A method, medium, device and product for dynamic geological modeling of a drill-and-blast tunnel
By dividing tunnel engineering into different construction periods and constructing hierarchical three-dimensional geological models, and adopting a dynamic update method, the problem of difficulty in balancing accuracy and efficiency in traditional three-dimensional geological modeling is solved, and efficient and low-cost tunnel construction risk management and decision support are achieved.
Patent Information
- Application Number
- CN202610311039.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-14
- Publication Date
- 2026-08-04
AI Technical Summary
Traditional 3D geological modeling methods cannot be dynamically updated, making it difficult to balance the high precision requirements during construction with the efficiency requirements of long-term planning. The model is out of touch with the actual geological conditions, and the modeling cost is high, the accuracy is insufficient, and it is difficult to quickly provide feedback on new geological information.
The tunnel project is divided into multiple construction phases along the axis, and three-dimensional geological models of the tunnel with different precision and resolution are constructed. Non-uniform rational B-spline surfaces and Kriging interpolation are used to construct the stratigraphic interface model. The model is dynamically updated with the construction progress, and the configuration of modeling resources is optimized through hierarchical and incremental update methods.
It enables intelligent allocation of limited modeling resources, reduces modeling costs and computational burden, improves work efficiency and model timeliness during construction, provides high-precision real-time geological analysis support, and reduces construction risks.
Smart Images
Figure CN122510470A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tunnel geological modeling technology, and in particular to a dynamic geological modeling method, medium, equipment and products for drill-and-blast tunnels. Background Technology
[0002] Traditional 3D geological modeling relies heavily on geological survey data before construction. Once the model is established, it becomes a static reference and cannot be dynamically updated as construction progresses. However, underground geological conditions are highly uncertain and hidden, and the early model often deviates from the actual situation, which can easily lead to the risk of collapse or sudden water inrush.
[0003] Traditional modeling uses a uniform resolution across the entire area, making it difficult to balance the high-precision requirements of the current construction phase with the efficiency needs of long-term planning phases. This results in excessively high modeling costs, insufficient accuracy in key areas, and misleading construction decisions. New geological information revealed during construction is difficult to quickly feed back into the model, requiring manual intervention for updates, which is inefficient and lacks standardized update logic, leading to a disconnect between the model and actual geological conditions. Summary of the Invention
[0004] The purpose of this invention is to address the problem that existing methods struggle to simultaneously meet the high-precision requirements of the current construction phase and the efficiency demands of future planning phases. This invention proposes a dynamic geological modeling method for drill-and-blast tunnels, comprising the following steps: S1. Divide the tunnel project into multiple construction phases along the axis; S2. Based on the tunnel length and diameter range for different construction periods, construct three-dimensional geological models of the tunnel with different accuracy and resolution. The construction process of the three-dimensional geological model of the tunnel is as follows: Based on geological exploration data and borehole data, a three-dimensional model of the stratigraphic interface was constructed. Based on borehole core data, lithological information from advanced drilling, and geophysical data, lithological information is coupled to a three-dimensional model of the formation interface to form a three-dimensional lithological distribution. By associating the physical and mechanical parameters of soil and rock with the three-dimensional distribution of lithology, three-dimensional values of the physical and mechanical parameters are assigned, forming an attribute parameter field model. S3. The three-dimensional geological model of the tunnel is dynamically updated as the construction progresses.
[0005] Furthermore, based on geological exploration data and borehole data, a three-dimensional model of the formation interface is constructed using non-uniform rational B-spline surface representation or triangular mesh interpolation method.
[0006] Furthermore, based on borehole core data, lithological information from advanced drilling, and geophysical data, the Kriging interpolation algorithm is used to couple the lithological information to a three-dimensional model of the formation interface.
[0007] Furthermore, based on the tunnel length and diameter range for different construction periods, the modeling area is divided into three levels of sub-regions. Each sub-region constructs a 3D geological model of the tunnel with different precision and resolution. The three levels of sub-region division are as follows: First-level sub-region: The range of tunnel length and m times the tunnel diameter during the current construction period, where m is greater than 1; Second-level sub-region: The range of tunnel length and n times the tunnel diameter along the tunnel axis from the 2nd to the 4th construction phases, where n is greater than 1; Third-level sub-region: Other areas of the project; Three-dimensional geological models of the tunnel are constructed from the first to the third level sub-regions according to their accuracy and resolution, from high to low.
[0008] Furthermore, the confidence level of the tunnel 3D geological model for each sub-region is calculated using the following formula:
[0009] Where C represents the confidence level of the three-dimensional geological model of the tunnel. , and Represents the weight parameters. This represents the root mean square error of the formation interface. Indicates the accuracy of lithological classification. This represents the coefficient of variation of the attribute parameter.
[0010] Furthermore, the three-dimensional geological model of the tunnel in the first-level sub-region also needs to be embedded with the three-dimensional geological model of the adverse geological body.
[0011] The present invention also proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described dynamic geological modeling method for drill-and-blast tunnels.
[0012] The present invention also proposes an electronic device, including a processor and a memory, wherein the processor and the memory are interconnected, wherein the memory is used to store a computer program, the computer program including computer-readable instructions, and the processor is configured to invoke the computer-readable instructions to execute the above-described dynamic geological modeling method for drill-and-blast tunnels.
[0013] The present invention also proposes a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the above-described dynamic geological modeling method for drill-and-blast tunnels.
[0014] The beneficial effects of the technical solution provided by this invention are: (1) The tunnel is divided into modeling areas with different precision and resolution. Starting from the current construction period, the tunnel 3D geological model is divided into three levels of precision and resolution from high to low. The current construction section adopts high precision and resolution, the medium precision area covers the short-term prospect area, and the low precision area is used to describe the long-term geological background. This gradient precision layout realizes the intelligent allocation and intensive use of limited modeling resources, effectively avoiding the redundant calculation and time cost caused by using the same high-resolution model in the whole area, and taking into account the high precision requirements of the current construction section and the efficiency requirements of the long-term planning section.
[0015] (2) As the construction period progresses, the model is updated incrementally, eliminating the need to rebuild the global model with each update. This incremental modeling and updating method significantly reduces the computational burden of model maintenance, enabling the model to keep up with the construction progress at a lower cost. While ensuring the model's timeliness, it also greatly improves the efficiency of the entire modeling process, providing a technical path for real-time geological modeling and dynamic analysis of long-distance, deep-buried tunnels. Attached Figure Description
[0016] Figure 1 This is a flowchart of a dynamic geological modeling method for drill-and-blast tunnels according to an example of the present invention; Figure 2 This is a schematic diagram of an improved MAPS method according to an example of the present invention. Figure 2 (a) is a schematic diagram of mudstone interlayers. Figure 2 (b) is a schematic diagram of a 3×m smooth window; Figure 3 This is a block diagram of an electronic device according to an exemplary embodiment of the present invention. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described below with reference to the accompanying drawings.
[0018] A flowchart of the dynamic geological modeling method for drill-and-blast tunnels according to an example of the present invention is shown below. Figure 1 Specifically, it includes: S1. Divide the tunnel project to be modeled into multiple construction periods along the axis.
[0019] S2. Based on the tunnel length and diameter range for different construction periods, construct 3D geological models of the tunnel with different precision and resolution. Divide the modeling area into three levels of sub-regions according to the tunnel length and diameter range for different construction periods. Construct a 3D geological model of the tunnel with different precision and resolution for each sub-region. The 3D geological models of the tunnels are constructed from the first to the third level of sub-regions, with precision and resolution decreasing. The division method of the three levels of sub-regions is as follows: First-level sub-region (high precision and high resolution): along the tunnel axis, the tunnel length during the current construction period; on the cross section, the adjacent surrounding rock area covering the tunnel outline extension m times the tunnel diameter (m is a coefficient greater than 1, which can be set according to the rock mass stability, usually 1-3).
[0020] Second-level sub-region (medium precision and medium resolution): This region serves the mid-term construction planning and preliminary risk assessment. Its spatial range follows the high precision zone and extends along the tunnel axis to the tunnel length of the 2nd to 4th construction periods after the current construction period. The radial range of the surrounding rock is n times the tunnel diameter (m is a coefficient greater than 1, and n is usually 1-5).
[0021] Third-level sub-region (low precision and low resolution): Other areas of the project.
[0022] The tunnel's three-dimensional geological model includes a three-dimensional model of the stratigraphic interfaces, a three-dimensional distribution of lithology, and a property parameter field model. The process of constructing the tunnel's three-dimensional geological model is as follows: (1) Based on geological exploration data and borehole data, a three-dimensional model of the stratigraphic interface is constructed using non-uniform rational B-spline surface representation or triangular mesh interpolation method. If there are fault zones in the modeling area, they should also be represented in the model.
[0023] (2) Based on borehole core data, lithological information from advanced drilling and geophysical data, the Kriging interpolation algorithm is used to couple lithological information to a three-dimensional model of the formation interface to form a three-dimensional distribution of lithology.
[0024] Kriging interpolation is a weighted improvement algorithm for interpolating data within a known space. Essentially, it's an improvement on the IDW interpolation algorithm, better reflecting the actual spatial distribution characteristics of the interpolated surface. When calculating the weights, it uses a variogram function with distance as the independent variable to demonstrate the structure and random distribution of the variable in three-dimensional space. Depending on the application, there are various variations of Kriging interpolation, such as ordinary Kriging, logarithmic Kriging, co-Kriging, and generalized Kriging. The ordinary Kriging interpolation algorithm is as follows.
[0025]
[0026] in, For point The estimated value at that location, , The first The weight coefficients and values of each known point. satisfy And it satisfies the conditions of unbiased estimation. , For point If the actual value at that location is given, then the following system of equations exists:
[0027] in, , Points , The semi-mutation value, It is a Lagrange multiplier.
[0028] In the three-dimensional lithological distribution, an improved MAPS method is used to perform fuzzy boundary processing on the lithological transition zone. The principle of the improved MAPS method is as follows: Assume the single-layer mudstone distribution code in the lithofacies model is as follows: 0 represents the background or other lithofacies types, 1 represents mudstone interlayers, and the number of 1s is L, representing the mudstone length. Using... Figure 2 The smoothing of the 3×m smoothing window shown in Figure (b) Figure 2 The mudstone interlayer shown in (a) is represented by a smooth window of 3×m:
[0029] When the center point K of the window moves Figure 2 When the left end of the mudstone interlayer in (a) Figure 2 In (a), the red 1 on the left represents mudstone, with a weight of [value missing]. The weight of background or other lithofacies types is +sum('1'), where sum('1') represents the sum of all values of 1 within the smoothing window. This is to ensure the smoothing of the mudstone interlayers at both ends ( Figure 2 The red 1) will not be affected by other phase types ( Figure 2 In code 0), the weight of other lithologies should be less than that of mudstone. Therefore, the smoothing window should satisfy K > sum('1'). This ensures that the ends and middle of the mudstone interlayer are not smoothed out during the smoothing process, and the entire mudstone interlayer can be preserved intact. To ensure that the center point K of the window moves to... Figure 2 When the red 1 point is on the left side of (a), the window edge point exist Figure 2 The right side of (a) and the left side of the red 1 need to satisfy n+1. L, where m = 2 × n + 1, is needed for the above weights to be calculated correctly, and mudstone interlayers are retained. Therefore, if we want to avoid retaining mudstone interlayers shorter than L during the smoothing process, we must let n + 1 > L, and... >K-sum('1') ensures that the weights of all mudstone interlayers are less than the weights of the background phase during the window movement process, thus smoothing them out.
[0030] (3) The physical and mechanical parameters of the soil and rock mass (such as cohesion, internal friction angle, permeability coefficient, etc.) are associated with the three-dimensional distribution of lithology. The laboratory test and in-situ test data are integrated by random forest algorithm, and the three-dimensional assignment of physical and mechanical parameters is realized by Kriging interpolation method to form an attribute parameter field model.
[0031] Attribute field modeling employs a deep learning-based enhanced kriging method. The core idea is to directly input spatial coordinates into a neural network, allowing the network to learn a non-stationary mapping from spatial location to the target attribute value. Unlike traditional kriging, which relies on a fixed variogram, this method leverages the powerful fitting capabilities of deep neural networks to implicitly learn the complex, spatially varying covariance structure of the data. Its network architecture typically includes a "basis function layer" that maps spatial coordinates to a high-dimensional feature space, followed by fully connected layers that ultimately output predicted values. This approach is particularly suitable for handling geological attribute distributions with strong non-stationary characteristics.
[0032] A high-precision and high-resolution 3D geological model of the tunnel is constructed in the first-level sub-region, focusing on the key sections currently under construction and those to be constructed soon. This region is the core support for risk prevention and refined construction decisions. The required resolution is ≤0.5m, stratigraphic interface error is ≤0.2m, and lithological classification accuracy is ≥95%. Within this region, complex spatial interpolation algorithms such as Kriging and radial basis functions are used, integrating multi-source, high-density exploration data to construct a detailed 3D model. This model needs to characterize in detail the stratigraphic interfaces, lithological spatial distribution, and the physical and mechanical parameter fields of key soil / rock masses (such as elastic modulus, cohesion, and internal friction angle). Tunnel construction, especially in areas with adverse geological development such as karst and fault fracture zones, faces significant geological risks such as collapses, water inrushes, and rock bursts. Effective prevention and control of these risks heavily rely on accurate prediction of the geological conditions ahead. This invention embeds a three-dimensional geological model of adverse geological bodies into a high-precision and high-resolution three-dimensional geological model of a tunnel constructed in the first-level sub-region. This model, including aspects such as the spatial morphology of karst caves, the attitude and fracture zone range of faults, and the water level and permeability characteristics of water-rich areas, allows for refined three-dimensional spatial identification and positioning, treating these as independent entities or attribute fields. The three-dimensional geological model of adverse geological bodies utilizes data from advanced drilling (bore diameter ≥ 50mm) and geophysical exploration (e.g., high-density electrical resistivity tomography) to model karst caves (three-dimensional cavities), groundwater levels (dynamic hydraulic gradient fields), etc., with an error ≤ 0.3m. This three-dimensional visualization and quantitative representation of adverse geological bodies makes hidden geological risks measurable. Construction and design personnel can not only be aware of the existence of risk bodies in advance but also intuitively analyze their spatial relationship with the tunnel structure, assess their potential impact, and provide direct and accurate data and graphical support for targeted disaster prevention and control measures (e.g., advance grouting range, reinforced support methods, drainage scheme design) and timely adjustments to construction techniques (e.g., excavation method conversion, blasting parameter optimization). This changes the previous passive risk response model that mainly relied on experience-based judgment and local detection, and promotes the transformation of tunnel construction risk management towards proactive early warning and scientific decision-making based on full-cycle, high-precision digital models.
[0033] In the second-level sub-region, a high-precision and high-resolution 3D geological model of the tunnel is constructed, serving medium-term construction planning and preliminary risk assessment. A lithological classification accuracy of ≥85% is required. Within this region, computationally efficient algorithms such as inverse distance weighted (IDW) and linear interpolation are employed to generate the model using relatively sparse exploration data. This level of model focuses on reflecting major stratigraphic boundaries, macroscopic lithological distribution patterns, and spatial trends of some key physical and mechanical parameters, significantly reducing the computational burden while ensuring a certain level of reliability.
[0034] Construct a 3D tunnel geological model with low precision and resolution in the third-level sub-region, which covers the entire model background range outside the aforementioned two regions. Adopt the simplest interpolation or trend surface analysis method, only outlining the distribution contours of the regional stratigraphic framework and major lithological units, aiming to provide the global geological background with the least time and computational cost, and avoiding data redundancy and interference caused by irrelevant details.
[0035] Calculate the confidence level of the 3D tunnel geological model for each sub-region using the following formula:
[0036] where C represents the confidence level of the 3D tunnel geological model, 、 and represent weight parameters, represents the root mean square error of the stratigraphic interface, represents the accuracy rate of lithological classification, represents the coefficient of variation of attribute parameters.
[0037] S3. Dynamically incrementally update the 3D tunnel geological model according to the construction progress.
[0038] Every time a construction period of the tunnel is completed, new geological information is collected and entered into the dynamic database, automatically triggering the model update process: Taking the geological information revealed at the current completed construction period (e.g., the kth construction period) as the true value, calibrate the high-resolution and high-precision regional model corresponding to this construction period, and correct the stratigraphic interface, lithological distribution, and attribute parameters.
[0039] Taking the kth construction period as the benchmark, update the medium-resolution and medium-precision regional models corresponding to the (k + 1)th to (k + 3)th construction periods along the axis forward, increase the weight of the modeling data in this region, and optimize the interpolation results.
[0040] Except for the above calibration and update ranges, the remaining low-resolution regional models remain unchanged until they enter the medium-resolution or high-resolution regional ranges.
[0041] When the construction reaches the kth (k < x) construction period, where x is the total construction period, the model status is as follows: High-precision model area: Located in the tunnel axis direction, covering the completed (k - 1) construction periods. This area is fully constrained by the revealed geological information (face sketch, in-tunnel geophysical exploration, etc.), and the geological interface, rock mass grade, tectonic features, etc. are depicted most precisely.
[0042] Medium-precision model area: Located in front of the high-precision area, covering the prospective section of the 3 construction periods from the (k + 1)th to (k + 3)th along the tunnel axis direction. This area is mainly established based on the previous exploration and certain extrapolation rules, with moderate precision, providing a forward look for short-term construction decisions.
[0043] Low-precision model area: This area covers the remaining range of the model outside the above-mentioned areas. It relies on more macroscopic and sparse initial exploration data and has the lowest precision. It mainly maintains the continuity of the regional geological background.
[0044] Once the k-th construction phase is completed and all corresponding geological information of the tunnel face and construction survey data for that phase are acquired, the model initiates a complete update iteration. This upgrades the area adjacent to the excavated section within the original medium-precision modeling area, specifically the region along the length of the k-th construction phase, to a high-precision model. This process comprehensively utilizes the latest revealed geological profiles of the tunnel face, geophysical interpretation results from the construction phase, and possible borehole data to accurately reconstruct and calibrate the locations of geological interfaces, lithological boundaries, and the boundaries of adverse geological bodies.
[0045] The 3D geological body, k+3 construction period lengths following the original medium-precision area within the original low-precision modeling region, is upgraded to a medium-precision model. This update is based on the updated high-precision model boundary conditions and geological trends, combined with advanced forecast data (such as TSP and long-range ground-penetrating radar detection) for constrained extrapolation, thus updating the geological predictions ahead of mid-term excavation.
[0046] After the update, the original high-precision area remains high-precision; the new medium-precision area (upgraded from the original low-precision part) and the un-upgraded part of the original medium-precision area together constitute the new medium-precision look-ahead area; the remaining area remains the low-precision background area. The model as a whole "grows" forward along the tunnel axis and is optimized by one construction period step.
[0047] This invention employs a hierarchical model and an incremental model update method, achieving multiple advantages: First, it optimizes the allocation of modeling resources, concentrating computational resources and data density on the "front" and "nearby" areas most directly impacting construction safety and efficiency, overcoming the efficiency bottleneck caused by excessive modeling in irrelevant areas in traditional methods. Second, it ensures consistently forward-looking risk forecasts. The model is updated continuously as construction progresses, ensuring the highest accuracy geological forecasts are always available within at least a length L in front of the construction face, providing a reliable basis for dynamic design, construction process adjustments, and risk contingency planning. Third, it enhances the model's engineering practicality. The hierarchical model better meets the information needs of engineers at different levels of decision-making, providing matching support from precise support design near the working face to macro-level planning of medium- and long-term construction organization.
[0048] In one exemplary embodiment, a computer-readable storage medium is included, which stores a computer program that, when executed by a processor, implements the above-described dynamic geological modeling method for drill-and-blast tunnels.
[0049] Please see Figure 3In one exemplary embodiment, the device further includes an electronic device including at least one processor, at least one memory, and at least one communication bus.
[0050] The memory stores a computer program, which includes computer-readable instructions. The processor calls the computer-readable instructions stored in the memory through the communication bus to execute the aforementioned dynamic geological modeling method for tunnels using the drill-and-blast method.
[0051] In one exemplary embodiment, a computer program product is proposed, including a computer program / instruction that, when executed by a processor, implements the steps of the above-described dynamic geological modeling method for drill-and-blast tunnels.
[0052] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A dynamic geological modeling method for drill-and-blast tunnels, characterized in that, Includes the following steps: S1. Divide the tunnel project into multiple construction phases along the axis; S2. Based on the tunnel length and diameter range for different construction periods, construct three-dimensional geological models of the tunnel with different accuracy and resolution. The construction process of the three-dimensional geological model of the tunnel is as follows: Based on geological exploration data and borehole data, a three-dimensional model of the stratigraphic interface was constructed. Based on borehole core data, lithological information from advanced drilling, and geophysical data, lithological information is coupled to a three-dimensional model of the formation interface to form a three-dimensional lithological distribution. By associating the physical and mechanical parameters of soil and rock with the three-dimensional distribution of lithology, three-dimensional values of the physical and mechanical parameters are assigned, forming an attribute parameter field model. S3. The tunnel's three-dimensional geological model is dynamically and incrementally updated according to the construction progress.
2. The dynamic geological modeling method for drill-and-blast tunnels according to claim 1, characterized in that, Based on geological exploration data and borehole data, a three-dimensional model of the formation interface is constructed using non-uniform rational B-spline surface representation or triangular mesh interpolation method.
3. The dynamic geological modeling method for drill-and-blast tunnels according to claim 1, characterized in that, Based on borehole core data, lithological information from advanced drilling, and geophysical data, the Kriging interpolation algorithm is used to couple lithological information into a three-dimensional model of the formation interface.
4. The dynamic geological modeling method for drill-and-blast tunnels according to claim 1, characterized in that, Based on the tunnel length and diameter range at different construction stages, the modeling area is divided into three levels of sub-regions. Each sub-region constructs a 3D geological model of the tunnel with different precision and resolution. The division of the three levels of sub-regions is as follows: First-level sub-region: The range of tunnel length and m times the tunnel diameter during the current construction period, where m is greater than 1; Second-level sub-region: The range of tunnel length and n times the tunnel diameter along the tunnel axis from the 2nd to the 4th construction phases, where n is greater than 1; Third-level sub-region: Other areas of the project; Three-dimensional geological models of the tunnel are constructed from the first to the third level sub-regions according to their accuracy and resolution, from high to low.
5. The dynamic geological modeling method for drill-and-blast tunnels according to claim 4, characterized in that, The confidence level of the tunnel 3D geological model for each sub-region is calculated using the following formula: Where C represents the confidence level of the three-dimensional geological model of the tunnel. , and Represents the weight parameters. This represents the root mean square error of the formation interface. Indicates the accuracy of lithological classification. This represents the coefficient of variation of the attribute parameter.
6. The dynamic geological modeling method for drill-and-blast tunnels according to claim 4, characterized in that, The three-dimensional geological model of the tunnel in the first-level sub-region also needs to be embedded with the three-dimensional geological model of the adverse geological body.
7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.
8. An electronic device, characterized in that, The device includes a processor and a memory, the processor being interconnected with the memory, wherein the memory is used to store a computer program, the computer program including computer-readable instructions, and the processor is configured to invoke the computer-readable instructions to perform the method as described in any one of claims 1 to 6.
9. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the method described in any one of claims 1 to 6.