Method and System for Improving the Accuracy of a Transparent Geological Model Based on Tunneling Navigation Data Elements

Abstract

The present invention belongs to the technical field of intelligent mines, and discloses a method and system for improving the accuracy of a transparent geological model based on tunneling navigation data elements. The method performs multi-source data collection of tunneling navigation data and advanced exploration geological data, preprocesses the collected multi-source data, converts the data formats from different sources into a unified database format, and performs data warehousing; completes multi-source data fusion; uses an irregular triangular network - generalized triangular prism model to build a three-dimensional dynamic geological model; iteratively optimizes the obtained three-dimensional dynamic geological model, and obtains a three-dimensional dynamic transparent geological model for output and visualization. Through multi-source data fusion and dynamic modeling, the present invention improves the spatial accuracy and real-time response ability of the model, and provides more reliable geological information support for rapid coal mine tunneling.

Description

Technical Field

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

[0002] In the field of intelligent mines, a transparent geological model is one of the core technologies for realizing intelligent and precise coal mine mining, and its accuracy directly affects tunneling path planning, safety risk assessment, and resource mining efficiency. It can intuitively present the undulating shape of coal seams and provide key support for coal mine tunneling, resource planning, and safety management. At present, most geological models rely on static geological exploration data (such as topographic data, borehole data, logging data), which have the disadvantages of long collection cycles, low spatial resolution, and poor real-time performance, resulting in geological models being difficult to accurately reflect the actual geological situation. The application of tunneling navigation technology provides real-time and refined data support for geological modeling. Navigation equipment (such as lidar, inertial navigation systems, and geophysical prospecting while tunneling) can obtain multi-dimensional data including tunneling paths, coal (rock) seam properties, position information, etc.

[0003] However, in the prior art, there are lack of effective data fusion and real-time modeling methods, resulting in the failure to give full play to the role of tunneling data. Summary of the Invention

[0004] To overcome the problems in the related art, the disclosed embodiments of the present invention provide a method and system for improving the accuracy of a transparent geological model based on tunneling navigation data elements, specifically relating to a method for improving the accuracy of a transparent geological model based on tunneling navigation data elements, which is applicable to 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 tunneling navigation data elements includes:

[0006] S1, performing multi-source data collection of tunneling navigation data and advanced exploration geological data, where the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and collection frequency, and the advanced exploration geological data includes borehole data, geophysical exploration data, and historical geological data;

[0007] S2, preprocessing the collected multi-source data, converting the data formats from different sources into a unified database format, and performing data warehousing;

[0008] S3, calling the preprocessed multi-source data, and completing multi-source data fusion by using data vectorization, weight distribution of multi-source data, vector space fusion of multi-source data, and returning the weight distribution model space of the vectorized fused multi-source data;

[0009] S4. Based on the multi-source data fusion result, use the triangulated irregular network - generalized triangular prism model to build a 3D dynamic geological model;

[0010] S5. Iteratively optimize the obtained 3D dynamic geological model, and obtain a 3D dynamic transparent geological model for output and visualization.

[0011] In step S2, preprocess the collected multi-source data, including: data cleaning and denoising; clean and denoise the collected tunneling navigation data, remove outliers and noise data, and use transformation to attenuate the underground environmental noise, and the expression is:

[0012] ;

[0013] In the formula, is the attenuation value of the underground environmental noise, is the tunneling navigation signal at time is the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Reynolds denoising value;

[0014] After performing noise interference attenuation on the output data , use inverse transformation to return the data after noise attenuation.

[0015] In step S2, convert the data formats from different sources into a unified database format and perform data warehousing; including:

[0016] Data coordinate alignment: Perform coordinate transformation and alignment on the tunneling navigation data and traditional geological data so that the data is represented in a unified geological coordinate system, and use control points for spatial registration;

[0017] Data format unification: Convert the point cloud, raster, and vector data formats from different sources into a unified database format;

[0018] The data warehousing includes: adopting the ETL (Extract, Transform, Load) method, and according to the unified data conversion rule set, implementing the structural conversion from the business system database to the geological database; the ETL method of data extraction, transformation and loading adopts two methods: active data push and passive data extraction; use the database connection tool PostgreSQL to establish a connection with the database, construct corresponding data insertion statements according to the database table structure and data format, insert the data into the database, and verify the data integrity and accuracy after warehousing.

[0019] In step S3, data vectorization includes: performing vectorization processing on the acquired multi-source data and mapping it to the same vector space; for image data, extracting the features of the image through texture analysis and edge detection, and then performing vectorization; for sound data, extracting features through Mel Frequency Cepstral Coefficients (MFCC) and spectral features and then performing vectorization.

[0020] Weight assignment for multi-source data includes: using a weight assignment model to fuse the multi-source data of tunneling navigation data and advanced detection geological data. The expression is:

[0021] ;

[0022] In the formula, is the calculated weight of the th input element, is the maximum weight assignment function, is the linear calculation function, is the value after preprocessing of the th input element.

[0023] In step S3, the vector space fusion of multi-source data includes:

[0024] After weight reallocation, data fusion is performed. The expression is:

[0025] ;

[0026] In the formula, is the data fused and output after calculation with the new weight;

[0027] The vectorized fusion of multi-source data returning to the weight assignment model space includes: returning the vectorized and fused data to the weight assignment model space. The expression is:

[0028] ;

[0029] In the formula, is the element converted to the model space, is the conversion factor, is the balance coefficient.

[0030] In step S4, using the Triangulated Irregular Network - Generalized Triangular Prism model for 3D dynamic geological model building includes:

[0031] Step 1: Encrypt the tunneling cloud data through a smooth discrete interpolation algorithm;

[0032] Step 2: For the point cloud data with the same number, use the point-by-point insertion method to generate an irregular triangular network model of each geological interface to describe the changes and characteristics of the interface;

[0033] Step 3: Based on the irregular triangular network model, search for formation information in a top-down order; by comparing and connecting the formation information between every three point cloud data sets, generate an initial generalized triangular prism model;

[0034] Step 4: For areas with complex geological conditions, expand the initial generalized triangular prism model into smaller generalized triangular prism models, or degenerate it into tetrahedron models and pyramid models, so that the initial generalized triangular prism model or tetrahedron models and pyramid models are consistent with the actual geological phenomena;

[0035] Step 5: By corresponding and fusing the triangular facets in the irregular triangular network model with the geological volume elements in the generalized triangular prism model, construct an irregular triangular network - generalized triangular prism model hybrid space model;

[0036] Step 6: Verify and optimize the constructed irregular triangular network - generalized triangular prism model hybrid model;

[0037] Step 7: Use the irregular triangular network - generalized triangular prism model hybrid model for 3D geological geometric modeling, and simultaneously perform data analysis and mining to obtain useful information from the irregular triangular network - generalized triangular prism model hybrid model.

[0038] In Step 5, by corresponding and fusing the triangular facets in the irregular triangular network model with the geological volume elements in the generalized triangular prism model, an irregular triangular network - generalized triangular prism model hybrid space model is constructed, including:

[0039] (1) Input the triangular facets in the irregular triangular network model and sort them into the triangular network domain. Let the triangular facet be , representing the central point coordinates;

[0040] (2) Select the first triangular network channel set in the data arrangement , representing the sampling point serial number of a single - channel record, representing the survey line serial number in the triangular network channel set, and select adjacent tunneling data based on the central point. Calculate the refraction coefficient sequence of the underground medium where the central point tunnels at different offset distances based on the triangular network diffraction equation;

[0041] (3) Establish a geological volume element model in the generalized triangular prism model where the dominant frequency information is consistent with the triangular network channel set :

[0042] (I) Conduct a frequency spectrum analysis on the zero - offset tunneling channel of the triangular network channel set to obtain the dominant frequency of the triangular network diffraction wave of the channel set;

[0043] (II) In the fusion simulation, the Ricker wavelet with zero phase is used as the diffracted wavelet of the geological body in the triangular prism model. Therefore, the main frequency of the triangular grid diffracted wave is substituted into the Ricker wavelet formula, and the expression is:

[0044] ;

[0045] In the formula, is the element model of the geological body in the generalized triangular prism model with respect to time, is the Ricker wavelet containing the main frequency of the triangular grid diffracted wave, is time;

[0046] Discretization is performed on time to obtain the element model of the geological body in the generalized triangular prism model;

[0047] (4) According to the obtained element model of the geological body in the generalized triangular prism model and the refraction coefficient for fusion, the expression is:

[0048] ;

[0049] In the formula, is the geological body channel set, is the element model of the geological body in the generalized triangular prism model from the time after discretization to the initial value;

[0050] The geological body channel set in the generalized triangular prism model is fusion-simulated;

[0051] (5) A fusion correlation analysis is performed on the geological body channel set in the generalized triangular prism model obtained by simulation and the input triangular grid channel set , and the expression is:

[0052] ;

[0053] In the formula, is the correlation value of the depth position of the deformation interface, is the time;

[0054] When performing the fusion correlation calculation, truncation is performed on the range of the correlation calculated according to the depth position of the deformation interface to obtain the correlation of each event axis at different offsets;

[0055] (6) Set the threshold variable Q to 0.2, and mark the tunneling channel numbers with the cross-axis correlation lower than the threshold, which are defined as distortion points; starting from this point, the cross-axis shows severe distortion, so the data after the cross-axis distortion points are removed for each cross-axis, realizing the dynamic correction and stretching distortion removal processing of the triangular mesh channel set and the geological body channel set in the generalized triangular prism model; (7) Change the center point, and repeat steps (2)-(6) until the distortion removal is realized for all the triangular mesh channel sets and the geological body channel sets in the generalized triangular prism model, and finally obtain the irregular triangular mesh-generalized triangular prism model hybrid space model.

[0056] (7) Change the center point, and repeat steps (2)-(6) until the distortion removal is realized for all the triangular mesh channel sets and the geological body channel sets in the generalized triangular prism model, and finally obtain the irregular triangular mesh-generalized triangular prism model hybrid space model.

[0057] In step (2), based on the triangular mesh diffraction equation, calculate the refraction coefficient sequence of the underground medium tunneling at the center point at different offsets , including:

[0058] (a) Determine the depth position of the underground deformation interface at the center point from the tunneling geological layer data and depth gauge in the tunneling data , representing the depth of the interface ;

[0059] (b) Convert the density curve and time difference curve in the tunneling data to obtain the density curve , the longitudinal wave velocity curve and the shear wave velocity curve, and calculate the deformation parameters of the media on both sides of each interface according to the depth position of the deformation interface to obtain the center point deformation parameter sequence ;

[0060] (c) Substitute the deformation parameter sequence into the triangular mesh diffraction equation to calculate the refraction coefficient of each deformation interface, representing the incident angle, and at this time reflects the change relationship of the refraction coefficient with the incident angle;

[0061] (d) According to the survey line step length in the triangular mesh, convert the refraction angle so that reflects the change relationship with the offset, and the conversion formula is:

[0062] ;

[0063] In the formula, is the transformation value of the incident angle with the offset, is the interval distance;

[0064] Thus, the refraction coefficient sequence of the underground medium at the center point at different offsets is obtained , the triangular network diffraction equation is as follows:

[0065] ;

[0066] In the formula, are the first incident angle and the second incident angle respectively, are the first longitudinal wave velocity and the second longitudinal wave velocity respectively, are the first refraction angle and the second refraction angle respectively, are the first shear wave density and the second shear wave density respectively, are the first shear wave velocity and the second shear wave velocity respectively, are the refraction coefficients of the deformation interface between longitudinal waves and longitudinal waves, and the refraction coefficients of the deformation interface between longitudinal waves and shear waves respectively.

[0067] In step S5, the three-dimensional dynamic geological model is iteratively optimized, including:

[0068] During the tunneling process, the model is updated every 10 m or when encountering bad geological bodies such as faults, collapse columns, and scouring zones. The iterative formula is expressed as:

[0069] ;

[0070] In the formula, is the model state of the next position ; is a function indicating how the model is updated according to the current position and geological conditions during the tunneling process; is the current position of the model state; is the interval distance; is the situation of encountering bad geological bodies, triggering an update;

[0071] The acquisition of the three-dimensional dynamic transparent geological model includes: model cutting and slicing, data mapping and conversion, and output generation and optimization;

[0072] Among them, model cutting and slicing: cutting or slicing the three-dimensional dynamic geological model to extract the data in the tunneling area;

[0073] Data mapping and conversion: mapping the three-dimensional dynamic geological model data into the data structure of the target output format, and performing conversion and processing, including color mapping, texture mapping, and application of the lighting model;

[0074] Output generation and optimization: generating the corresponding output file, that is, the three-dimensional dynamic transparent geological model, according to the selected output format, and performing optimization processing such as compression and format adjustment;

[0075] The visualization includes: extracting isosurfaces through the MC algorithm, performing ray casting, and using texture mapping algorithms for volume rendering to render a dynamic transparent geological model; at the same time, using cluster analysis to identify and extract similar geological features, including low-resistance anomaly bodies in front of the tunneling face.

[0076] 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. This system implements the method for improving the accuracy of a transparent geological model based on tunneling navigation data elements. The system includes:

[0077] A multi-source data acquisition module for performing multi-source data acquisition of tunneling navigation data and advanced exploration geological data. Among them, the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency, and the advanced exploration geological data includes borehole data, geophysical exploration data, and historical geological data;

[0078] A data preprocessing and warehousing module for preprocessing the acquired multi-source data, converting the data formats from different sources into a unified database format, and performing data warehousing;

[0079] A multi-source data fusion module for calling the preprocessed multi-source data, using data vectorization, weight assignment for multi-source data, vector space fusion of multi-source data, and returning the weight assignment model space by vectorizing and fusing multi-source data to complete multi-source data fusion;

[0080] A three-dimensional dynamic geological model modeling module, based on the multi-source data fusion result, using the irregular triangular network - generalized triangular prism model for three-dimensional dynamic geological model modeling;

[0081] A three-dimensional dynamic transparent geological model acquisition and visualization module for iteratively optimizing the obtained three-dimensional dynamic geological model, obtaining a three-dimensional dynamic transparent geological model for output and visualization.

[0082] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: Through multi-source data fusion and dynamic modeling, the present invention improves the spatial accuracy and real-time response ability of the model, providing more reliable geological information support for rapid coal mine tunneling.

[0083] Real-time dynamic update: Using the real-time data generated during tunneling, dynamically adjusting the geological model to improve the adaptability to complex geological conditions.

[0084] Multi-source data fusion: Combining traditional geological data with real-time navigation data to achieve data complementarity and enhance the reliability and accuracy of the model.

[0085] High-resolution local modeling: Conducting refined modeling for key areas in front of the tunneling face to provide reliable and effective geological support for accurate tunneling.

[0086] Intelligent decision-making support: Through dynamic models and 3D visualization technology, it provides a scientific basis for the optimization of tunneling paths and risk early warning. Description of the Drawings

[0087] The drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0088] Figure 1 is a flowchart 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;

[0089] Figure 2 is a schematic diagram of a system for improving the accuracy of a transparent geological model based on tunneling navigation data elements provided by an embodiment of the present invention;

[0090] In the figure: 1. Multi-source data acquisition module; 2. Data preprocessing and warehousing module; 3. Multi-source data fusion module; 4. 3D dynamic geological model modeling module; 5. 3D dynamic transparent geological model acquisition and visualization module. Detailed Embodiments

[0091] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be given with reference to the accompanying drawings. Many specific details are set forth in the following description to facilitate a thorough 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 departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0092] Embodiment 1. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements provided by an embodiment of the present invention includes:

[0093] S1. Perform multi-source data acquisition of tunneling navigation data and advanced exploration geological data. Among them, the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency, and the advanced exploration geological data includes borehole data, geophysical exploration data, and historical geological data;

[0094] S2. Preprocess the acquired multi-source data, convert the data formats from different sources into a unified database format, and perform data warehousing;

[0095] S3. Call the preprocessed multi-source data, and use data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and return the weight allocation model space of the vectorized fusion multi-source data to complete multi-source data fusion;

[0096] S4. Based on the multi-source data fusion result, use the irregular triangular network - generalized triangular prism model to build a three-dimensional dynamic geological model;

[0097] S5. Iteratively optimize the obtained three-dimensional dynamic geological model, and obtain a three-dimensional dynamic transparent geological model for output and visualization.

[0098] Embodiment 2. As another implementation manner of the present invention, 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;

[0099] Step 1. Data collection.

[0100] Before tunneling, anticipate the cross-section in front of the tunneling through borehole data and geophysical exploration data (such as three-dimensional seismic data). During the tunneling process, use roadway geological survey real-time data and collect relevant parameters of tunneling equipment (position information, dip angle, speed, lithology, etc. of the tunneling path; point cloud data of the roadway morphology by lidar; undulation and structure data of the front coal seam by geophysical exploration during tunneling) as the basic data source for model construction.

[0101] Exemplarily, data collection includes tunneling navigation data and advanced exploration geological data;

[0102] (1) Tunneling navigation data.

[0103] The types of equipment used include lidar, inertial navigation system, seismic monitoring system during tunneling, etc. The collected content includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and collection frequency; among them, the spatial position information includes the three-dimensional coordinates, dip angle, azimuth angle, etc. of the tunneling path. The hidden disaster-causing parameters include water accumulation abnormal areas, faults, etc. The dynamic tunneling data includes tunneling speed, tunneling distance, and equipment status. The collection frequency includes the equipment recording data at a frequency of 1 second / time to ensure the spatio-temporal continuity of the data.

[0104] (2) Advanced exploration geological data.

[0105] The data types include borehole data, geophysical exploration data (such as electrical method, magnetic method, diffraction wave detection data of triangular network), and historical geological data. The main content includes borehole columnar diagrams and cross-sectional diagrams; the borehole columnar diagrams include lithology, distribution characteristics of coal (rock) layers, etc. The cross-sectional diagrams include the positions of hidden abnormal bodies, fault morphologies, etc.

[0106] Step 2. Data preprocessing, including:

[0107] 2.1. Data cleaning and denoising;

[0108] Correct the equipment drift and environmental noise (such as vibration or tunneling equipment interference) existing in the tunneling navigation data, that is, clean and denoise the collected tunneling navigation data, eliminate outliers and noise data, and ensure data quality. Among them, s-transform is used for denoising, which can attenuate the underground environmental noise, and the expression is:

[0109] ;

[0110] In the formula, is the attenuation value of the underground environmental noise, is the tunneling navigation signal at time is the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Reynolds denoising value;

[0111] After performing noise interference attenuation on the output data use inverse transform to return the data after noise attenuation.

[0112] Perform coordinate transformation and alignment on the tunneling navigation data and traditional geological data to ensure that the data is represented in a unified geological coordinate system.

[0113] 2.2, Data coordinate alignment; Perform coordinate transformation and alignment on the tunneling navigation data and traditional geological data to ensure that the data is represented in a unified geological coordinate system.

[0114] Exemplarily, convert the local coordinate system of the tunneling navigation data to a global coordinate system (such as the UTM coordinate system) consistent with the traditional geological data.

[0115] Use control points (such as known borehole positions) for spatial registration to ensure that the alignment error of multi-source data is less than 0.1 meter.

[0116] 2.3, Data format unification; Convert data formats from different sources (such as point cloud, raster, vector) to a unified database format for convenient subsequent fusion and analysis.

[0117] Step 3, Data storage in database;

[0118] Adopt the data extraction, transformation, and loading (Extract-Transform-Load, ETL) method to achieve the structural transformation of the business system database to the geological database according to the unified data transformation 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 corresponding data insertion statements (such as the INSERT INTO statement) according to the database table structure and data format, and insert the data into the database. After the data is stored in the database, verify the integrity and accuracy of the data, and use constraints and triggers to ensure the integrity and accuracy of the database.

[0119] Step 4: Multi-source data fusion, including:

[0120] Step 4.1 Data vectorization.

[0121] Perform vectorization processing on the multi-source data obtained in Step 1 and map it to the same vector space. For common image data, image features can be extracted through techniques such as texture analysis and edge detection, and then vectorized; for general sound data, features can be extracted through Mel-frequency cepstral coefficients (MFCC) and spectral features and then vectorized.

[0122] Step 4.2 Multi-source data weight assignment.

[0123] Use machine learning algorithms or weight assignment models to fuse multi-source data such as tunneling navigation data and geological data. The expression is:

[0124] ;

[0125] In the formula, is the calculated weight of the th input element, is the maximum weight assignment function, is the linear calculation function, is the value after preprocessing of the th input element

[0126] Analyze the applicability of multi-source data. When the geological environment changes complexly, increase the weight of tunneling navigation data; when the geological conditions are simple, increase the weight of traditional geological data.

[0127] Step 4.3 Vector space fusion of multi-source data.

[0128] After the weights are reallocated, the expression for data fusion is:

[0129] ;

[0130] In the formula, The output data is fused after calculation using new weights;

[0131] Step 4.4 vectorizes the fused multi-source data and returns it to the model space.

[0132] 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.

[0133] ;

[0134] In the formula, For elements converted to model space, is the conversion factor, is the equalization coefficient.

[0135] Step 5: Accurate three-dimensional dynamic geological model building.

[0136] 5.1, Three-dimensional dynamic geological model modeling algorithm.

[0137] 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.

[0138] Exemplarily, the three-dimensional dynamic geological model is constructed based on the irregular triangulated network-generalized triangular prism model, and the key steps include:

[0139] Step 1: Encrypt the tunneling cloud data using a smooth discrete interpolation algorithm.

[0140] 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.

[0141] 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.

[0142] Step 4: For areas with complex geological conditions, expand the generalized triangular prism model into smaller generalized triangular prism models, or degenerate it into tetrahedron models and pyramid models to ensure that the model is consistent with the actual geological phenomena.

[0143] Step 5: By corresponding and fusing the triangular facets in the irregular triangular mesh model with the geological volume elements in the generalized triangular prism model, an irregular triangular mesh - generalized triangular prism model hybrid spatial model is constructed.

[0144] Exemplarily, in Step 5, by corresponding and fusing the triangular facets in the irregular triangular mesh model with the geological volume elements in the generalized triangular prism model, the construction of the irregular triangular mesh - generalized triangular prism model hybrid spatial model includes:

[0145] (5.1) Input the triangular facets in the irregular triangular mesh model and sort them into the triangular mesh domain. Let the triangular facet be , representing the central point coordinates;

[0146] (5.2) Select the first triangular mesh channel set in the data arrangement , representing the sampling point serial numbers of a single - trace record, representing the survey line serial number in the triangular mesh channel set, and select adjacent tunneling data according to the central point. Calculate the refraction coefficient sequence of the underground medium where the central point tunnels at different offset distances based on the triangular mesh diffraction equation , including:

[0147] (a) Determine the depth position of the underground deformation interface of the central point , representing the interface 's depth;

[0148] (b) Make conversions on the density curve and time - difference curve in the tunneling data to obtain the density curve , the longitudinal wave velocity curve and the shear wave velocity curve, and calculate the deformation parameters of the media on both sides of each interface according to the depth position of the deformation interface to obtain the central point deformation parameter sequence ;

[0149] (c) Substitute the deformation parameter sequence into the triangular mesh diffraction equation to calculate the refraction coefficient of each deformation interface, representing the incident angle. At this time, reflects the variation relationship of the refraction coefficient with the incident angle;

[0150] (d) According to the survey line step length in the triangular network make a conversion of the refraction angle so that it reflects the variation relationship with the offset. The conversion formula is:

[0151] ;

[0152] In the formula, is the transformed value of the incident angle with respect to the offset, is the interval distance;

[0153] Thus, a sequence of refraction coefficients of the subsurface medium at the center point for different offsets is obtained , and the triangular network diffraction equation is as follows:

[0154] ;

[0155] In the formula, are the first incident angle and the second incident angle respectively, are the first longitudinal wave velocity and the second longitudinal wave velocity respectively, are the first refraction angle and the second refraction angle respectively, are the first shear wave density and the second shear wave density respectively, are the first shear wave velocity and the second shear wave velocity respectively, are the refraction coefficients of the deformation interface between longitudinal waves - longitudinal waves and the refraction coefficient of the deformation interface between longitudinal waves - shear waves respectively;

[0156] (5.3) Establish a geological body element model in the generalized triangular prism model where the dominant frequency information is consistent with the triangular network channel set :

[0157] (I) Conduct a frequency spectrum analysis on the zero-offset tunneling channel of the triangular network channel set to obtain the dominant frequency of the triangular network diffraction wave of the channel set ;

[0158] (II) In the fusion simulation, use a zero-phase Ricker wavelet as the diffraction wavelet of the geological body in the triangular prism model. Therefore, substitute the dominant frequency of the triangular network diffraction wave into the Ricker wavelet formula. The expression is:

[0159] ;

[0160] In the formula, is the geological body element model in the generalized triangular prism model at time is the Ricker wavelet containing the dominant frequency of the triangular network diffraction wave, is time;

[0161] For time Perform discrete processing to obtain the geological body element model in the generalized triangular prism model ;

[0162] (5.4)According to the obtained geological body element model in the generalized triangular prism model and the refraction coefficient perform fusion, and the expression is:

[0163] ;

[0164] In the formula, is the geological body channel set, is the geological body element model in the generalized triangular prism model from the time after discrete time processing to the initial value;

[0165] Fusion simulates the geological body channel set in the generalized triangular prism model ;

[0166] (5.5)Perform fusion correlation analysis on the simulated geological body channel set in the generalized triangular prism model and the input triangular mesh channel set The expression is:

[0167] ;

[0168] In the formula, is the depth position of the deformation interface to the correlation value, is the th moment;

[0169] When performing fusion correlation calculation, truncate according to the range of the depth position of the deformation interface to the correlation calculated to obtain the correlation of each in-phase axis at different offsets;

[0170] (5.6)Set the threshold variable Q to 0.2, mark the tunneling channel numbers with correlation lower than the threshold as distortion points; from this point, the in-phase axis appears severely distorted, so the data after the distortion point of each in-phase axis is removed to achieve dynamic correction stretching distortion removal processing of the triangular mesh channel set and the geological body channel set in the generalized triangular prism model;

[0171] (5.7)Change the center point, and repeat steps (5.2) - (5.6) until distortion removal is achieved for all triangular mesh channel sets and geological body channel sets in the generalized triangular prism model, and finally obtain an irregular triangular mesh - generalized triangular prism model hybrid space model.

[0172] ​Step 6: Verify and optimize the constructed Triangulated Irregular Network - Generalized Triprism model hybrid model to ensure that the model can accurately and efficiently describe three - dimensional spatial data.

[0173] Step 7: Use the Triangulated Irregular Network - Generalized Triprism model hybrid model for three - dimensional geological geometric modeling, and at the same time conduct data analysis and mining to obtain useful information from the model.

[0174] 5.2, Dynamic feedback of abnormal geological bodies.

[0175] Taking a fault as an example to illustrate the modeling method of abnormal bodies. When modeling a fault, the "overall method" is adopted to solve the problem of fault data modeling. The specific process is as follows:

[0176] (i) Calculate the fault plane; A fault consists of one or several fault points, and each point includes strike, dip angle, and throw. When only one fault point is recorded, extension data should also be included. The fault triangular plane can be calculated from the attributes of the initial fault point.

[0177] (ii) Calculate the coal - fault intersection line; The intersection line obtained by intersecting the fault plane with the irregular triangular network of the coal seam is the coal - fault intersection line. The throw of the corresponding points on the coal - fault intersection line is calculated from the throw of the initial fault point and the distribution ratio.

[0178] (iii) Divide the irregular triangular network of the coal seam; According to the normal and reverse attributes of the fault, the strata are divided into the hanging wall and footwall areas of the fault. The intersection line is inserted, and the coal seam triangles within the range are divided into hanging wall and footwall triangles.

[0179] (iv) Calculate the coal seam displacement; The displacement distance of each corner point of the hanging wall and footwall triangles in the XOY plane and the elevation change in the Z direction are calculated from the fault plane respectively.

[0180] (v) Influence area calculation; The fault influence area is calculated from the throw value of the coal - fault intersection line, and the default correction radius is 10m per meter of throw; The change in the Z - coordinate of the corner points of the triangles in the irregular triangular network of the strata within the fault influence area is calculated.

[0181] (vi) Calculate the coal seam displacement within the influence area; Calculate the change in the position of the corner points of the irregular triangular network of the strata in the XOY plane within the influence area of the normal fault.

[0182] Step 6, Iterative optimization of the three - dimensional dynamic geological model.

[0183] During the tunneling process, the model is updated every 10m or when encountering bad geological bodies such as faults, collapse columns, and scouring zones, to improve the dynamic response ability of the model.

[0184] The basic form of the iterative formula can be expressed as:

[0185] ;

[0186] In the formula, is the next position of the model state; is a function that represents how the model is updated during tunneling based on the current position and geological conditions. Usually, this function can be a numerical model of geological engineering that takes into account the geometric model and properties. is the interval distance, where is the interval distance for each update; Geological Condition refers to the situation of encountering unfavorable geological bodies (such as faults, collapse columns, etc.), which triggers special update rules; is the current position of the model state.

[0187] Another exemplary case is that when the tunneling advance distance reaches 10 meters or there are significant changes in the navigation data, such as revealing a fault, the model is updated. Through the input of tunneling navigation data, the model within the custom range of the new data is updated using the follow-up grid technology to meet the timely update of the mine strata model. The working face model can utilize measurement data, inertial navigation data, and seismic data during tunneling to quickly update locally or globally, dynamically improving the accuracy of the coal seam model in front of the mining face. The basic resolution is 0.5 meters, and in geologically complex areas, it is increased to 0.1 meters to capture more detailed geological feature changes. The latest data is loaded in real-time and the model is reconstructed, while retaining the historical model for reference.

[0188] Error analysis: Compare the model prediction results with the actual tunneling data. According to the actual tunneling situation, the predicted geological model is corrected, and based on this, the geological model of the unexcavated area is analyzed. The process of model verification analysis and iterative optimization is as follows: typical research field area complex three-dimensional geological body → original geological data → three-dimensional geological comprehensive database → initial three-dimensional dynamic geological model → accuracy evaluation model → visual display of model accuracy (three-dimensional spatial distribution model of geological structure uncertainty) → detect whether the model accuracy meets the requirements → if not, perform model error correction → corrected three-dimensional dynamic geological model → re-detect whether the model accuracy meets the requirements, if not, perform model error correction; if it meets the requirements, conduct actual engineering experiments and verification in the research field area → correct the accuracy evaluation model → model prediction and actual engineering analysis and application.

[0189] Step 7, output and visualization of the three-dimensional dynamic transparent geological model.

[0190] Using data integration and real-time update capabilities, 3D dynamic geological modeling and visualization technology, a dynamic transparent geological model is generated. This includes: real-time display through a 3D visualization platform, facilitating decision-makers to intuitively understand the geological situation and optimize the tunneling plan. Output of the dynamic transparent geological model is carried out using 3D dynamic geological model slicing and data mapping techniques, and volume rendering such as the MC algorithm and Ray Casting is used to achieve the goal of 3D visualization.

[0191] Exemplarily, the output of the 3D dynamic transparent geological model includes model cutting and slicing, data mapping and conversion, output generation and optimization;

[0192] Among them, model cutting and slicing: The 3D dynamic geological model is cut or sliced to extract data in the tunneling area.

[0193] Data mapping and conversion: Map the 3D dynamic geological model data into the data structure of the target output format (common formats such as.obj,.fbx,.shp,.dwg, etc.), and perform necessary conversions and processing, such as color mapping, texture mapping, and application of lighting models.

[0194] Output generation and optimization: Generate the corresponding output file according to the selected output format, that is, the 3D dynamic transparent geological model, and perform optimization processing such as compression and format adjustment to improve the readability and usability of the output file.

[0195] Visualization technology:

[0196] Isosurface extraction is carried out through the Marching Cubes (MC) algorithm, and volume rendering is performed using the Ray Casting and texture mapping algorithms for rendering the dynamic transparent geological model. At the same time, clustering analysis is used to identify and extract similar geological features, such as low-resistance anomaly bodies in front of tunneling.

[0197] Example 3, as Figure 2 shown, a system for improving the accuracy of a transparent geological model based on tunneling navigation data elements includes:

[0198] A multi-source data acquisition module 1 for performing multi-source data acquisition of tunneling navigation data and advanced exploration geological data. Among them, the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency, and the advanced exploration geological data includes borehole data, geophysical exploration data, and historical geological data;

[0199] A data preprocessing and warehousing module 2 for preprocessing the acquired multi-source data, converting the data formats from different sources into a unified database format, and performing data warehousing;

[0200] The multi-source data fusion module 3 is used to call the preprocessed multi-source data, and complete multi-source data fusion by using data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and vectorizing and fusing multi-source data to return to the weight allocation model space.

[0201] The three-dimensional dynamic geological model modeling module 4 is based on the multi-source data fusion result, and uses the irregular triangular network - generalized triangular prism model to perform three-dimensional dynamic geological model modeling.

[0202] The three-dimensional dynamic transparent geological model acquisition and visualization module 5 is used to iteratively optimize the obtained three-dimensional dynamic geological model, and acquire and output and visualize the three-dimensional dynamic transparent geological model.

[0203] As mentioned above, only the relatively optimal specific implementation manners of the present invention are described, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, improvement, etc. made by any person skilled in the technical field within the technical scope disclosed by the present invention, as long as it is made within the spirit and principle of the present invention, shall be covered by 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-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; 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 point 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; 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: ： Wherein, is the model state of the next position ; is a function indicating how the model is updated according to the current position and geological conditions during tunneling; is the model state of the current position ; is the interval distance; is the situation of encountering poor geological bodies, triggering an update; 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 according to the selected output format, i.e., a three-dimensional dynamic transparent geological model, and perform compression and format adjustment optimization processing; Visualization, including: Extracting isosurfaces through the MC algorithm and performing volume rendering through ray casting and texture mapping algorithms for rendering the dynamic transparent geological model; At the same time, using cluster analysis to identify and extract similar geological features, including low-resistance anomaly bodies in front of the tunneling face.

2. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements according to claim 1, wherein In step S2, preprocess the collected multi-source data, including: cleaning and denoising the collected tunneling navigation data, removing outliers and noise data, and using transformation to attenuate the underground environmental noise, and the expression is: ； Wherein, is the attenuation value of the underground environmental noise, is the tunneling navigation signal at time the absolute value of the noise frequency, is the Reynolds denoising value for attenuation, is the Reynolds denoising value; For the output data After performing noise interference attenuation, use Inverse transformation to return the data after noise attenuation.

3. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements according to claim 1, wherein, In step S2, convert the data formats from different sources into a unified database format and perform data warehousing, including: Data coordinate alignment: Perform coordinate transformation and alignment on the tunneling navigation data and geological data so that the data is represented in a unified geological coordinate system, and use control points for spatial registration; Unify data formats: Convert the point cloud, raster, and vector data formats from different sources into a unified database format; Data warehousing includes: Adopting the ETL (Extract, Transform, Load) method, and implementing the structural transformation from the business system database to the geological database according to the unified data conversion rule set; The ETL method of data extraction, transformation, and loading adopts two methods: active data push and passive data extraction; Use the database connection tool PostgreSQL to establish a connection with the database, construct the corresponding data insertion statements according to the database table structure and data format, insert the data into the database, and verify the data integrity and accuracy after warehousing.

4. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements according to claim 1, characterized in that, In step S3, data vectorization, including: Perform vectorization processing on the obtained multi-source data and map it to the same vector space; For image data, extract the features of the image through texture analysis and edge detection, and then perform vectorization; For sound data, extract the features through the Mel Frequency Cepstral Coefficient (MFCC) and spectral features, and then perform vectorization; Multi-source data weight assignment, including: Using the weight assignment model to fuse the multi-source data of the tunneling navigation data and the advanced detection geological data, and the expression is: ； Wherein, is the calculation weight of the th input element, is the maximum weight distribution function, is the linear calculation function, is the value after preprocessing of the th input element.

5. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements according to claim 4, characterized in that In step S3, the vector space fusion of multi-source data, including: After the weights are re-assigned, perform data fusion, and the expression is: ； In the formula, is the data fused and output after calculating with the new weights; Return the vectorized fused multi-source data to the weight assignment model space, including: Return the vectorized fused data to the weight assignment model space, and the expression is: ； Wherein, is the element converted to the model space, is the conversion factor, is the equilibrium coefficient.

6. The method for improving the accuracy of a transparent geological model based on tunneling navigation data elements according to claim 1, characterized in that In step five, correspond and fuse the triangular facets in the Triangulated Irregular Network (TIN) model with the geological volume elements in the Generalized Triangular Prism (GTP) model to construct a hybrid space model of the TIN-GTP model, including: (1) Input the triangular facets in the irregular triangular network model and sort them into the triangular network domain. Let the triangular facet be , be the central point coordinates; (2) Select the first triangular network channel set in the data arrangement , is the sampling point serial number of the single-channel record, is the survey line serial number in the triangular network channel set, and adjacent tunneling data is selected according to the center point. Based on the triangular network diffraction equation, the refraction coefficient sequence of the underground medium where the center point tunnels at different offset distances is calculated ; (3)Establish a geological volume model in the generalized triangular prism model where the main frequency information is consistent with the triangular mesh channel set : (I) Perform spectral analysis on the zero-offset tunneling channels of the triangular network channel set to obtain the main frequency of the diffracted waves of the triangular network in the channel set ; (II) In the fusion simulation, a zero-phase Ricker wavelet is used as the diffracted wave of the geological body in the triangular prism model. Therefore, the main frequency of the triangular network diffracted wave is substituted into the Ricker wavelet formula, and the expression is as follows: ； In the formula, is the geological body element model in the generalized triangular prism model of time, is the Ricker wave containing the main frequency of the triangular grid diffracted wave , is time; For time perform discrete processing to obtain the geological body element model in the generalized triangular prism model ; (4)Based on the geological body element model in the obtained generalized triangular prism model and the refraction coefficient perform fusion, and the expression is: ； In the formula, is the geological body channel set, is the geological body element model in the generalized triangular prism model from the time after discrete processing to the initial value; Fusion simulates the geological body channel set in the generalized triangular prism model ; (5) Perform a fusion correlation analysis on the geological body channel set in the simulated generalized triangular prism model and the input triangular mesh channel set The expression is as follows: ； In the formula, is the depth position of the deformation interface with respect to the correlation value, is the moment; When performing the fusion correlation calculation, truncate the range of correlation calculation according to the depth position of the deformation interface to obtain the correlation of each in-phase axis at different offsets;​​​​ (6) Set the threshold variable Q to 0.2, and mark the serial numbers of the tunneling channels with cross-axis correlation lower than the threshold, which are defined as distortion points; implement the dynamic correction stretching distortion excision processing for the triangular mesh channel set and the geological body channel set in the generalized triangular prism model; For the geological body channel set in the triangular mesh channel set and the generalized triangular prism model, perform dynamic correction stretching distortion excision processing; (7) Change the center point and repeat steps (2)-(6) until distortion excision is achieved for all triangular network channel sets and geological body channel sets in the generalized triangular prism model, and finally obtain a hybrid space model of the TIN-GTP model.

7. The method for improving the accuracy of the transparent geological model based on the tunneling navigation data elements according to claim 6, characterized in that In step (2), calculate the refraction coefficients of the underground medium excavated by the central point at different offsets based on the triangular network diffraction equation , including: (a) Determine the depth position of the underground deformation interface at the center point based on the tunneling geological layer data and depth gauge in the tunneling data. , is the interface depth. (b)Convert the density curve and the time difference curve in the tunneling data to obtain the density curve , the longitudinal wave velocity curve and the shear wave velocity curve, and calculate the deformation parameters of the media on both sides of each interface according to the depth position of the deformation interface to obtain the central point deformation parameter sequence ; (c) Substitute the deformation parameter sequence into the triangular network diffraction equation to calculate the refractive index of each deformed interface , is the incident angle, and at this time reflects the variation relationship of the refractive index with the incident angle; (d) According to the survey line step length in the triangular network perform conversion on the refraction angle so that reflects the variation relationship with the offset. The conversion formula is: ； In the formula, is the transformation value of the incident angle with respect to the offset, is the interval distance; Thus, a refraction coefficient sequence of the underground medium at the center point with different offsets is obtained. , and the triangular network diffraction equation is as follows: ； Wherein, are the first incident angle and the second incident angle respectively, are the first longitudinal wave velocity and the second longitudinal wave velocity respectively, are the first refraction angle and the second refraction angle respectively, 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 respectively, are the refraction coefficient of the deformation interface between longitudinal waves and longitudinal waves, and the refraction coefficient of the deformation interface between longitudinal waves and shear waves respectively.

8. A system for improving the accuracy of a transparent geological model based on tunneling navigation data elements, characterized in that, This system implements the method for improving the accuracy of the transparent geological model based on the tunneling navigation data elements described in any one of claims 1-7. This system includes: Multi-source data acquisition module (1), which is used to perform multi-source data acquisition of tunneling navigation data and advanced detection geological data. Among them, the tunneling navigation data includes spatial position information, hidden disaster-causing parameters, dynamic tunneling data, and acquisition frequency, and the advanced detection geological data includes borehole data, geophysical exploration data, and historical geological data; Data preprocessing and storage module (2), which is used to preprocess the acquired multi-source data, convert the data formats from different sources into a unified database format, and perform data storage; Multi-source data fusion module (3), which is used to call the preprocessed multi-source data, and complete multi-source data fusion by using data vectorization, multi-source data weight allocation, vector space fusion of multi-source data, and returning the weight allocation model space by vectorizing and fusing multi-source data; Three-dimensional dynamic geological model modeling module (4), which is based on the multi-source data fusion result and uses the irregular triangular network-generalized triangular prism model to perform three-dimensional dynamic geological model modeling; Three-dimensional dynamic transparent geological model acquisition and visualization module (5), which is used to iteratively optimize the obtained three-dimensional dynamic geological model, and acquire and output and visualize the three-dimensional dynamic transparent geological model.

Images