-source data to return to the weight allocation model space to complete multi-source data fusion; S4, based on the multi-source data fusion results, three-dimensional dynamic geological model building is carried out using the irregular triangulated network-generalized triangular prism model; S5, iteratively optimizing the obtained three-dimensional dynamic geological model, obtaining a three-dimensional dynamic transparent geological model for output and visualization.

2. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 1 is characterized in that: In step S2, the collected multi-source data is preprocessed, including: cleaning and denoising the collected excavation navigation data, removing abnormal values ​​and noise data, and denoising by Transformation is used to attenuate the underground environmental noise. The expression is: ； In the formula, is the attenuation value of the underground environmental noise, for The excavation navigation signal at all times, is the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Renault denoising value; For output data After noise interference attenuation, use The inverse transform returns the noise-attenuated data.

3. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 1, characterized in that: In step S2, the data formats from different sources are converted into a unified database format and the data is stored in the database, including: Data coordinate alignment: coordinate conversion and alignment of tunneling navigation data and geological data, so that the data are expressed in a unified geological coordinate system, and spatial registration is performed using control points; Data format unification: convert point cloud, raster, and vector data formats from different sources into a unified database format; Data entry includes: using data extraction, conversion and loading ETL methods, according to a unified data conversion rule set, to achieve structural conversion from the business system database to the geological database; data extraction, conversion and loading ETL methods use two methods: active data push and passive data extraction; using the database connection tool PostgreSQL to establish a connection with the database, and constructing corresponding data insertion statements according to the database table structure and data format, inserting the data into the database, and verifying the data integrity and accuracy after entry.

4. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 1, characterized in that: In step S3, data vectorization includes: vectorizing the acquired multi-source data and mapping them to the same vector space; extracting image features through texture analysis and edge detection for image data, and then vectorizing; extracting features through Mel frequency cepstral coefficients MFCC and spectrum features for sound data, and then vectorizing; The weight allocation of multi-source data includes: using the weight allocation model to fuse the multi-source data of tunneling navigation data and advanced geological data, and the expression is: ； In the formula, For the The calculated weights of the input elements, is the maximum weight distribution function, is a linear calculation function, For the The preprocessed value of the input elements.

5. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 4 is characterized in that: In step S3, vector space fusion of multi-source data includes: After the weights are redistributed, data fusion is performed and the expression is: ； In the formula, The output data is fused after calculation using new weights; Vectorizing and fusing multi-source data to return to the weight distribution model space includes: returning the vectorized fused data to the weight distribution model space, and the expression is: ； In the formula, For elements converted to model space, is the conversion factor, is the equalization coefficient.

6. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 1, characterized in that: In step S4, a three-dimensional dynamic geological model is built using an irregular triangulated network-generalized triangular prism model, including: Step 1: Encrypt the tunneling cloud data using a smooth discrete interpolation algorithm; Step 2: For the point cloud data with the same number, the irregular triangulated network model of each geological interface is generated by point-by-point insertion method to describe the changes and characteristics of the interface; Step 3: Based on the irregular triangulated network model, the stratigraphic information is searched in a top-down order; the initial generalized triangular prism model is generated by comparing and connecting the stratigraphic information between each group of three point cloud data; Step 4: For areas with complex geological conditions, the initial generalized triangular prism model is expanded into a smaller generalized triangular prism model, or degenerated into a tetrahedron model and a pyramid model, so that the initial generalized triangular prism model or tetrahedron model and pyramid model are consistent with the actual geological phenomena; Step 5: correspond and merge the triangular face elements in the irregular triangulated network model with the geological volume elements in the generalized triangular prism model to construct a hybrid space model of the irregular triangulated network and the generalized triangular prism model; Step 6: Verify and optimize the constructed irregular triangulated network-generalized triangular prism model hybrid model; Step 7: Use the irregular triangulated network-generalized triangular prism model hybrid model to perform three-dimensional geological geometry modeling, and perform data analysis and mining at the same time to obtain useful information from the irregular triangulated network-generalized triangular prism model hybrid model.

7. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 6 is characterized in that: In step 5, the triangular face elements in the irregular triangulated network model are matched and merged with the geological volume elements in the generalized triangular prism model to construct an irregular triangulated network-generalized triangular prism model hybrid space model, including: (1) Input the triangular face elements in the irregular triangulated network model and sort them into the triangulated network domain. Let the triangular face elements be , is the center point coordinate; (2) Select the first triangulated channel set in the data arrangement , is the sampling point number of a single channel record, is the survey line number in the triangulated channel concentration, and the adjacent excavation data is selected according to the center point. The refractive coefficient sequence of the underground medium of the center point excavation at different offset distances is calculated based on the triangulated diffraction equation. ; (3) Establishing a geological volume element model in a generalized triangular prism model whose main frequency information is consistent with the triangulated network channel set : (I) Perform spectrum analysis on the zero-offset tunneling channel of the triangulated network channel set to obtain the main frequency of the triangulated network diffraction wave of the channel set ; (II) In the fusion simulation, the zero-phase Ricker wavelet is used as the geological body diffraction wavelet in the triangular prism model, so the main frequency of the triangular network diffraction wave Substituting into the Ricker wavelet formula, the expression is: ； In the formula, for The geological volume element model in the generalized triangular prism model of time, The main frequency of the diffraction wave containing the triangular network The Ricker wavelet, For time; About Time Discrete the geological volume element model in the generalized triangular prism model ; (4) Based on the geological volume element model obtained in the generalized triangular prism model and refractive index To do fusion, the expression is: ； In the formula, is the geological channel set, for After time discretization The geological volume element model in the generalized triangular prism model from time to initial value; Fusion simulation of geological channel set in generalized triangular prism model ; (5) The geological channel set in the simulated generalized triangular prism model With the input triangulation channel set For fusion correlation analysis, the expression is: ； In the formula, is the depth position correlation value of the deformation interface, For the time; When calculating the fusion correlation, the correlation is calculated according to the depth position of the deformation interface. Computational The range is truncated to obtain the correlation of each event axis at different offsets; (6) Set the threshold variable Q to 0.2 and calculate the correlation of each in-phase axis. The tunneling channel numbers below the threshold are marked and defined as distortion points; dynamic correction stretching distortion removal processing is implemented for the triangulated channel set and the geological channel set in the generalized triangular prism model; (7) Change the center point and repeat steps (2) to (6) until distortion removal is achieved for all triangulated channel sets and geological body channel sets in the generalized triangular prism model, and finally obtain an irregular triangulated network-generalized triangular prism model hybrid space model.

8. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 1, characterized in that: In step (2), the refractive index sequence of the underground medium at different offset distances at the center point is calculated based on the triangulated diffraction equation. ,include: (a) Determine the depth position of the underground deformation interface at the center point based on the excavation geological layer data and depth gauge in the excavation data , For the interface Depth (b) Convert the density curve and time difference curve in the tunneling data to obtain the density curve , P-wave velocity curve and shear wave velocity curve, and according to the depth position of the deformation interface Calculate the deformation parameters of the media on both sides of each interface and obtain the center point deformation parameter sequence ; (c) Substitute the deformation parameter sequence into the triangulated diffraction equation to calculate the refractive index of each deformation interface , is the incident angle, then It reflects the relationship between the refractive index and the angle of incidence; (d) According to the line step length in the triangulation network Convert the refraction angle to Reflecting the change relationship with the offset distance, the conversion formula is: ； In the formula, is the transformation value of the incident angle with the offset distance, is the interval distance; Thus, the refractive index sequence of the central point underground medium at different offset distances is obtained. , the triangulated diffraction equation is as follows: ； In the formula, are the first incident angle, the second incident angle, are the first longitudinal wave velocity, the second longitudinal wave velocity, are the first refraction angle, the second refraction angle, are the first shear wave refraction coefficient and the second shear wave refraction coefficient respectively. are the first shear wave velocity and the second shear wave velocity, They are respectively the refractive index of the deformation interface between longitudinal wave and longitudinal wave, and the refractive index of the deformation interface between longitudinal wave and transverse wave.

9. The method for improving the accuracy of transparent geological models based on tunneling navigation data elements according to claim 7, characterized in that: In step S5, the three-dimensional dynamic geological model is iteratively optimized, including: During the excavation process, the model update is triggered every 10m or when encountering faults, collapse columns, and unfavorable geological bodies in the scour zone. The iterative formula is expressed as: ； In the formula, For the next position The model status; is a function that represents how the model is updated according to the current position and geological conditions during the excavation process; For the current position The model status; is the interval distance; To trigger updates when encountering bad geological bodies; Obtaining a three-dimensional dynamic transparent geological model, including: model cutting and slicing, data mapping and conversion, output generation and optimization; model cutting and slicing: cutting or slicing the three-dimensional dynamic geological model to extract data from the excavation area; Data mapping and conversion: Map the 3D dynamic geological model data into the data structure of the target output format, and perform conversion and processing, including color mapping, texture mapping and lighting model application; Output generation and optimization: Generate the corresponding output file, i.e., 3D dynamic transparent geological model, according to the selected output format, and perform optimization processing such as compression and format adjustment; Visualization, including: isosurface extraction through MC algorithm and volume rendering through ray casting and texture mapping algorithm, which are used to render dynamic transparent geological models; at the same time, cluster analysis is used to identify and extract similar geological features, including low-resistance anomalies in front of the excavation.

10. A system for improving the accuracy of transparent geological models based on excavation navigation data elements, characterized in that: The system implements the method for improving the accuracy of a transparent geological model based on tunneling navigation data elements as described in any one of claims 1 to 9, and the system comprises: A multi-source data acquisition module (1) is used for multi-source data acquisition of tunneling navigation data and advanced geological detection data, wherein the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency, and the advanced geological detection data includes drilling data, geophysical data, and historical geological data; The data preprocessing and storage module (2) is used to preprocess the collected multi-source data, convert the data formats from different sources into a unified database format, and store the data in the database; A multi-source data fusion module (3) is used to call the pre-processed multi-source data, and complete the multi-source data fusion by using data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and vectorized fusion of multi-source data to return to the weight allocation model space; A three-dimensional dynamic geological model building module (4) is used to build a three-dimensional dynamic geological model based on the multi-source data fusion result and the irregular triangular network-generalized triangular prism model; The three-dimensional dynamic transparent geological model acquisition and visualization module (5) is used to iteratively optimize the obtained three-dimensional dynamic geological model, obtain the three-dimensional dynamic transparent geological model for output and visualization.

Images