Coal mine roof rock stratum fracture inversion modeling method based on drill hole group

Abstract

The invention relates to a coal mine roof rock stratum fracture inversion modeling method based on a drill hole group, which comprises the following steps: collecting drill hole group rock core data of a coal mine roof sampling area, and calculating the fracture linear density of each drill hole; acquiring a conversion factor, and converting the fracture linear density value into a local fracture volume density value based on the conversion factor; performing spatial interpolation on the local fracture volume density value to generate a three-dimensional space global fracture volume density intensity field of the coal mine roof rock stratum; taking the three-dimensional space global fracture volume density intensity field as a three-dimensional space fracture density constraint, and generating a coal mine roof rock discrete fracture network model database; and screening the coal mine roof rock stratum discrete fracture network model meeting the preset precision from the coal mine roof rock stratum discrete fracture network model database as final output, and completing coal mine roof rock stratum inversion modeling. The method is high in modeling precision, water inrush early warning and support design can be directly guided, and the safety of the coal mine roof is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine safety engineering and geological modeling, and in particular to a method for inversion modeling of coal mine roof rock strata fractures based on a drilling group. Background Art

[0002] The fracture system in coal mine roof strata is a key geological factor controlling surrounding rock stability, water inrush risk, and gas migration. Its complexity is reflected in its multi-scale nature, strong spatial heterogeneity, and dynamic evolution. Existing fracture modeling techniques have significant limitations in data acquisition, model construction, and engineering application, as follows:

[0003] Data dependence limitations: It mainly relies on surface outcrops or local tunnel wall mapping, which makes it difficult to reflect the true distribution of deep rock fractures. Especially in mining areas with thick overburden, there is a significant difference between surface data and the actual underground fracture network.

[0004] Insufficient modeling accuracy: Existing discrete fracture network (DFN) models often assume that the fracture density is uniformly distributed, ignoring the spatial variability of fractures caused by geological structure, weathering, and mining, resulting in large deviations in fracture distribution predictions.

[0005] Insufficient utilization of drilling data: Engineering drill cores are obtained in large quantities during coal mine exploration. Although they contain linear fracture density (P10) information, traditional methods do not convert them into three-dimensional volume fracture density (P32) fields, resulting in the value of a large amount of core data not being fully tapped.

[0006] Single verification method: Model verification mostly relies on a small number of verification points and lacks systematic error analysis (such as mean absolute error Mae), making it difficult to ensure the global reliability of the model. Summary of the Invention

[0007] The purpose of the present invention is to provide a coal mine roof rock stratum fracture inversion modeling method based on a borehole group. By mining the fracture information carried by a large amount of drill core data, a DFN model constrained by the borehole group fracture information is established to achieve accurate prediction of the spatial distribution of roof fractures.

[0008] To achieve the above object, the present invention provides the following solutions:

[0009] A method for inverse modeling of fractures in coal mine roof strata based on a drill hole cluster, comprising:

[0010] Collect core data from a cluster of drill holes in the coal mine roof sampling area and calculate the linear density of fractures in each drill hole;

[0011] Obtaining a conversion factor, and converting the fracture linear density value into a local fracture volume density value based on the conversion factor;

[0012] Perform spatial interpolation on the local fracture volume density values to generate the three-dimensional global fracture volume density intensity field of the coal mine roof strata;

[0013] The three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database of coal mine roof strata.

[0014] A discrete fracture network model of the coal mine roof stratum that meets the preset accuracy is selected from the coal mine roof stratum discrete fracture network model database as the final output to complete the coal mine roof stratum fracture inversion modeling.

[0015] Preferably, collecting core data of a group of boreholes in the coal mine roof sampling area and calculating the linear density of fractures in each borehole comprises:

[0016] Arrange a plurality of boreholes in a grid pattern in the coal mine roof sampling area;

[0017] High-resolution scanning of the cores from each borehole was performed to distinguish between natural fractures and mechanical fractures based on color and morphological characteristics;

[0018] The number of natural fractures carried by the core was counted, and the linear density of fractures in each drilling section was calculated.

[0019] Preferably, counting the number of natural fractures carried by the core and calculating the fracture linear density of each drilling section includes:

[0020] Along the borehole axis, with the core box length as the reference unit, the natural fractures carried by the core are counted in sections;

[0021] A digital calibration tool was used to record the spatial position of each natural fracture in a three-dimensional coordinate system, and the fracture linear density of each borehole section was calculated.

[0022] Preferably, the conversion factor is obtained by simulating a virtual borehole using a finite volume discrete fracture network and establishing a linear conversion relationship between fracture linear density and fracture volume density, including:

[0023] Construct a cube model and set the initial fracture volume density value;

[0024] Randomly generating a discrete fracture network model with Bingham-distributed fracture orientations and power-law distribution within the cubic model;

[0025] Arranging virtual boreholes in the discrete fracture network model, and calculating average fracture linear density values of the virtual boreholes;

[0026] The relationship curve between the fracture volume density and the fracture linear density is drawn through linear regression analysis, and the slope of the relationship curve is determined, which is the conversion factor.

[0027] Preferably, the spatial interpolation of the local fracture volume density value adopts a geostatistical interpolation method, and the geostatistical interpolation method adopts ordinary kriging method, including:

[0028] Fitting the experimental variogram based on all local fracture volume density values;

[0029] The coal mine roof modeling area is divided into a hexahedral grid, and the local fracture volume density value is interpolated into a fracture volume density field in three-dimensional space based on the ordinary kriging method using the experimental variation function;

[0030] The invalid data are eliminated by combining the bedrock surface and the lower limit surface of the borehole to generate the three-dimensional global fracture volume density intensity field.

[0031] Preferably, the three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database of the coal mine roof strata, including:

[0032] Based on GPS-RTK mapping, tunnel wall mapping and borehole TV data, the fracture occurrence in the coal mine roof modeling area is obtained;

[0033] Based on the fracture occurrence, the fractures in the coal mine roof modeling area are divided into several groups using an adaptive probabilistic pattern recognition algorithm, and the distribution parameters of the fracture groups are optimized using the Kolmogorov-Smirnov test.

[0034] According to the fracture volume density field, fracture orientation and size parameters, several groups of coal mine roof stratum discrete fracture network model libraries with drill hole group constraints are generated.

[0035] Preferably, a discrete fracture network model database of coal mine roof strata is generated, and the Bingham bivariate distribution model of fracture orientation and the power law distribution model of fracture size are combined for data supplementation.

[0036] Preferably, a coal mine roof stratum discrete fracture network model that meets a preset accuracy is selected from a coal mine roof stratum discrete fracture network model database as the final output, including:

[0037] Several verification boreholes are selected in the coal mine roof modeling area, and the average absolute error between the fracture linear density values simulated by the verification boreholes in the discrete fracture network model of the coal mine roof rock strata and the actual fracture linear density values is calculated. The discrete fracture network model of the coal mine roof rock strata whose average absolute error meets the preset value is selected as the final output, that is, an acceptable discrete fracture network model of the coal mine roof rock strata.

[0038] The beneficial effects of the present invention are:

[0039] The present invention solves the problem of insufficient spatial heterogeneity representation in traditional fracture models by integrating borehole core data, geostatistical interpolation, and discrete fracture network (DFN) modeling technology. The present invention identifies natural fractures based on borehole core information and calculates the fracture linear density (P10); calibrates the conversion factor (CF) between P10 and fracture volume density (P32) through finite volume uncertainty DFN modeling; uses ordinary kriging to interpolate the local P32 of the borehole group to generate a three-dimensional global P32 intensity field for the coal mine roof; combines the fracture orientation of the Bingham bivariate distribution and the fracture size of the power law distribution as additional supplementary data, and uses the global P32 field as a constraint condition to establish a DFN model for the coal mine roof strata; and selects the optimal model by verifying the measured and simulated P10 errors of the boreholes. The present invention has high accuracy and can directly guide water inrush warning and support design, significantly improving the safety of coal mine roofs. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0041] Figure 1 This is a flow chart of a method for inversion modeling of coal mine roof rock fractures based on a drill hole group according to an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the distribution of drill hole groups on the top of a coal mine according to an embodiment of the present invention;

[0043] Figure 3 Schematic diagram of calculation of drill core fracture P10 according to an embodiment of the present invention;

[0044] Figure 4 Schematic diagram of converting drilling hole P10 into P32 according to an embodiment of the present invention;

[0045] Figure 5 Schematic diagram of matching of a drilling part P32 and a corresponding grid unit according to an embodiment of the present invention;

[0046] Figure 6 This is a three-dimensional full-area P32 distribution map of the coal seam roof according to an embodiment of the present invention;

[0047] Figure 7 Schematic diagram of the DFN model of the coal seam roof according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] This embodiment provides a method for inversion modeling of fractures in coal mine roof strata based on a drill hole cluster, including:

[0051] Collect core data from a cluster of drill holes in the coal mine roof sampling area and calculate the linear density of fractures in each drill hole;

[0052] Obtaining a conversion factor, and converting the fracture linear density value into a local fracture volume density value based on the conversion factor;

[0053] Perform spatial interpolation on the local fracture volume density values to generate the three-dimensional global fracture volume density intensity field of the coal mine roof strata;

[0054] The three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database of coal mine roof strata.

[0055] A discrete fracture network model of the coal mine roof stratum that meets the preset accuracy is selected from the coal mine roof stratum discrete fracture network model database as the final output to complete the coal mine roof stratum fracture inversion modeling.

[0056] Specifically, this embodiment solves the problem of insufficient representation of spatial heterogeneity by traditional fracture models by integrating borehole core data, geostatistical interpolation and discrete fracture network (DFN) modeling technology. The present invention identifies natural fractures based on borehole core information and calculates fracture linear density (P10); calibrates the conversion factor (CF) between P10 and fracture volume density (P32) through finite volume uncertainty DFN modeling; uses ordinary Kriging to interpolate the local P32 of the borehole group to generate the three-dimensional global P32 intensity field of the coal mine roof; combines the fracture orientation of the Bingham bivariate distribution and the fracture size of the power law distribution as additional supplementary data, and uses the global P32 field as a constraint condition to establish a DFN model of the coal mine roof rock layer; and screens the optimal model by verifying the measured and simulated P10 errors of the borehole. This embodiment has high accuracy and can directly guide water inrush warning and support design, significantly improving the safety of the coal mine roof.

[0057] Furthermore, core data of a group of boreholes in the coal mine roof sampling area is collected to calculate the linear density of fractures in each borehole, including:

[0058] Arrange a plurality of boreholes in a grid pattern in the coal mine roof sampling area;

[0059] High-resolution scanning of the cores from each borehole was performed to distinguish between natural fractures and mechanical fractures based on color and morphological characteristics;

[0060] The number of natural fractures carried by the core was counted, and the linear density of fractures in each drilling section was calculated.

[0061] Furthermore, the number of natural fractures carried by the core was counted, and the fracture linear density of each drilling section was calculated, including:

[0062] Along the borehole axis, with the core box length as the reference unit, the natural fractures carried by the core are counted in sections;

[0063] A digital calibration tool was used to record the spatial position of each natural fracture in a three-dimensional coordinate system, and the fracture linear density of each borehole section was calculated.

[0064] Furthermore, the conversion factor is obtained by simulating a virtual borehole using a finite volume discrete fracture network and establishing a linear conversion relationship between fracture linear density and fracture volume density, including:

[0065] Construct a cube model and set the initial fracture volume density value;

[0066] Randomly generating a discrete fracture network model with Bingham-distributed fracture orientations and power-law distribution within the cubic model;

[0067] Arranging virtual boreholes in the discrete fracture network model, and calculating average fracture linear density values of the virtual boreholes;

[0068] The relationship curve between the fracture volume density and the fracture linear density is drawn through linear regression analysis, and the slope of the relationship curve is determined, which is the conversion factor.

[0069] Furthermore, a geostatistical interpolation method is used to spatially interpolate the local fracture volume density value. The geostatistical interpolation method uses ordinary kriging, including:

[0070] Fitting the experimental variogram based on all local fracture volume density values;

[0071] The coal mine roof modeling area is divided into a hexahedral grid, and the local fracture volume density value is interpolated into a fracture volume density field in three-dimensional space based on the ordinary kriging method using the experimental variation function;

[0072] The invalid data are eliminated by combining the bedrock surface and the lower limit surface of the borehole to generate the three-dimensional global fracture volume density intensity field.

[0073] Furthermore, the three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database of the coal mine roof strata, including:

[0074] Based on GPS-RTK mapping, tunnel wall mapping and borehole TV data, the fracture occurrence in the coal mine roof modeling area is obtained;

[0075] Based on the fracture occurrence, the fractures in the coal mine roof modeling area are divided into several groups using an adaptive probabilistic pattern recognition algorithm, and the distribution parameters of the fracture groups are optimized using the Kolmogorov-Smirnov test.

[0076] According to the fracture volume density field, fracture orientation and size parameters, several groups of coal mine roof stratum discrete fracture network model libraries with drill hole group constraints are generated.

[0077] Furthermore, a discrete fracture network model database of coal mine roof strata was generated, and the Bingham bivariate distribution model of fracture orientation and the power-law distribution model of fracture size were combined to supplement the data.

[0078] Furthermore, a discrete fracture network model of the coal mine roof stratum that meets the preset accuracy is selected from the coal mine roof stratum discrete fracture network model database as the final output, including:

[0079] Several verification boreholes are selected in the coal mine roof modeling area, and the average absolute error between the fracture linear density values simulated by the verification boreholes in the discrete fracture network model of the coal mine roof rock strata and the actual fracture linear density values is calculated. The discrete fracture network model of the coal mine roof rock strata whose average absolute error meets the preset value is selected as the final output, that is, an acceptable discrete fracture network model of the coal mine roof rock strata.

[0080] The following combination Figure 1-Figure 7 The present embodiment provides a method for inversion modeling of coal mine roof rock fractures based on a drilling group, as shown in FIG. Figure 1 As shown, the specific steps include:

[0081] Step S1, collecting core data of the roof drilling group;

[0082] The distribution of the top drilling holes in the coal mine of this embodiment is shown in Figure 2The projection position (X, Y) of each borehole on the surface is accurately located through the plane coordinate system, and the Gauss-Krüger projection or the independent coordinate system of the mining area is used to achieve centimeter-level precision positioning, forming a two-dimensional benchmark for spatial analysis of core data; the borehole depth (Z) is based on the vertical position of the core sampling point recorded by the elevation measurement system, and the depth scale is constructed with the bedrock surface as the reference plane. The two together form a three-dimensional data framework of "plane coordinates-vertical depth".

[0083] Step S2, calculating the core fracture line density P10;

[0084] Based on high-resolution core scan images (≥600dpi), a morphological feature interpretation method was used to distinguish mechanical fracture surfaces from natural fracture surfaces using the dual indicators of "dark red iron-manganese impregnation traces" and "smooth surface associated with secondary mineral filling." Along the borehole axis, using the specific length of the core box as the reference unit, the effective fracture surfaces that passed the distinction were counted in sections. A digital calibration tool was used to record the spatial position (X, Y, Z) of each fracture surface in a three-dimensional coordinate system to ensure that the fracture count and spatial positioning were completed simultaneously. Figure 3 , according to the formula: Where N i is the number of segmented cracks, L i The length of the core segment is denoted by . For cores at the end of non-integer segments, linear interpolation is used for normalization to ensure comparability of data from different boreholes. The calculated results are integrated with the borehole spatial coordinates to form a structured data record format (X, Y, Z, P10), where X and Y are the plane coordinates of the borehole mouth; Z is the elevation of the center of the core segment (based on the bedrock surface); and P10 is accurate to two decimal places.

[0085] Step S3, P10 is converted into crack volume density P32;

[0086] See Figure 4 , a 20m×20m×20m cube benchmark model was established as the standard test space, and a discrete fracture network (DFN) model library containing different preset P32 values was randomly generated using the Monte Carlo method. Virtual borehole groups were arranged in a 2×2 matrix inside the model, and the borehole spacing was strictly maintained at 5m to simulate the actual sampling density of the project. Virtual drilling sampling was performed on each DFN model, and the P10 values of the four boreholes were counted. The P10-P32 relationship curve was drawn, and the slope of the curve was determined to be the conversion factor CF (i.e., the slope of the regression curve), and finally the conversion formula was established

[0087] Step S4, dividing the roof rock layer into grid units;

[0088] Based on the actual geological coordinates of the coal seam roof (using an independent coordinate system for the mining area, with an X / Y coordinate error of ≤0.1m and Z-axis elevation referenced to the 1985 National Elevation Datum), a three-dimensional geological framework model with meter-level accuracy was constructed by integrating drill hole cluster data (generally ≥100 engineering drill holes), seismic exploration profiles, and roadway exposure information. A gridding benchmark was then defined based on the standard length of the core box. Using a uniform meshing technique, the modeling area was divided into 250×200×20 hexahedral elements. A 10m×10m plane grid size corresponds to the engineering specification for drill hole spacing, and a 5m vertical grid thickness matches the sedimentary cycle characteristics of the roof strata.

[0089] Step S5, P32 point cloud data grid unit matching;

[0090] The spatial coordinates (X, Y, Z) of the engineering borehole group (≥200 holes) and the corresponding P32 values are imported into the geological model through 3D geological modeling software (such as GOCAD or Petrel), and the Kriging interpolation algorithm is used to achieve spatial matching between the borehole point cloud and the 10m×10m×5m grid unit. In the gridding process, based on the variogram parameter Where h is the lag distance, N(h) is the number of sample logarithms in the distance interval, and a spatial correlation model is established. The P32 value of each grid center point is weighted by searching a spherical neighborhood with a radius of 5 m to ensure that the grid cells within 1 m from the borehole directly inherit the measured value of the borehole, while the peripheral cells are allocated the interpolation results according to the exponential decay function. The final generated local P32 distribution map of the roof is shown in Figure 5 .

[0091] Step S6, interpolation of point P32 data;

[0092] The spatial variability characteristics of P32 were analyzed and the nugget value (C0) was calculated to characterize the random measurement error and microvariation, the sill value (C) reflects the intensity of regional total variation, and the range (a) defines the range of spatial autocorrelation. Then, the ordinary kriging interpolation method was used with the exponential model: A spatial weight matrix was constructed, and a 5m spherical search radius was set to perform interpolation calculations on 50m×50m grid cells. P32 values were assigned to blank grids through linear unbiased estimation, and the P32 distribution map of the entire coal seam floor was finally generated. Figure 6 .

[0093] Step S7, DFN modeling other data supplement;

[0094] The two basic parameter systems of fracture occurrence and fracture size are linked through bidirectional data flow. In terms of fracture occurrence, geometric parameters based on the Bingham bivariate distribution model need to be added:

[0095] The fracture occurrence is divided into four groups of dominant orientations through cluster analysis, and the probability density function is established. Where C is the normalization constant, k1 and k2 are distribution parameters, and the distribution fit was verified by KS test (P value > 0.05).

[0096] In terms of crack size, the power law distribution parameters need to be input: the minimum equivalent diameter d min =2m, maximum diameter d max =50m, distribution index b = 1.9343, its cumulative probability function is

[0097] Step S8, random DFN model database M i ;

[0098] The global P32 values generated by the boreholes were used as data constraints to generate the M-group roof DFN model according to the geological grid method.

[0099] Step S9: DFN model identification and verification.

[0100] For each DFN model, virtual drilling was performed at the verification borehole location (strictly matching the measured borehole spatial coordinates), and the simulated line density P10 value was statistically compared with the measured P10 value for verification. The model accuracy was evaluated by calculating the mean absolute error (Mae). If Mae ≤ 15.84%, it was determined to be an accurate coal seam floor DFN model; otherwise, the loop mechanism (i = i + 1) was triggered to regenerate. The output of the acceptable coal seam roof DFN model is shown in Figure 7 .

[0101] This method supports the real-time import of new borehole data and iterative model updates to adapt to the dynamic changes in coal mining. It also converts the P10 data of engineering drill cores into a three-dimensional P32 field for the first time, breaking the traditional modeling's reliance on surface data. It also constrains the spatial distribution of fracture density through ordinary Kriging interpolation, improving the accuracy of the uniform DFN model by 35%.

[0102] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A method for inversion modeling of coal mine roof rock fractures based on a drilling group, characterized in that: include: Collect core data from a cluster of drill holes in the coal mine roof sampling area and calculate the linear density of fractures in each drill hole; Obtaining a conversion factor, and converting the fracture linear density value into a local fracture volume density value based on the conversion factor; Perform spatial interpolation on the local fracture volume density values to generate the three-dimensional global fracture volume density intensity field of the coal mine roof strata; The three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database of coal mine roof strata. A discrete fracture network model of the coal mine roof stratum that meets the preset accuracy is selected from the coal mine roof stratum discrete fracture network model database as the final output to complete the coal mine roof stratum fracture inversion modeling.

2. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 1, characterized in that: Collecting core data of the drill hole group in the coal mine roof sampling area and calculating the linear density of fractures in each drill hole, including: Arrange a plurality of boreholes in a grid pattern in the coal mine roof sampling area; High-resolution scanning of the cores from each borehole was performed to distinguish between natural fractures and mechanical fractures based on color and morphological characteristics; The number of natural fractures carried by the core was counted, and the linear density of fractures in each drilling section was calculated.

3. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 2, characterized in that: Count the number of natural fractures carried by the core and calculate the linear density of fractures in each drilling section, including: Along the borehole axis, with the core box length as the reference unit, the natural fractures carried by the core are counted in sections; A digital calibration tool was used to record the spatial position of each natural fracture in a three-dimensional coordinate system, and the fracture linear density of each borehole section was calculated.

4. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 1, characterized in that: The conversion factor is obtained by simulating a virtual borehole using a finite volume discrete fracture network and establishing a linear conversion relationship between fracture linear density and fracture volume density, including: Construct a cube model and set the initial fracture volume density value; Randomly generating a discrete fracture network model with Bingham-distributed fracture orientations and power-law distribution within the cubic model; Arranging virtual boreholes in the discrete fracture network model, and calculating average fracture linear density values of the virtual boreholes; The relationship curve between the fracture volume density and the fracture linear density is drawn through linear regression analysis, and the slope of the relationship curve is determined, which is the conversion factor.

5. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 1, characterized in that: The local fracture volume density value is spatially interpolated using a geostatistical interpolation method, wherein the geostatistical interpolation method uses an ordinary kriging method, including: Fitting the experimental variogram based on all local fracture volume density values; The coal mine roof modeling area is divided into a hexahedral grid, and the local fracture volume density value is interpolated into a fracture volume density field in three-dimensional space based on the ordinary kriging method using the experimental variation function; The invalid data are eliminated by combining the bedrock surface and the lower limit surface of the borehole to generate the three-dimensional global fracture volume density intensity field.

6. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 1, characterized in that: The three-dimensional global fracture volume density intensity field is used as the three-dimensional fracture density constraint to generate a discrete fracture network model database for coal mine roof strata, including: Based on GPS-RTK mapping, tunnel wall mapping and borehole TV data, the fracture occurrence in the coal mine roof modeling area is obtained; Based on the fracture occurrence, the fractures in the coal mine roof modeling area are divided into several groups using an adaptive probabilistic pattern recognition algorithm, and the distribution parameters of the fracture groups are optimized using the Kolmogorov-Smirnov test. According to the fracture volume density field, fracture orientation and size parameters, several groups of coal mine roof stratum discrete fracture network model libraries with drill hole group constraints are generated.

7. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 6, characterized in that: A discrete fracture network model database of coal mine roof strata is generated, and the Bingham bivariate distribution model of fracture orientation and the power-law distribution model of fracture size are combined to supplement the data.

8. The method for inversion modeling of coal mine roof rock fractures based on a drilling group according to claim 1, characterized in that: The coal mine roof stratum discrete fracture network model that meets the preset accuracy is selected from the coal mine roof stratum discrete fracture network model database as the final output, including: Several verification boreholes are selected in the coal mine roof modeling area, and the average absolute error between the fracture linear density values simulated by the verification boreholes in the discrete fracture network model of the coal mine roof rock strata and the actual fracture linear density values is calculated. The discrete fracture network model of the coal mine roof rock strata whose average absolute error meets the preset value is selected as the final output, that is, an acceptable discrete fracture network model of the coal mine roof rock strata.

Images