Method, device, equipment and program product for representing borehole fracture hydraulic structure with increased dimensionality

Abstract

The method comprises the following steps: constructing a borehole fracture geometry database; constructing a three-dimensional geometric model of the borehole fracture and surface structure characteristic data of the borehole fracture; based on the three-dimensional geometric model, the surface structure characteristic data and the geometric parameters, a fracture flow model of the borehole fracture is constructed, and hydraulic characteristic simulation parameters of the borehole fracture are simulated; pressure test data is obtained; the fracture flow model is corrected; based on the hydraulic characteristic test parameters and the corrected fracture flow model, a filling factor is backstepped; a plurality of sets of random combinations of fracture parameters are generated by using a Monte Carlo method; and the plurality of sets of random combinations of fracture parameters and the corresponding equivalent permeability are respectively taken as the input and label of a permeability prediction model for training, so that the trained permeability prediction model is used for the multidimensional characterization of the hydraulic structure of the borehole fracture in deep grouting design.

Description

Technical Field

[0001] This disclosure relates to the fields of geotechnical engineering and hydraulic engineering technology, specifically to a method, device, equipment, and program product for deep grouting design and for enhancing the characterization of borehole fracture hydraulic structures based on multi-source data fusion and artificial intelligence inversion. Background Technology

[0002] In water conservancy projects, dam foundation seepage prevention, underground cavern support and seepage prevention, deep rock engineering, and deep grouting reinforcement projects, rock fractures are the core structure controlling groundwater seepage, grout migration, and diffusion. Their spatial distribution, geometric morphology, interface characteristics, and hydraulic conductivity directly determine the grouting scheme design, grouting parameter selection, and engineering treatment effect. In current engineering practice, borehole drilling is usually the main detection method. Methods such as borehole television imaging, borehole acoustic scanning, and conventional water pressure tests are used to obtain basic information about fractures. This allows for the preliminary identification of the spatial location, attitude, aperture, and other geometric parameters of the fractures, and the construction of two-dimensional or simplified three-dimensional fracture geometric models, providing a foundation for grouting design.

[0003] However, the true seepage characteristics and injectability of rock fractures are not determined by the aperture alone, but are controlled by a combination of factors, including fracture interface roughness, type and density of filling material, fracture spatial distribution characteristics, in-situ stress field, and fracture surface contact state. Traditional borehole detection and analysis methods generally have significant limitations: on the one hand, they can only obtain two-dimensional geometric information of fractures, making it difficult to quantitatively characterize the three-dimensional morphology and roughness characteristics of fractures; on the other hand, they lack the mapping relationship between fracture geometric parameters and hydraulic characteristics and grouting controllability, making it impossible to accurately predict fracture seepage capacity and grout diffusion patterns.

[0004] In actual grouting projects, the aforementioned shortcomings often lead to problems such as inaccurate judgment of the grouting range, reliance on experience in selecting grouting pressure and grout mix ratio, excessive grout diffusion, or incomplete grouting. These issues not only affect the anti-seepage and reinforcement effects of the project but also easily cause material waste and project delays. With the rapid development of deep geotechnical engineering, smart water conservancy, and intelligent construction, the traditional experience-based "drilling observation-qualitative judgment-experience design" model is no longer sufficient to meet the requirements of high-precision and intelligent grouting design.

[0005] Therefore, the industry urgently needs a technical method that can upgrade crack information from two-dimensional geometric representation to multi-dimensional structural-hydraulic-controllability integrated representation. Summary of the Invention

[0006] This disclosure provides a method, apparatus, equipment, and program product for characterizing the dimensional enhancement of borehole fracture hydraulic structures.

[0007] In a first aspect, this disclosure provides a method for dimensional characterization of borehole fracture hydraulic structures, comprising: A borehole fracture geometric database is constructed based on image data of borehole fractures within the target engineering area; the borehole fracture geometric database includes the geometric parameters of the borehole fractures; Based on the surface point cloud data of the borehole fracture, a three-dimensional geometric model of the borehole fracture and surface structure feature data of the borehole fracture are constructed, the surface structure feature data including the roughness parameter of the borehole fracture. Based on the aforementioned three-dimensional geometric model, surface structure feature data, and geometric parameters, a fracture flow model of the borehole fracture is constructed to simulate the hydraulic characteristics of the borehole fracture, including the equivalent permeability of the borehole fracture. A water pressure test is conducted on the boreholes within the target engineering area to obtain water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fractures; the hydraulic characteristic test parameters include the test permeability of the borehole fractures; The fracture flow model is corrected by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. Based on the hydraulic characteristic test parameters and the corrected fracture flow model, the filling factor is calculated by reverse calculation. Based on the geometric parameters, surface structure feature data, and the filling factor, multiple sets of random combinations of fracture parameters are generated using the Monte Carlo method. The equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters is obtained using the corrected fracture flow model. The multiple sets of fracture parameters are randomly combined and their corresponding equivalent permeability is used as the input and label of the permeability prediction model for training. The permeability prediction model obtained by training is used for the dimensional characterization of the hydraulic structure of the borehole fracture in deep grouting design.

[0008] Furthermore, the method also includes: The hydraulic structure index of the borehole fracture is calculated using the geometric parameters, roughness parameters, filler factor, and hydraulic pressure test parameters. The grouting controllability index of the borehole fracture is calculated using the hydraulic structure index and the grouting engineering adjustment function. The borehole fractures are classified into different types to guide grouting design using the grouting controllability index.

[0009] Furthermore, the method also includes: For other borehole fractures within the target engineering area, the geometric parameters, surface structure feature data, and hydraulic characteristic test parameters of the other borehole fractures are obtained using the above steps. The filling factor of the other borehole fractures is obtained by back-calculating using the hydraulic characteristic test parameters of the other borehole fractures and the corrected fracture flow model; The permeability prediction model obtained through training is used to process the geometric parameters, surface structure feature data, and filling factor of the other borehole fractures to obtain the equivalent permeability of the other borehole fractures.

[0010] Furthermore, each of the multiple sets of random combinations of fracture parameters includes multiple sets of random combinations of fracture parameters generated using the Monte Carlo method based on the geometric parameters, surface structure feature data, and the filling factor, including: The aperture distribution of the borehole fracture is determined using the geometric parameters, and the aperture distribution includes the mean, standard deviation, and distribution type of the aperture. The roughness statistical distribution of the borehole fracture is calculated using the roughness parameters. The roughness statistical distribution includes the mean, fluctuation range, and distribution type of JRC. Determine the filler factor distribution, which includes the value range and distribution type of the filler factor; Using the Monte Carlo method, multiple sets of random combinations of fracture parameters are generated by randomly sampling from the aperture distribution, roughness statistical distribution and filling factor distribution. Each set of combinations includes an aperture value, a JRC value and a filling factor.

[0011] Furthermore, the geometric parameters include the spatial location, orientation, dip angle, and initial aperture of the fracture; and / or, The roughness parameters include the joint roughness coefficient JRC, fractal dimension, and surface power spectral density PSD; and / or, The hydraulic characteristic simulation parameters also include the local velocity distribution and pressure loss characteristics of the borehole fractures; and / or, The hydraulic characteristic test parameters also include the injection pressure and flow rate of the borehole fracture; and / or, The value range of the filling factor is 0 to 1, where 0 indicates that the borehole fracture is completely filled with dense, impermeable material, 1 indicates that the borehole fracture is an open fracture without filling, and other values ​​indicate that the borehole fracture is in a partially filled state.

[0012] Furthermore, the fracture flow model is corrected by comparing and analyzing the simulated hydraulic characteristics parameters and the experimental hydraulic characteristics parameters, including: Replace the opening in the fracture flow model with the equivalent hydraulic opening calculated based on the hydraulic characteristic test parameters; Adjust the inlet and outlet pressure boundary conditions of the fracture flow model; Adjust the roughness correction coefficient or local resistance coefficient to control the error between the simulated hydraulic characteristics parameters and the corresponding experimental hydraulic characteristics parameters within a set threshold range, thereby establishing the measured constraints of fracture hydraulic parameters composed of the injection pressure-flow correspondence, permeability range, and equivalent hydraulic opening.

[0013] Secondly, this disclosure provides a borehole fracture hydraulic structure dimensional enhancement characterization device, comprising: The first construction module is configured to construct a borehole fracture geometric database based on image data of borehole fractures within the target engineering area; the borehole fracture geometric database includes the geometric parameters of the borehole fractures; The second construction module is configured to construct a three-dimensional geometric model of the borehole fracture and surface structural feature data of the borehole fracture based on the surface point cloud data of the borehole fracture. The surface structural feature data includes the roughness parameters of the borehole fracture. The simulation module is configured to simulate the hydraulic characteristics of the borehole fracture based on the three-dimensional geometric model, surface structure feature data, and geometric parameters by constructing a fracture flow model of the borehole fracture; the hydraulic characteristics simulation parameters include the equivalent permeability of the borehole fracture. The first acquisition module is configured to conduct a water pressure test on the borehole in the target engineering area and acquire water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fracture; the hydraulic characteristic test parameters include the test permeability of the borehole fracture; The calibration module is configured to calibrate the fracture flow model by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. The reverse calculation module is configured to back-calculate the filling factor based on the hydraulic characteristic test parameters and the corrected fracture flow model. The generation module is configured to generate multiple sets of random combinations of fracture parameters based on the geometric parameters, surface structure feature data, and the filling factor using the Monte Carlo method. The second acquisition module is configured to use the corrected fracture flow model to obtain the equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters. The training module is configured to use the random combination of the multiple sets of fracture parameters and the corresponding equivalent permeability as inputs and labels for training the permeability prediction model. The permeability prediction model obtained by training is used for the dimensional characterization of the borehole fracture hydraulic structure in deep grouting design.

[0014] The function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above function.

[0015] In one possible design, the above-described device includes a memory and a processor. The memory stores one or more computer instructions that support the device in performing the corresponding methods described above, and the processor is configured to execute the computer instructions stored in the memory. The device may also include a communication interface for communicating with other devices or communication networks.

[0016] Thirdly, embodiments of this disclosure provide an electronic device including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the method described in any of the above aspects.

[0017] Fourthly, embodiments of this disclosure provide a computer-readable storage medium for storing computer instructions used by any of the above-described devices, which, when executed by a processor, are used to implement the methods described in any of the above aspects.

[0018] Fifthly, embodiments of this disclosure provide a computer program product comprising computer instructions which, when executed by a processor, are used to implement the methods described in any of the preceding aspects.

[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0020] Other features, objects, and advantages of this disclosure will become more apparent from the following detailed description of non-limiting embodiments, taken in conjunction with the accompanying drawings. In the drawings: Figure 1 A flowchart is shown for a method for characterizing the dimensionality of borehole fracture hydraulic structures according to an embodiment of the present disclosure.

[0021] Figure 2 A flowchart illustrating an example of borehole fracture augmentation characterization according to an embodiment of the present disclosure is shown.

[0022] Figure 3 A three-dimensional scanning schematic diagram of a fracture interface according to an embodiment of the present disclosure is shown.

[0023] Figure 4 This is a schematic diagram of an electronic device suitable for implementing a method for characterizing the hydraulic structure of a borehole fracture according to an embodiment of the present disclosure.

[0024] Figure 5 A schematic diagram of a fracture hydraulic structure augmentation model according to an embodiment of the present disclosure is shown. Detailed Implementation

[0025] In the following, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings to enable those skilled in the art to readily implement them. Furthermore, for clarity, portions unrelated to the description of the exemplary embodiments have been omitted from the drawings.

[0026] In this disclosure, it should be understood that terms such as “comprising” or “having” are intended to indicate the presence of features, figures, steps, behaviors, components, parts or combinations thereof disclosed in this specification, and do not preclude the possibility of the presence or addition of one or more other features, figures, steps, behaviors, components, parts or combinations thereof.

[0027] It should also be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0028] To overcome the shortcomings of existing technologies, such as the limited acquisition of two-dimensional geometric information by fracture detection, reliance on experience-based grouting, and the disconnect between borehole information and grouting design, this disclosure provides a method for enhancing the dimensionality of borehole fracture hydraulic structure for deep grouting design. This method constructs the fracture hydraulic structure index (HSI) and the fracture grouting controllability index (GCI) through multi-source data fusion, 3D modeling, numerical simulation, and artificial intelligence inversion. This achieves a quantitative connection mechanism between precise fracture characterization and grouting parameter design, forming an intelligent process of "borehole detection - 3D modeling - hydraulic analysis - AI inversion - grouting optimization." It builds a precise bridge between borehole information and grouting design, improving the accuracy, reliability, and intelligence level of deep grouting design.

[0029] The details of the embodiments of this disclosure are described in detail below through specific examples.

[0030] Figure 1 A flowchart illustrating a method for characterizing the dimensional augmentation of borehole fracture hydraulic structures according to an embodiment of this disclosure is shown. Figure 1 As shown, the method for dimensional augmentation characterization of borehole fracture hydraulic structures includes the following steps: In step S101, a borehole fracture geometric database is constructed based on image data of borehole fractures within the target engineering area; the borehole fracture geometric database includes the geometric parameters of the borehole fractures; In step S102, based on the surface point cloud data of the borehole fracture, a three-dimensional geometric model of the borehole fracture and surface structure feature data of the borehole fracture are constructed, the surface structure feature data including the roughness parameters of the borehole fracture. In step S103, based on the three-dimensional geometric model, surface structure feature data, and geometric parameters, a fracture flow model of the borehole fracture is constructed to simulate the hydraulic characteristics of the borehole fracture, and the hydraulic characteristics simulation parameters include the equivalent permeability of the borehole fracture. In step S104, a water pressure test is conducted on the borehole within the target engineering area to obtain water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fracture; the hydraulic characteristic test parameters include the test permeability of the borehole fracture; In step S105, the fracture flow model is corrected by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. In step S106, the filling factor is calculated by back-calculating based on the hydraulic characteristic test parameters and the corrected fracture flow model. In step S107, based on the geometric parameters, surface structure feature data, and the filling factor, multiple sets of random combinations of crack parameters are generated using the Monte Carlo method; In step S108, the equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters is obtained using the corrected fracture flow model; In step S109, the multiple sets of fracture parameters are randomly combined and the corresponding equivalent permeability is used as the input and label of the permeability prediction model for training. The permeability prediction model obtained by training is used for the dimensional characterization of the borehole fracture hydraulic structure in deep grouting design.

[0031] The following examples illustrate in detail the method for dimensional characterization of borehole fracture hydraulic structures for deep grouting design proposed in this disclosure, such as... Figure 2 As shown, the method includes the following steps: Step 1: Borehole Image Acquisition Drill holes are arranged in the target engineering area, and image data of the fractures in the boreholes are obtained through borehole television or borehole imaging system. Based on the image data, the spatial location, orientation and dip angle, initial aperture and other geometric parameters of the fractures in the target engineering area are identified, and a borehole fracture geometric database is established to provide a basic geometric basis for subsequent fracture characterization.

[0032] For example: Boreholes are drilled at predetermined intervals in the target area of ​​the dam foundation, with a depth of 50–80 m. A borehole television imaging system is used to acquire images of the entire borehole interior, with a resolution of no less than 1080P and an acquisition speed of 0.5 m / min. The acquired image data is processed using image analysis software to identify 12 main fractures within the borehole. The spatial location of each fracture is recorded (including borehole depth, i.e., the length of the borehole axis from the borehole opening to the fracture location; and borehole circumference angle, i.e., the angle of the fracture relative to the reference direction on the borehole cross-section), orientation (azimuth angle of the intersection of the fracture surface and the horizontal plane, ranging from 30° to 150°), dip angle (the maximum angle between the fracture surface and the horizontal plane, ranging from 45° to 85°), and initial opening (the average vertical distance between the two walls of the fracture under natural conditions, ranging from 0.2 to 1.5 mm). A borehole fracture geometric database is established to store the basic geometric parameters of each fracture. For the fracture filling state (the degree to which the fracture interior is occupied by the filling material and its physical properties), the identification is aided by borehole acoustic scanning or core observation, which serves as the basic parameter for subsequent uncertainty analysis.

[0033] Step 2: 3D scanning of the fracture interface The fracture interface exposed by the borehole is subjected to three-dimensional scanning. Point cloud data of the fracture surface is obtained using a three-dimensional scanning device. A three-dimensional geometric model of the fracture interface is constructed based on the point cloud data to realistically restore the fracture surface morphology. The roughness parameters of the fracture interface are calculated based on the point cloud data, including but not limited to joint roughness coefficient (JRC), fractal dimension, and surface power spectral density (PSD), forming a fracture surface structural feature database and quantifying the structural characteristics of the fracture interface.

[0034] For example: In boreholes arranged in the target area of ​​the dam foundation, when the drill bit passes through a fractured zone, core samples containing natural fracture surfaces are obtained. Figure 3 (Mark 2 shows the rough fracture surface obtained after splitting the cylindrical rock core specimen). The rock core is an irregular cylinder, approximately 50 mm in diameter and 100–150 mm in length. Its surface naturally exhibits the undulating morphology, roughness characteristics, and infill material distribution of the actual underground fractures. Three typical fractures (aperture 0.5 mm, 0.8 mm, and 1.2 mm) were selected for three-dimensional scanning to cover the common fracture aperture ranges in engineering. (Refer to...) Figure 3 A high-precision three-dimensional laser scanning device 1 was used, with a scanning accuracy of ±0.01mm, to acquire point cloud data of the crack surface 2, with a point cloud density of not less than 100 points / cm². 2The point cloud data was denoised and stitched using point cloud processing software to construct three-dimensional geometric models of the fracture interfaces corresponding to the selected typical fractures. The roughness parameters of the three selected fractures were calculated based on the point cloud data. The JRC value was 6-12, the fractal dimension was 1.2-1.6, and the peak value of the PSD spectrum corresponded to wavelengths of 0.5-2cm, forming a database of geometric structural features of the fracture surface. The fracture geometric structural feature data included (1) basic geometric parameters: spatial position (hole depth, hole perimeter angle), attitude (strike, dip, dip angle), and geometric aperture. (2) surface structural parameters: joint roughness coefficient JRC, fractal dimension, surface power spectral density PSD, undulation amplitude, etc.

[0035] Step 3: Numerical simulation of fracture hydraulic characteristics Based on the three-dimensional geometric model and geometric structural feature data of the fracture obtained in step 2, a fracture seepage model is established in numerical simulation software such as the finite element method. This model can also be called a numerical simulation model. (Using COMSOL...) Taking Multiphysicsmo finite element software as an example, the specific operation is as follows: First, import the three-dimensional geometric model of the fracture interface obtained by scanning into COMSOL finite element software, and set the model boundary conditions and material properties according to the geometric parameters and roughness parameters obtained in steps 1 and 2. Since the original point cloud data may contain holes or defects, geometric repair is required to ensure that the model is closed and free of self-intersections. Then, extract the fluid computation domain (i.e., the space between the two walls of the fracture) and select an appropriate physical field interface (such as laminar flow or fracture flow interface) according to the flow characteristics. Then, apply pressure or velocity boundary conditions at the fracture inlet and outlet according to the actual engineering situation. Finally, perform mesh generation (manual densification is required at rough fractures), and set the solution type (such as steady state or transient) in the solver and click Calculate. The software will automatically iterate to solve the Navier-Stokes equation or Darcy equation until convergence, thereby outputting the numerical simulation results of the fracture such as equivalent permeability, velocity field, and pressure distribution, including but not limited to hydraulic characteristic simulation parameters such as equivalent permeability, local velocity distribution, and pressure loss characteristics, clarifying the intrinsic relationship between the fracture geometry and hydraulic characteristics.

[0036] In some embodiments, this intrinsic correlation is established by comparing and analyzing multiple sets of numerical simulation results (e.g., changing geometric parameters such as fracture aperture, JRC value, and fractal dimension respectively) to create a quantitative mapping relationship between geometric parameters and hydraulic parameters. This includes, but is not limited to, regression equations, empirical formulas, or sensitivity curves (e.g., the percentage decrease in equivalent permeability for every 1 increase in JRC value; the rate of decrease in pressure loss for every 0.1 mm increase in aperture). Thus, for a fracture with a specific geometric shape (e.g., aperture, roughness, undulation amplitude, fractal dimension, etc.), the corresponding equivalent hydraulic parameters (i.e., hydraulic characteristic simulation parameters) can be determined based on this quantitative mapping relationship. For example, simulation results may show that for every 1 increase in JRC value, the equivalent permeability decreases by approximately 15%; or that a higher roughness coefficient results in more pronounced local eddies and greater pressure loss.

[0037] In other embodiments, this intrinsic correlation can be used to: ① provide physical constraints on parameter values ​​for the Monte Carlo simulation in step 5; ② serve as prior knowledge for the AI ​​inversion model in step 6, verifying the physical rationality of the model's predictions; ③ directly estimate fracture hydraulic parameters quickly in engineering scenarios where high-precision prediction is not required, improving design efficiency. Simultaneously, the AI ​​model in step 6 automatically learns the mapping relationship between input and output parameters through training. This mapping relationship should maintain physical consistency with the intrinsic correlation revealed in step 3, thereby ensuring the reliability and interpretability of the AI ​​model.

[0038] For example: Using finite element numerical simulation software, based on the three-dimensional geometric model of the fracture interface constructed in step 2, a fracture seepage model 3 is established, with the fluid set as room temperature clean water with a density of 1000 kg / m³. 3 The dynamic viscosity is 1.002×10⁻⁶. -3 Pa·s; Through fluid dynamics calculations, the equivalent permeability of each fracture was obtained as 1.2 × 10⁻⁶ Pa·s. -6 ~5.8×10 -6 The flow rate was m / s, the local velocity distribution range was 0.01 to 0.08 m / s, and the pressure loss was 0.05 to 0.3 MPa / m, clarifying the correlation between fracture geometry and hydraulic characteristics.

[0039] Step 4: Data Fusion from Pressure Water Test A water pressure test is conducted in the borehole. The measured data of injection pressure, flow rate, and permeability of the fracture are obtained strictly according to engineering test specifications. This data includes hydraulic characteristic test parameters. The measured data of the water pressure test are compared and analyzed with the numerical simulation results from step 3. The numerical simulation model and equivalent hydraulic parameters in step 3 are corrected to establish measured constraints for the fracture hydraulic parameters and improve the accuracy of hydraulic parameter calculations. It should be noted that the model correction includes: first, the correction of the geometric parameters and boundary conditions of the numerical simulation model, such as inferring the equivalent hydraulic opening from the water pressure test data (rather than directly using the geometric opening), and adjusting the inlet and outlet pressure or flow boundary conditions to match the test conditions; second, the correction of the physical parameters of the numerical simulation model, such as adjusting the roughness correction coefficient and local resistance coefficient to match the equivalent permeability calculated by the simulation with the permeability coefficient converted from the water pressure test. Correcting equivalent hydraulic parameters means using the measured data from the water pressure test as the "true value" and then adjusting the uncertain parameters in the model (such as opening degree, roughness influence factor, etc.) until the error between the simulation results and the test results is controlled within an acceptable range.

[0040] It should also be noted that, in some embodiments, the correlation between injection pressure and injection flow rate (PQ curve), permeability (Lu value) and its nonlinear characteristics obtained from the pressure test, as well as the equivalent hydraulic aperture calculated based on the cubic law, can be used as measured constraints for the numerical simulation model. These constraints limit the reasonable range of values ​​for uncertain parameters such as equivalent permeability, fracture aperture, and roughness correction coefficient in the model, ensuring that the flow rate calculated by the corrected fracture flow model under the same pressure conditions matches the measured flow rate, and the simulated permeability is consistent with the measured permeability, thereby ensuring that the numerical simulation results truly reflect the actual hydraulic behavior of the fracture. Furthermore, based on the nonlinear characteristics of the PQ curve from the pressure test and the difference between permeability and equivalent hydraulic aperture, the fracture filling factor F can be inversely derived: when the PQ curve exhibits significant nonlinearity (i.e., the flow rate increase gradually decreases as pressure increases) and the equivalent hydraulic aperture is significantly smaller than the geometric aperture, it indicates the presence of filling material within the fracture. The filling factor F can be calculated using the formula F=b h / b g Estimate, where b h For the equivalent hydraulic opening (calculated from the pressure test), b g The geometric opening is measured in step 1; the value of F ranges from 0 to 1. The closer F is to 0, the denser the filling; the closer F is to 1, the less filling or no filling. This filling factor serves as an important input parameter for the subsequent Monte Carlo simulation and HSI index calculation in step 5.

[0041] For example: A water pressure test is conducted in the drilled borehole using the single-point method. The test pressure is 1.0 MPa, and the pressure stabilization time is no less than 30 minutes. The injection flow rate is recorded, and the permeability is calculated to be 0.5–2.8 Lu. Based on the three-dimensional geometric model of the fracture constructed in step 2, a single-fracture seepage model is established in the finite element software: the space between the two walls of the fracture is discretized into a computational mesh, a laminar flow physics field interface is selected, and the fluid (water) density is assigned 1000 kg / m³. 3 Dynamic viscosity 1.002×10 -3 A water injection pressure of 1.0 MPa was applied at the inlet of the fracture, with the outlet pressure set to atmospheric pressure and the wall surface set to no-slip condition. The velocity and pressure fields under this pressure were obtained by solving the Navier-Stokes equations, and the equivalent permeability and corresponding flow rate were then calculated. The permeability data and measured flow rate obtained from the pressure test were compared with the numerical simulation results. With the goal of controlling the error between the two within 5%, the boundary conditions in the model (such as the actual applied inlet pressure and outlet pressure) and equivalent hydraulic parameters (such as the equivalent hydraulic aperture and roughness correction coefficient) were adjusted to match the simulation results with the measured results. This established the measured constraints of the fracture hydraulic parameters, consisting of the water injection pressure-flow rate correspondence, the permeability range, and the equivalent hydraulic aperture.

[0042] In some embodiments, the fracture flow model is calibrated by comparing and analyzing the simulated hydraulic characteristics parameters and the experimental hydraulic characteristics parameters, including: Replace the opening in the fracture flow model with the equivalent hydraulic opening calculated based on the hydraulic characteristic test parameters; Adjust the inlet and outlet pressure boundary conditions of the fracture flow model; Adjust the roughness correction coefficient or local resistance coefficient to control the error between the simulated hydraulic characteristics parameters and the corresponding experimental hydraulic characteristics parameters within a set threshold range, thereby establishing the measured constraints of fracture hydraulic parameters composed of the injection pressure-flow correspondence, permeability range, and equivalent hydraulic opening.

[0043] Step 5: Monte Carlo stochastic simulation To address the uncertainties in parameters such as fracture roughness, aperture, and filling state, the Monte Carlo method is used to generate multiple random combinations of fracture parameters. Hydraulic simulation calculations are performed on each parameter combination to obtain the probability distribution of fracture seepage capacity and quantify the impact of fracture parameter uncertainties on seepage characteristics.

[0044] It should be noted that the Monte Carlo simulation in step 5 is mainly used for: ① generating a large number of training samples covering the range of parameter uncertainty during the AI ​​model training phase; ② pre-calculating the probability distribution, sensitivity coefficient, and design thresholds of fracture hydraulic parameters (such as the grouting material selection criteria corresponding to the P90 value). Once the above probability distribution and statistical results are established, they can be stored as a knowledge base. In practical engineering applications, when predicting the equivalent permeability of new fractures using the AI ​​model in step 6, it is not necessary to repeat the Monte Carlo simulation in step 5. Instead, the pre-calculated statistical results can be directly called (such as determining the quantile of the current predicted value, whether it exceeds the design threshold, etc.), providing a fast and reliable decision-making basis for grouting design.

[0045] For example, considering the statistical distribution of the fracture aperture (fluctuation range ±0.1 mm) measured in step 1, the JRC value (fluctuation range ±1 mm) calculated in step 2, and the infill factor (0.1–0.8) back-derived after pressure water test correction in step 4, determine the distribution type of each uncertainty parameter (such as aperture, roughness influence factor, etc.). Using these parameters as input, generate 1000 sets of random combinations of fracture parameters using the Monte Carlo method. The specific process is as follows: First, based on the mean and standard deviation of the aperture obtained from the statistical analysis of fracture geometric parameters in step 1, the statistical distribution of the roughness parameters calculated in step 2, and the distribution of the infill factor back-derived after pressure water test correction in step 4, determine the distribution type of each uncertainty parameter (aperture usually follows a normal distribution, while JRC and infill factor follow a uniform or triangular distribution); then, randomly sample from each distribution using the Monte Carlo method to generate 10 1000 sets of parameter combinations were used, each containing three parameters: aperture, JRC, and filling factor. Each parameter combination was then input into the numerical simulation model corrected in step 4, and the corresponding equivalent permeability was calculated using fluid dynamics. Finally, statistical analysis was performed on these 1000 simulation results to obtain the probability distribution of fracture seepage capacity, such as the probability density function, cumulative distribution function, mean, standard deviation, and specific quantiles (e.g., P10, P50, P90 values) of the equivalent permeability. The mean of the equivalent permeability was 3.5 × 10⁻⁶. -6 m / s, standard deviation is 0.8×10 -6m / s, the influence of quantitative parameter uncertainty on seepage characteristics. The probability distribution of the fracture seepage capacity specifically includes the following: (1) Probability density function (PDF): describes the relative probability of equivalent permeability (or other hydraulic parameters, such as equivalent hydraulic aperture) in each value interval, usually presented in the form of histogram or smooth curve. (2) Cumulative distribution function (CDF): describes the cumulative probability of equivalent permeability being less than or equal to a certain threshold, used to query the permeability value corresponding to a specific quantile. (3) Statistical characteristic values: mean (mathematical expectation): reflects the central trend of equivalent permeability; standard deviation: measures the degree of uncertainty of the parameter; variance: the square of the standard deviation; coefficient of variation (standard deviation / mean): a dimensionless uncertainty index; skewness and kurtosis: describe the symmetry and tail thickness of the distribution. (4) Specific quantiles: P10, P50, P90 values ​​(or other quantiles such as P5, P95), representing the equivalent permeability values ​​corresponding to cumulative probabilities of 10%, 50%, and 90%, respectively. The P90 value is often used in conservative engineering designs (i.e., there is a 90% certainty that the actual penetration rate will not exceed this value). (5) Confidence interval: The range of equivalent penetration rate values ​​at a given confidence level (e.g., 90% or 95%). (6) Exceedance probability curve: Describes the probability that the equivalent penetration rate exceeds a certain threshold, and is directly used for risk assessment. The above probability distribution results can be obtained through statistical analysis based on the 1000 sets of equivalent penetration rate simulation values ​​generated in step 5.

[0046] The resulting probability distribution of fracture seepage capacity is used for the following purposes: ① quantifying the impact of uncertainties in fracture aperture, roughness, and filling state on seepage capacity; ② providing a reliable basis for grouting design, such as calculating equivalent permeability exceeding 1×10⁻⁻⁻⁶. 5 ③ Assess the grouting and sealing risk by determining the probability of m / s; ④ Use the probability distribution as the label distribution of the training samples and input it into the artificial intelligence inversion model in step 6 to improve the robustness of the inversion to parameter uncertainties; ⑤ Determine the grouting material selection threshold. When the P90 value of the equivalent permeability exceeds the set value, automatically recommend the use of chemical grout instead of ordinary cement grout, thereby achieving risk-based grouting scheme optimization.

[0047] It should be noted that the probability distribution in step 5 is pre-calculated offline (for typical fractures or engineering areas). After the equivalent permeability prediction model is trained, it is not necessary to rerun the Monte Carlo simulation in actual engineering applications. However, the pre-calculated statistical results from step 5 (such as the five weighting coefficients obtained from sensitivity analysis, the P90 threshold, and the exceedance probability curve) can be used as design aids. These statistical results (such as "a 0.1 mm change in aperture leads to a 30% change in permeability" or "permeability exceeds 1×10⁻⁶") can be used to support the design. -5 The probability of m / s is 12%; the P90 value is 8 × 10 -6The "Recommended Chemical Slurry at m / s" can serve as a knowledge base and be invoked when applying AI models to make design decisions.

[0048] In summary, the implementation process of step 5 above can be described as follows: In some embodiments, each of the multiple sets of random combinations of fracture parameters includes generating multiple sets of random combinations of fracture parameters using the Monte Carlo method based on the geometric parameters, surface structure feature data, and the filling factor, including: The aperture distribution of the borehole fracture is determined using the geometric parameters, and the aperture distribution includes the mean, standard deviation, and distribution type of the aperture. The roughness statistical distribution of the borehole fracture is calculated using the roughness parameters. The roughness statistical distribution includes the mean, fluctuation range, and distribution type of JRC. Determine the filler factor distribution, which includes the value range and distribution type of the filler factor; Using the Monte Carlo method, multiple sets of random combinations of fracture parameters are generated by randomly sampling from the aperture distribution, roughness statistical distribution and filling factor distribution. Each set of combinations includes an aperture value, a JRC value and a filling factor.

[0049] Step 6: Artificial Intelligence Inversion An equivalent permeability prediction model is established using machine learning algorithms (such as neural networks or random forests). The model takes fracture geometric parameters, roughness parameters, and measured data from water pressure tests as inputs and fracture equivalent hydraulic parameters as outputs. Through model training, the model achieves the fusion of multi-source data and parameter inversion, completes the dimensionality representation of borehole fracture information, and realizes the accurate characterization of fracture hydraulic structure.

[0050] For example: (1) Based on the 1000 sets of fracture parameter combinations (each set includes aperture, JRC, and filling factor) generated by the Monte Carlo simulation in step 5 and their corresponding equivalent permeability calculation results, a training dataset is constructed. The probability distribution of the equivalent permeability in this dataset (including mean, standard deviation, P10, P50, P90, etc.) has been obtained by statistical analysis in step 5. This dataset is used as the training sample, where the input parameters are fracture geometric parameters, roughness parameters and measured data from the water pressure test, and the output parameter is the equivalent permeability. Since the training data fully covers the uncertainty range of each parameter and the output results retain the original probability distribution characteristics, the trained AI model naturally has robustness to parameter uncertainty. (2) Based on the same typical fracture, a BP neural network model is used, with the 1000 sets of simulation results generated in step 5 as the training data. Each set of simulation results includes input features (fracture aperture, JRC value, and filling factor) and output label (equivalent permeability). 800 sets of simulation results were selected as the training set and 200 sets as the test set. The model was trained until the error of the test set was less than 3%. The multi-source data fusion and parameter inversion were realized through the trained BP neural network model to complete the dimensionality enhancement of the fractured hydraulic structure.

[0051] It should also be noted that if more than one typical fracture is selected, such as the three fractures illustrated in step 2 above, then each of these three typical fractures (apertures of 0.5 mm and 1.2 mm) corresponds to 1000 sets of parameters. Therefore, all 3000 sets of parameters can be used to train the same AI model. In this case, the 3000 sets of parameters corresponding to these three fractures are merged to form a comprehensive training dataset covering different aperture ranges. A BP neural network model is used, with the fracture geometric parameters (aperture), roughness parameters (JRC), and filling factor in this dataset as input features, and the equivalent permeability as the output label. 80% of the data (2400 sets) is randomly selected as the training set, and 20% of the data (600 sets) is used as the test set. The model is trained until the error on the test set is less than 3%. Through the trained BP neural network model, a rapid mapping from fracture geometric parameters and roughness parameters to equivalent hydraulic parameters is achieved, completing the dimensionality-enhanced representation of the fracture hydraulic structure.

[0052] Step 7: Characterization of deep grouting parameters Based on the BP neural network model trained in step 6, the geometric parameters (direction, dip angle, aperture), roughness parameters (JRC, fractal dimension, PSD) and pressure test data of the fracture to be predicted are input into the model. The model automatically outputs the equivalent hydraulic parameters of the fracture (including equivalent permeability and equivalent hydraulic aperture). The input geometric parameters, roughness parameters, and the equivalent hydraulic parameters output by the model are combined to construct the dimensionality-enhanced characterization result of the fracture, and the grout diffusion potential index is calculated based on this. The characterization result is applied to the design of deep grouting engineering, providing a scientific basis for grouting parameter optimization and grout diffusion path prediction.

[0053] In the above embodiments, a fracture flow model was constructed by selecting a small number of typical borehole fractures, and a permeability prediction model was trained. At the same time, some other data required for subsequent practical engineering applications were also analyzed during the process.

[0054] In practical engineering applications, in some embodiments, the above method further includes: The hydraulic structure index of the borehole fracture is calculated using the geometric parameters, roughness parameters, filler factor, and hydraulic pressure test parameters. The grouting controllability index of the borehole fracture is calculated using the hydraulic structure index and the grouting engineering adjustment function. The borehole fractures are classified into different types to guide grouting design using the grouting controllability index.

[0055] It should be noted that the above steps can be applied to the selected typical borehole fractures, or to new borehole fractures. When applied to new borehole fractures, the geometric parameters, surface structure characteristics, and hydraulic characteristic test parameters of the new borehole fractures need to be obtained using the aforementioned steps, and the filling factor can be deduced in reverse. Then, the equivalent permeability of the other borehole fractures can be predicted using the trained permeability prediction model.

[0056] Therefore, in some embodiments, the method further includes: For other borehole fractures within the target engineering area, the geometric parameters, surface structure feature data, and hydraulic characteristic test parameters of the other borehole fractures are obtained using the above steps. The filling factor of the other borehole fractures is obtained by back-calculating using the hydraulic characteristic test parameters of the other borehole fractures and the corrected fracture flow model; The permeability prediction model obtained through training is used to process the geometric parameters, surface structure feature data, and filling factor of the other borehole fractures to obtain the equivalent permeability of the other borehole fractures.

[0057] Subsequently, HIS and GCI indices can be constructed for the corresponding borehole fractures (which may include typical borehole fractures and other borehole fractures), and the borehole fractures can be classified.

[0058] HIS and GCI are two-level index evaluation systems constructed in this disclosure to achieve precise integration of borehole information and grouting design: (1) First level: Fracture hydraulic structure index (HSI) HSI is used to comprehensively evaluate the seepage potential and hydraulic complexity of fractures. Its functional expression is: HSI=f(b,JRC,θ,F,P)=α1·b n +α2·JRC n +α3·θ n +α4·F n +α5·P n Where: b n To normalize the fracture aperture, the value is between 0 and 1, b n The geometric opening measured in step 1 is obtained through linear mapping; JRC n This is the normalized roughness coefficient, with a value between 0 and 1; JRC n The JRC value calculated in step 2 is obtained by linear mapping; θ n θ is the normalized orientation factor, representing the effect of the angle between the fracture direction and the grouting pressure gradient direction; n The cosine value of F is calculated based on the angle between the fracture direction and the grouting pressure gradient direction obtained in step 1, and then converted to obtain F. n The normalized filler factor takes values ​​between 0 and 1, where 0 represents complete filling and impermeability, and 1 represents no filling; F n The filler factor derived directly from the pressure test in step 4 is used; P n These are normalized pressure test parameters, which are normalized values ​​based on permeability; P n The permeability (Lu value) measured in step 4 is obtained by linear mapping; α1 to α5 are weighting coefficients, which are determined by training the artificial intelligence model in step 6, satisfying α1+α2+α3+α4+α5=1; the HSI value ranges from 0 to 1. The higher the HSI, the stronger the hydraulic conductivity of the fracture and the more significant the control effect on seepage.

[0059] It should be noted that the five normalization parameters (b) in the HSI index n JRC n θ n F n P nAll results are directly derived from the measurement or calculation results of steps 1, 2, and 4, and do not depend on the AI ​​model output of step 6. The equivalent hydraulic parameters (such as equivalent permeability) and HSI index output in step 6 are two different outputs of the same set of input parameters: the HSI index is used for rapid comprehensive evaluation and classification of fractured hydraulic structures, while the equivalent hydraulic parameters are used for precise quantitative calculation of seepage. The two are independent and complementary, and can be selected or used in combination according to engineering needs.

[0060] It should also be noted that, based on the AI ​​model trained in step 6 (or the 3000 sets of Monte Carlo simulation data generated in step 5), a global sensitivity analysis method (such as the Sobol index method or random forest feature importance assessment) is used to quantitatively calculate each input parameter (b). n JRC n θ n F n P n The contribution of the HSI index to the equivalent permeability is normalized to obtain the HSI weighting coefficients α1 to α5. In this embodiment, the weighting coefficients determined by sensitivity analysis are α1=0.25, α2=0.20, α3=0.15, α4=0.20, and α5=0.20, respectively. Once determined by sensitivity analysis, the HSI weighting coefficients α1 to α5 are fixed and will not change with different input parameters of the fracture to be predicted. This is because the weighting coefficients reflect the average relative importance of the influence of each parameter on hydraulic characteristics (for example, under the rock mass conditions of this engineering area, the influence of aperture change on permeability is generally greater than the influence of roughness change). This is an inherent law determined by the geological conditions of this area, rather than a dynamically adjusted parameter for a specific fracture. The HSI index changes with different fractures based on the five normalized parameters themselves (b n JRC n θ n F n P n The weighting coefficient is a fixed multiplier.

[0061] (2) Second level: Grouting controllability index (GCI) Based on HSI, GCI further incorporates grouting engineering factors to intuitively evaluate the groutability of fractures and the difficulty of grouting control. Its functional expression is as follows: GCI = HSI × β(M, ΔP, D) β(M,ΔP,D)=γ1·M n +γ2·ΔP n +γ3·D n Where β is the grouting engineering adjustment function; M nThe normalized grouting material characteristic parameters are a comprehensive index based on the ratio of grout viscosity, particle size, and crack aperture; ΔP n To normalize the grouting pressure difference, which is the difference between the design grouting pressure and the groundwater pressure; D n The normalized borehole intersection coefficient represents the influence of the intersection angle and position of the borehole and the fracture; γ1 to γ3 are engineering adjustment weights, which are determined based on specific engineering experience.

[0062] Based on the GCI value, the fractures are classified into three categories, which directly guides the grouting design: Easy-to-inject and controllable cracks (GCI≥0.6): The grouting material is easy to inject and its diffusion is controllable, using conventional grouting parameters; Injectable but controlled cracks (0.3≤GCI<0.6): Injectable but diffusion needs to be controlled, using intermittent grouting and step-by-step pressurization process; Difficult-to-inject and difficult-to-control cracks (GCI<0.3): Injection is difficult or diffusion is uncontrollable, requiring special measures such as high-pressure grouting and modified grout.

[0063] Through the above two-level index system, the complex fracture characterization results are transformed into intuitive engineering decision indicators, truly realizing the intelligent connection between borehole information and grouting design.

[0064] For example: Based on the multidimensional parameters of the fracture obtained from the inversion in step 6, construct an augmented representation of each fracture. Calculate the HSI value for each fracture, ranging from 0.3 to 0.8. Based on the training results of the artificial intelligence model in step 6, determine the HSI weight coefficients α1 to α5 to be 0.25, 0.20, 0.15, 0.20, and 0.20, respectively.

[0065] Further calculation of the controllability index (GCI) for fracture grouting. Based on the grouting design parameters for this project, the grouting material properties M... n =0.8 (cement-based grout, with a moderate ratio of particle size to crack aperture), design grouting pressure difference ΔP n =0.7 (grouting pressure 1.5MPa, groundwater pressure 0.5MPa), borehole delivery coefficient D n =0.9 (large angle borehole crossing of fracture). Based on engineering experience, the weights of the grouting engineering adjustment function β are determined as γ1=0.4, γ2=0.3, and γ3=0.3. β is calculated as 0.4×0.8+0.3×0.7+0.3×0.9=0.8, thus GCI=HSI×0.8, with a range of 0.24~0.64.

[0066] Cracks are classified based on their GCI (Gross Injection Coefficient) values: cracks with GCI ≥ 0.6 are easily groutable (HSI ≥ 0.75 in this project), and grouting parameters are optimized for these cracks, with grouting pressure set at 1.2–1.5 MPa and grouting rate at 5–10 L / min; cracks with GCI ≤ 0.3 < 0.6 are groutable (HSI 0.375–0.75), and the grouting pressure is adjusted to 0.9–1.2 MPa, using an intermittent grouting process; cracks with GCI < 0.3 are difficult to grout (HSI < 0.375), and high-pressure grouting (1.5–2.0 MPa) combined with modified grout is used. This method achieves precise integration of borehole information and grouting design, providing a scientific basis for deep grouting design of the dam foundation.

[0067] Reference Figure 4 The schematic diagram of the fracture hydraulic structure augmented model of this invention illustrates the process from the three-dimensional geometric model of the fracture interface constructed in step 2 to the streamline distribution, velocity field, and pressure field contour map inside the fracture obtained by numerical simulation in step 3. Using COMSOL Multiphysics numerical simulation results as an example, this diagram intuitively presents the progressive relationship between the fracture geometry and hydraulic properties, providing a visual reference for understanding the technical solution of this invention.

[0068] Compared with the prior art, this disclosure has the following advantages: (1) Breaking through the limitation that traditional borehole images can only acquire two-dimensional geometric information, the technology integrates multiple technologies to expand the information of fractures from "two-dimensional geometric information" to "multi-dimensional structure-hydraulic-grouting controllability", comprehensively capturing the geometric, structural, hydraulic characteristics and grouting controllability of fractures, filling the technical gap of traditional methods.

[0069] (2) The three-dimensional interface roughness parameter is introduced into the fracture hydraulic evaluation system. Combined with numerical simulation technology, the calculation accuracy of fracture hydraulic parameters is effectively improved, and the problem that traditional methods cannot quantitatively describe the characteristics of fracture interfaces is solved.

[0070] (3) The Monte Carlo stochastic simulation method is introduced to address the uncertainty of fracture parameters and quantify the impact of parameter fluctuations on seepage capacity, making the characterization results more scientific and practical, and in line with the actual needs of engineering.

[0071] (4) The use of artificial intelligence algorithms to achieve efficient fusion and parameter inversion of multi-source data simplifies the characterization process, improves the characterization efficiency, and enhances the intelligence level of fracture hydraulic structure identification.

[0072] (5) The constructed fracture hydraulic structure index (HSI) and two-level index evaluation system can intuitively evaluate the control effect of fractures on grout diffusion and transform the complex fracture characterization results into intuitive engineering decision indicators. (6) The core value lies in building an intelligent bridge between borehole information and grouting design, completely replacing the traditional crude process of "drilling to see cracks → experience judgment", realizing the intelligent closed loop of "drilling → three-dimensional structural hydraulic model → AI → grouting parameters", solving the pain points of the traditional method of disconnecting borehole information and grouting design, relying on experience and lacking accuracy. It is a key link in the intelligentization of deep grouting engineering, providing more reliable basis for crack structure identification and parameter estimation for deep grouting engineering, effectively optimizing grouting design, improving engineering construction quality and safety, and has broad engineering application value.

[0073] It should be noted that although the application scenario exemplified above is a dam foundation, the method disclosed herein can actually be applied to various different scenarios. Taking underground cavern seepage prevention grouting projects as an example, the drilling depth in this method can be 30-60m. A borehole acoustic imaging system combined with a borehole television system is used for image acquisition to improve the accuracy of fracture identification; portable 3D scanning equipment is used for 3D scanning, which is suitable for operation in the confined space of underground caverns; the artificial intelligence inversion uses a random forest model to improve the efficiency and accuracy of multi-source data fusion; a stress field factor is added to the HSI index calculation, and the characterization results are optimized in combination with the stress state of the underground cavern. At the same time, the weight coefficient of the grouting engineering adjustment function β is adjusted according to the stress state of the underground cavern, so that the grouting design is more in line with the actual underground cavern project, further strengthening the intelligent connection between borehole information and grouting design.

[0074] With the addition of the stress field factor, the relevant data and models in steps 1-7 above will undergo the following adaptive adjustments: In step 1, it is necessary to supplement the collection or obtain the stress field parameters of the borehole area through in-situ stress testing, including the direction of the maximum principal stress, the direction of the minimum principal stress, and the stress magnitude, as input data for subsequent analysis; In step 3, it is necessary to introduce the stress-aperture coupling relationship into the fracture seepage model, that is, to calculate the dynamic change of fracture aperture under a given stress field condition (for example, according to the Barton-Bandis model, an increase in normal stress leads to a decrease in aperture), thereby obtaining an equivalent permeability and pressure loss that better reflects the actual stress state; In step 5, it is necessary to adjust the stress field factor... Substances (such as the magnitude and orientation angle of normal stress) are included as new uncertain parameters in the generation range of random parameter combinations, and their probability distribution is set (such as normal distribution or interval uniform distribution). In step 6, a stress field factor (such as normalized principal stress value or stress anisotropy coefficient) needs to be added to the input feature vector, and the corresponding data dimension in the training samples is increased accordingly, so that the machine learning model can learn the influence law of stress field on hydraulic parameters. In step 7, the expression of fracture hydraulic structure index HSI needs to add a stress field factor S, that is, HSI=f(b,JRC,θ,F,P,S), and its weighted normalized form is extended to HSI=α1·b n +α2·JRCn +α3·θ n +α4·F n +α5·P n +α6·S n The sum of α1 to α6 is 1; meanwhile, the weighting coefficients γ1 to γ3 of the grouting engineering adjustment function β need to be adjusted according to the stress state of the underground cavern, for example, by appropriately increasing the grouting pressure difference ΔP in high-stress areas. n The weights are adjusted to reflect the inhibitory effect of stress on grout diffusion. Through these adjustments, the method of this disclosure can be adapted to fracture hydraulic characterization and grouting design under different stress environments.

[0075] The following are embodiments of the apparatus disclosed herein, which can be used to execute embodiments of the method disclosed herein.

[0076] According to an embodiment of the present disclosure, a borehole fracture hydraulic structure augmentation characterization device can be implemented as part or all of an electronic device through software, hardware, or a combination of both. The borehole fracture hydraulic structure augmentation characterization device includes: The first construction module is configured to construct a borehole fracture geometric database based on image data of borehole fractures within the target engineering area; the borehole fracture geometric database includes the geometric parameters of the borehole fractures; The second construction module is configured to construct a three-dimensional geometric model of the borehole fracture and surface structural feature data of the borehole fracture based on the surface point cloud data of the borehole fracture. The surface structural feature data includes the roughness parameters of the borehole fracture. The simulation module is configured to simulate the hydraulic characteristics of the borehole fracture based on the three-dimensional geometric model, surface structure feature data, and geometric parameters by constructing a fracture flow model of the borehole fracture; the hydraulic characteristics simulation parameters include the equivalent permeability of the borehole fracture. The first acquisition module is configured to conduct a water pressure test on the borehole in the target engineering area and acquire water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fracture; the hydraulic characteristic test parameters include the test permeability of the borehole fracture; The calibration module is configured to calibrate the fracture flow model by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. The reverse calculation module is configured to back-calculate the filling factor based on the hydraulic characteristic test parameters and the corrected fracture flow model. The generation module is configured to generate multiple sets of random combinations of fracture parameters based on the geometric parameters, surface structure feature data, and the filling factor using the Monte Carlo method. The second acquisition module is configured to use the corrected fracture flow model to obtain the equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters. The training module is configured to use the random combination of the multiple sets of fracture parameters and the corresponding equivalent permeability as input and labels for training the permeability prediction model. The permeability prediction model obtained by training is used for the dimensional characterization of the borehole fracture hydraulic structure in deep grouting design.

[0077] The borehole fracture hydraulic structure augmentation characterization device in this embodiment corresponds to the borehole fracture hydraulic structure augmentation characterization method described above. For specific details, please refer to the description of the borehole fracture hydraulic structure augmentation characterization method described above, which will not be repeated here.

[0078] Figure 5 This is a schematic diagram of an electronic device suitable for implementing a method for characterizing the hydraulic structure of a borehole fracture according to an embodiment of the present disclosure.

[0079] like Figure 5 As shown, the electronic device 500 includes a processing unit 501, which can be implemented as a CPU, GPU, FPGA, NPU, or other processing unit. The processing unit 501 can execute various processes according to any of the methods described above in this disclosure, based on a program stored in the read-only memory (ROM) 502 or a program loaded from the storage portion 508 into the random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device 500. The processing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0080] The following components are connected to I / O interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to I / O interface 505 as needed. A removable medium 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 510 as needed so that computer programs read from it can be installed into storage section 508 as needed.

[0081] In particular, according to embodiments of this disclosure, any of the methods described above in the embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program containing program code for performing any of the methods in the embodiments of this disclosure. In such an embodiment, the computer program can be downloaded and installed from a network via communication section 509, and / or installed from removable medium 511.

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0083] The units or modules described in the embodiments of this disclosure can be implemented in software or hardware. The described units or modules can also be located in a processor, and the names of these units or modules do not necessarily constitute a limitation on the unit or module itself.

[0084] In another aspect, this disclosure also provides a computer-readable storage medium, which may be a computer-readable storage medium included in the apparatus described in the above embodiments; or it may be a standalone computer-readable storage medium not assembled into a device. The computer-readable storage medium stores one or more programs that are used by one or more processors to perform the methods described in this disclosure.

[0085] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

Claims

1. A method for characterizing the dimensional augmentation of hydraulic structures in borehole fractures, characterized in that: include: A geometric database of borehole fractures was constructed based on image data of borehole fractures within the target engineering area; The borehole fracture geometry database includes the geometric parameters of the borehole fracture; Based on the surface point cloud data of the borehole fracture, a three-dimensional geometric model of the borehole fracture and surface structure feature data of the borehole fracture are constructed, the surface structure feature data including the roughness parameter of the borehole fracture. Based on the aforementioned three-dimensional geometric model, surface structure feature data, and geometric parameters, a fracture flow model of the borehole fracture is constructed to simulate the hydraulic characteristics of the borehole fracture, including the equivalent permeability of the borehole fracture. A water pressure test is conducted on the boreholes within the target engineering area to obtain water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fractures; the hydraulic characteristic test parameters include the test permeability of the borehole fractures; The fracture flow model is corrected by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. Based on the hydraulic characteristic test parameters and the corrected fracture flow model, the filling factor is calculated by reverse calculation. Based on the geometric parameters, surface structure feature data, and the filling factor, multiple sets of random combinations of fracture parameters are generated using the Monte Carlo method. The equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters is obtained using the corrected fracture flow model. The multiple sets of fracture parameters are randomly combined and their corresponding equivalent permeability is used as the input and label of the permeability prediction model for training. The permeability prediction model obtained by training is used for the dimensional characterization of the hydraulic structure of the borehole fracture in deep grouting design.

2. The method according to claim 1, characterized in that: The method further includes: The hydraulic structure index of the borehole fracture is calculated using the geometric parameters, roughness parameters, filler factor, and hydraulic pressure test parameters. The grouting controllability index of the borehole fracture is calculated using the hydraulic structure index and the grouting engineering adjustment function. The borehole fractures are classified into different types to guide grouting design using the grouting controllability index.

3. The method according to claim 1, characterized in that: The method further includes: For other borehole fractures within the target engineering area, the geometric parameters, surface structure feature data, and hydraulic characteristic test parameters of the other borehole fractures are obtained using the above steps. The filling factor of the other borehole fractures is obtained by back-calculating using the hydraulic characteristic test parameters of the other borehole fractures and the corrected fracture flow model; The permeability prediction model obtained through training is used to process the geometric parameters, surface structure feature data, and filling factor of the other borehole fractures to obtain the equivalent permeability of the other borehole fractures.

4. The method according to claim 1, characterized in that: Each of the multiple sets of random combinations of fracture parameters includes multiple sets of random combinations of fracture parameters generated using the Monte Carlo method based on the geometric parameters, surface structure feature data, and the filling factor, including: The aperture distribution of the borehole fracture is determined using the geometric parameters, and the aperture distribution includes the mean, standard deviation, and distribution type of the aperture. The roughness statistical distribution of the borehole fracture is calculated using the roughness parameters. The roughness statistical distribution includes the mean, fluctuation range, and distribution type of JRC. Determine the filler factor distribution, which includes the value range and distribution type of the filler factor; Using the Monte Carlo method, multiple sets of random combinations of fracture parameters are generated by randomly sampling from the aperture distribution, roughness statistical distribution and filling factor distribution. Each set of combinations includes an aperture value, a JRC value and a filling factor.

5. The method according to any one of claims 1 to 4, characterized in that: The geometric parameters include the spatial location, orientation, dip angle, and initial aperture of the fracture; and / or, The roughness parameters include the joint roughness coefficient JRC, fractal dimension, and surface power spectral density PSD; and / or, The hydraulic characteristic simulation parameters also include the local velocity distribution and pressure loss characteristics of the borehole fractures; and / or, The hydraulic characteristic test parameters also include the injection pressure and flow rate of the borehole fracture; and / or, The value range of the filling factor is 0 to 1, where 0 indicates that the borehole fracture is completely filled with dense, impermeable material, 1 indicates that the borehole fracture is an open fracture without filling, and other values ​​indicate that the borehole fracture is in a partially filled state.

6. The method according to any one of claims 1 to 4, characterized in that: The fracture flow model is corrected by comparing and analyzing the simulated hydraulic characteristics parameters and the experimental hydraulic characteristics parameters, including: Replace the opening in the fracture flow model with the equivalent hydraulic opening calculated based on the hydraulic characteristic test parameters; Adjust the inlet and outlet pressure boundary conditions of the fracture flow model; Adjust the roughness correction coefficient or local resistance coefficient to control the error between the simulated hydraulic characteristics parameters and the corresponding experimental hydraulic characteristics parameters within a set threshold range, thereby establishing the measured constraints of fracture hydraulic parameters composed of the injection pressure-flow correspondence, permeability range, and equivalent hydraulic opening.

7. A dimension-enhancing characterization device for borehole fracture hydraulic structures, characterized in that: include: The first construction module is configured to build a borehole fracture geometry database based on image data of borehole fractures within the target engineering area; The borehole fracture geometry database includes the geometric parameters of the borehole fracture; The second construction module is configured to construct a three-dimensional geometric model of the borehole fracture and surface structural feature data of the borehole fracture based on the surface point cloud data of the borehole fracture. The surface structural feature data includes the roughness parameters of the borehole fracture. The simulation module is configured to simulate the hydraulic characteristics of the borehole fracture based on the three-dimensional geometric model, surface structure feature data, and geometric parameters by constructing a fracture flow model of the borehole fracture; the hydraulic characteristics simulation parameters include the equivalent permeability of the borehole fracture. The first acquisition module is configured to conduct a water pressure test on the borehole in the target engineering area and acquire water pressure test data; the test data includes hydraulic characteristic test parameters of the borehole fracture; the hydraulic characteristic test parameters include the test permeability of the borehole fracture; The calibration module is configured to calibrate the fracture flow model by comparing and analyzing the hydraulic characteristic simulation parameters and the hydraulic characteristic test parameters. The reverse calculation module is configured to back-calculate the filling factor based on the hydraulic characteristic test parameters and the corrected fracture flow model. The generation module is configured to generate multiple sets of random combinations of fracture parameters based on the geometric parameters, surface structure feature data, and the filling factor using the Monte Carlo method. The second acquisition module is configured to use the corrected fracture flow model to obtain the equivalent permeability corresponding to the random combination of the multiple sets of fracture parameters. The training module is configured to use the random combination of the multiple sets of fracture parameters and the corresponding equivalent permeability as inputs and labels for training the permeability prediction model. The permeability prediction model obtained by training is used for the dimensional characterization of the borehole fracture hydraulic structure in deep grouting design.

8. An electronic device, characterized in that: It includes a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to implement the method of any one of claims 1 to 6.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that: When executed by a processor, the computer instructions implement the method of any one of claims 1 to 6.

10. A computer program product comprising computer instructions, characterized in that: When executed by a processor, the computer instructions implement the method of any one of claims 1 to 6.

Images