Method for constructing three-dimensional fracture network model of coal body based on CT (Computed Tomography) scanning

Abstract

The invention discloses a coal body three-dimensional fracture network model construction method based on CT scanning, and the method comprises the steps: obtaining the three-dimensional image data of a coal sample based on CT scanning, and extracting the geometric features of a coal sample fracture through an image processing algorithm; and then constructing a random network model of the fractures according to the extracted geometrical characteristics of the fractures, simulating distribution and connection of fracture networks by using the three-dimensional random network model, and screening and optimizing the constructed model until a three-dimensional fracture network model meeting requirements is formed as a final model. According to the method, the geometric distribution of the fractures can be accurately described, the heterogeneity of the coal body is simulated, and the shapes of the fractures are adjusted according to actual conditions; the method overcomes the limitation that a traditional model generally cannot describe geometrical characteristics of fractures in detail and cannot accurately reflect a fracture network, finally realizes accurate reflection of three-dimensional distribution of fractures in a coal body, and provides accurate data for subsequent prediction of coalbed methane extraction efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of coalbed methane mining, and in particular is a method for constructing a three-dimensional fracture network model of a coal body based on CT scanning. Background Art

[0002] As a byproduct of coal formation, coalbed methane has the potential to be an efficient and clean energy source, but it is a strong greenhouse gas and also poses a threat to the safe production of coal mines. In view of the common problems of strong adsorption and low permeability of coal seams in my country, coalbed methane extraction efficiency is one of the important means of coal mine safety, among which the coal body fracture model can be used to predict coalbed methane extraction efficiency.

[0003] Traditional coal fracture models generally assume that coal seams are homogeneous or equivalent homogeneous media, ignoring the spatial heterogeneity of fractures and pores in the coal. These simplified assumptions cannot reflect the actual complexity of the coal fracture network, especially in describing the coalbed methane mining process, and cannot accurately capture the geometric distribution of fractures and the heterogeneity of the coal. CT scanning can provide three-dimensional images of coal samples, revealing the fractures, pores and their distribution inside the coal body, but the fracture model reconstructed by CT scanning is heavily dependent on physical samples. The larger the scanning volume, the longer the scanning time, and high-resolution scanning requires more scanning layers and higher calculation accuracy. When performing large-scale scanning, the shape and size of the fractures may be simplified due to the lower resolution, and in some cases, small fractures may be ignored or not accurately identified, so it cannot fully reflect the fracture distribution of large-scale coal seams.

[0004] In view of the above problems, it is urgent to seek a method for constructing a fracture model that can accurately reflect the heterogeneity of coal body fractures and pore space, is not limited by the scale of coal seam samples, and can handle a large range of coal seams, so as to solve the problems of single fracture distribution in coal body model, homogeneity of pore space and high cost of reconstructing fracture model by CT scanning. Summary of the invention

[0005] In response to the problems existing in the above-mentioned prior art, the present invention provides a method for constructing a three-dimensional fracture network model of coal body based on CT scanning, which can accurately reflect the heterogeneity of coal body fractures and pore space, is not limited by the scale of coal seam samples, and can process fracture models of large-scale coal seams.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: a method for constructing a three-dimensional fracture network model of a coal body based on CT scanning, comprising the following steps:

[0007] Step 1, obtaining a CT scan image of a coal sample: selecting a coal sample of a certain volume, and performing CT scanning on the coal sample to obtain a two-dimensional CT image, and then preprocessing the image to obtain the required CT scan image of the coal sample;

[0008] Step 2: Extracting geometric features of coal sample cracks: extracting crack regions from the coal sample CT scan image obtained in step 1, and performing morphological analysis on the crack regions, and then calculating geometric feature parameters of the cracks in each crack region;

[0009] Step 3: Construct a three-dimensional fracture network model: Based on the volume of the coal sample in step 1 and combined with the geometric characteristic parameters of each fracture in step 2, the Monte-Carle algorithm is used to generate model parameters for each fracture; then, a three-dimensional fracture network model is constructed based on the generated model parameters, and after optimizing the model, the required three-dimensional fracture network model of the coal body is obtained.

[0010] Further, the step 1 is specifically as follows:

[0011] 1.1. Sampling the target coal body to obtain the required volume of coal sample, and physically treating the coal sample including drying, decontamination and calibration to ensure the representativeness of the coal sample and the consistency of experimental conditions;

[0012] 1.2. Perform CT scanning on the coal sample to obtain a two-dimensional CT image. The scanning data is reasonably adjusted according to the physical properties of the coal sample (such as density, porosity and crack size). The spatial resolution should meet the resolution requirements of the minimum crack geometric features and select a volume size of 10μm or less. The scanning voltage and current are adjusted according to the density of the coal sample. High-density coal samples use higher voltages (120-180kV) and low-density coal samples use lower voltages (80-100kV). The current range is controlled at 100-200μA to optimize image contrast. The scanning step size needs to balance time efficiency and data accuracy, and a small step size of 0.1° / step is selected to ensure crack continuity. The exposure time is optimized according to the image noise and clarity test results, and is usually selected between 50 and 200ms. Through pre-scanning and multiple sets of experimental parameter tests, it can be ensured that the final parameter settings can not only scan efficiently, but also obtain clear and accurate two-dimensional CT images, providing reliable data for subsequent crack feature extraction.

[0013] 1.3. The scanned two-dimensional CT images are preprocessed by denoising, image smoothing and enhancement to ensure that the data quality meets the requirements of subsequent analysis. If errors occur in the scanned images, the CT scan is performed again; the errors include excessive noise or serious artifacts that mask the crack characteristics; blurred images, inability to clearly display the details of the crack characteristics; abnormal grayscale distribution that fails to correctly reflect the contrast between the cracks and the matrix in the coal sample; data loss or missing, blank areas, data truncation or scanning faults, affecting the integrity of the crack structure; geometric distortion, distortion or offset of the crack morphology.

[0014] Further, the step 2 is specifically as follows:

[0015] 2.1. Image segmentation: Perform fracture image segmentation on the two-dimensional CT image processed in step 1, extract the fracture area in the coal sample, and process the image using Avizo 3D;

[0016] 2.2. Crack morphology analysis: Perform morphological analysis on the extracted crack area and calculate the geometric characteristic parameters of the crack, including the mean value of the crack inclination μ a , standard deviation of crack inclination σ a , mean crack tendency μ b , standard deviation of crack tendency σ b , mean crack radius μ L , standard deviation of crack radius σ L and the fracture volume density ρ.

[0017] Further, the step three is specifically as follows:

[0018] 3.1. Model parameter generation: According to step 2, the geometric characteristic parameters of the fractures are extracted. In the coal sample with the volume determined in step 1, the number of fractures is calculated by the formula N=ρV, where N is the number of fractures and V is the volume of the coal sample. The Monte-Carle algorithm is used to generate model parameters for each fracture, as follows:

[0019] Crack radius Li, i = 1, 2, 3, ..., N; according to the mean crack radius μ L and the standard deviation of crack radius σ L Generate and assign values, which conform to normal distribution;

[0020] Its probability density function is:

[0021] Crack inclination angle a i , i = 1, 2, 3, ..., N; according to the mean value of the crack inclination μ a and the standard deviation of fracture inclination σ a Generate and assign values, which conform to normal distribution;

[0022] Its probability density function is:

[0023] Crack tendency b i , i = 1, 2, 3, ..., N; according to the mean value of crack tendency μ b and the standard deviation of crack tendency σ b Generate and assign values, which conform to normal distribution;

[0024] Its probability density function is:

[0025] 3.2. Construction of three-dimensional fracture network model: According to the fracture parameters (fracture radius Li, fracture inclination ai, fracture tendency bi) extracted in step 3.1, the three-dimensional fracture network model is constructed by controlling the COMSOL software using Matlab scripts. Specifically, N points with uniform spatial distribution are randomly generated within the volume of the coal sample as the centers of the fracture disks, and these points are numbered in the order of 1 to N; then, the corresponding radius size is generated for each fracture disk according to the fracture radius in the fracture parameters, and the disks are rotated and adjusted in turn according to the inclination and tendency data in the fracture parameters to match their directional attributes; finally, all adjusted disks are integrated into the coal sample volume through geometric Boolean operations to form a complete three-dimensional fracture network model, which provides a basis for subsequent numerical calculations;

[0026] 3.3. Optimization of the three-dimensional fracture network model: After constructing the three-dimensional fracture network model, if some disk parts exceed the volume range of the coal sample, delete the exceeding parts; after completing the adjustment of the three-dimensional fracture network model, calculate the fracture geometric parameters of the three-dimensional fracture network model, including the mean fracture radius μS, the standard deviation of the fracture radius σS, the mean fracture inclination μα, the standard deviation of the fracture inclination σα, the mean fracture inclination μβ, the standard deviation of the fracture inclination σβ and the fracture density λ, and compare them with the geometric characteristic parameters of the fracture obtained in step 2.2 (mean fracture inclination μa, standard deviation of fracture inclination σa, mean fracture inclination μb, standard deviation of fracture inclination σb, mean fracture radius μL, standard deviation of fracture radius σL and fracture body density ρ) and perform error calculation; if the current three-dimensional fracture If the errors of the geometric parameters of the network model after comparison do not exceed 5%, the three-dimensional fracture network model is determined to be a three-dimensional fracture network model of the coal body, which is used to subsequently determine the distribution of coal body fractures; if the error of at least one geometric parameter in the current three-dimensional fracture network model after comparison exceeds 5%, the model parameters in step 3.1 and the fracture network construction data in step 3.2 are deleted, and the model parameter generation in step 3.1 and the fracture network construction in step 3.2 are repeated; during the construction process, if some disks exceed the volume range of the coal sample, the exceeding parts are still deleted; then, the geometric parameters of the three-dimensional fracture network model are recalculated, and the error calculation is performed again with the geometric characteristic parameters of the fractures obtained in step 2.2 until a three-dimensional fracture network model that meets the error requirements is generated.

[0027] Compared with the prior art, the present invention obtains three-dimensional image data of coal samples based on CT scanning, and extracts the geometric features of coal sample fractures through image processing algorithms; then, based on the extracted fracture geometric features, a random network model of fractures is constructed, and the three-dimensional random network model is used to simulate the distribution and connection of the fracture network, and the constructed model is screened and optimized until a three-dimensional fracture network model that meets the requirements is formed as the final model, and the model is used to analyze the permeability and storage data of coal samples. The method of the present invention can accurately describe the geometric distribution of fractures, simulate the heterogeneity of coal bodies, and adjust the shape of fractures according to actual conditions. It overcomes the limitations of traditional models that are usually unable to describe the geometric features of fractures in detail and cannot accurately reflect the fracture network, and finally achieves accurate reflection of the three-dimensional distribution of fractures inside the coal body, providing accurate data for subsequent prediction of coalbed methane extraction efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is the overall flow chart of the present invention;

[0029] Figure 2 A fracture generation diagram of the three-dimensional fracture network model of the coal body in the embodiment;

[0030] Figure 3 A perspective view of a three-dimensional fracture network model of a coal body in an embodiment;

[0031] Figure 4 This is a comparison diagram of the effect of the fracture tendency after CT scanning and the fracture tendency of the three-dimensional fracture network model of the coal body in the embodiment;

[0032] Figure 5 This is a comparison diagram of the effects of the fracture inclination after CT scanning and the fracture angle of the three-dimensional fracture network model of the coal body in the embodiment. DETAILED DESCRIPTION

[0033] The present invention will be further described below.

[0034] like Figure 1 As shown, the present invention comprises the following steps:

[0035] Step 1: Obtain CT scan image of coal sample: select a certain volume of coal sample, and perform CT scan on the coal sample to obtain a two-dimensional CT image, and then pre-process the image to obtain the required CT scan image of the coal sample, specifically:

[0036] 1.1. Sampling the target coal body to obtain the required volume of coal sample, and physically treating the coal sample including drying, decontamination and calibration to ensure the representativeness of the coal sample and the consistency of the experimental conditions. The process of this embodiment is as follows: sampling the coal seam area with representative geological conditions, using drilling equipment to obtain columnar coal samples (such as 50 mm in diameter and 100 mm in height), gently removing the dirt, dust and debris on the surface of the coal sample with a soft brush to avoid damaging the fracture structure, gently rinsing the surface of the coal sample with deionized water to remove residual impurities, placing the cleaned coal sample in a ventilated and cool place, air-drying it naturally for 48 hours, or drying it at low temperature in a constant temperature oven (temperature of 40 to 50°C for 24 hours), ensuring that the moisture content in the coal sample meets the experimental requirements, using a vernier caliper or a laser measuring instrument to accurately measure the diameter and height of the coal sample, and recording the geometric parameters of the coal sample.

[0037] 1.2. Perform CT scanning on the coal sample to obtain a two-dimensional CT image. The scanning data is reasonably adjusted according to the physical properties of the coal sample (such as density, porosity and crack size). The process of this embodiment is as follows: the dried, cleaned and calibrated coal sample is placed in the sample slot of the CT scanning device to ensure that the coal sample does not move during the scanning process and that the surface of the coal sample is flat to avoid image distortion or artifacts during the scanning process. The CT spatial resolution should meet the resolution requirements of the minimum crack geometric features and select a volume size of 10 μm or less. The scanning voltage and current are adjusted according to the density of the coal sample. High-density coal sample sampling A higher voltage (120-180 kV) is used, while a lower voltage (80-100 kV) is used for low-density coal samples; the current range is controlled at 100-200 μA to optimize image contrast; the scanning step size needs to balance time efficiency and data accuracy, and a small step size of 0.1° / step is selected to ensure crack continuity; the exposure time is optimized based on the image noise and clarity test results, and is usually selected between 50 and 200 ms; through pre-scanning and multiple sets of experimental parameter tests, it can be ensured that the final parameter setting can not only scan efficiently, but also obtain clear and accurate two-dimensional CT images, providing reliable data for subsequent crack feature extraction.

[0038] 1.3. The scanned two-dimensional CT images are preprocessed by denoising, image smoothing and enhancement to ensure that the data quality meets the requirements of subsequent analysis. If errors occur in the scanned images, the CT scan is performed again; the errors include excessive noise or serious artifacts that mask the crack characteristics; blurred images, inability to clearly display the details of the crack characteristics; abnormal grayscale distribution that fails to correctly reflect the contrast between the cracks and the matrix in the coal sample; data loss or missing, blank areas, data truncation or scanning faults, affecting the integrity of the crack structure; geometric distortion, distortion or offset of the crack morphology.

[0039] Step 2: Extraction of geometric features of coal sample cracks: Extract the crack area from the coal sample CT scan image obtained in step 1, perform morphological analysis on the crack area, and then calculate the geometric feature parameters of the cracks in each crack area, specifically:

[0040] 2.1. Image segmentation: Perform fracture image segmentation on the two-dimensional CT image processed in step 1, extract the fracture area in the coal sample, and use Avizo 3D to process the image; the process of this embodiment is: use the noise filtering function in Avizo 3D to denoise the scanned image to reduce the impact of noise on the image quality, adjust the brightness and contrast of the image according to the density and material properties of the coal sample to make the boundary between the fracture and the matrix clearer, select the "Thresholding" tool, and segment the fracture area by adjusting the image grayscale value, and use the Remove Small Objects function to remove these irrelevant small objects to improve the accuracy of fracture recognition.

[0041] 2.2. Crack morphology analysis: Perform morphological analysis on the extracted crack area and calculate the geometric characteristic parameters of the crack, including the mean value of the crack inclination μ a , standard deviation of crack inclination σ a , mean crack tendency μ b , standard deviation of crack tendency σ b , mean crack radius μ L , standard deviation of crack radius σ L and the fracture volume density ρ.

[0042] Step 3: Construction of three-dimensional fracture network model: Based on the volume of the coal sample in step 1 and the geometric characteristic parameters of each fracture in step 2, the Monte-Carle algorithm is used to generate model parameters for each fracture; then, a three-dimensional fracture network model is constructed based on the generated model parameters, and after optimizing the model, the required three-dimensional fracture network model of the coal body is obtained, specifically:

[0043] 3.1. Model parameter generation: According to step 2, the geometric characteristic parameters of the fractures are extracted. In the coal sample with the volume determined in step 1, the number of fractures is calculated by the formula N=ρV, where N is the number of fractures and V is the volume of the coal sample. The Monte-Carle algorithm is used to generate model parameters for each fracture, as follows:

[0044] Crack radius Li, i = 1, 2, 3, ..., N; according to the mean crack radius μ L and the standard deviation of crack radius σ L Generate and assign values, which conform to normal distribution;

[0045] Its probability density function is:

[0046] Crack inclination angle a i , i = 1, 2, 3, ..., N; according to the mean value of the crack inclination μ a and the standard deviation of fracture inclination σ a Generate and assign values, which conform to normal distribution;

[0047] Its probability density function is:

[0048] Crack tendency b i , i = 1, 2, 3, ..., N; according to the mean value of crack tendency μ b and the standard deviation of crack tendency σ b Generate and assign values, which conform to normal distribution;

[0049] Its probability density function is:

[0050] In this embodiment, first, according to the fracture statistical characteristics (mean and standard deviation of fracture inclination, dip, radius and fracture body density) provided by experimental data or literature, the coal sample volume V and the number of fractures N are calculated using Matlab software; then, the radius, dip and dip of each fracture are randomly generated from the normal distribution using the Monte-Carlo algorithm to ensure that they meet the set mean and standard deviation. The generated fracture parameters will be used to construct a three-dimensional fracture network model of the coal body. Through these steps, a fracture network that meets the actual coal body characteristics can be simulated to provide data support for engineering applications such as coalbed methane mining.

[0051] 3.2. Construction of three-dimensional fracture network model: According to the fracture parameters (fracture radius Li, fracture inclination ai, fracture tendency bi) extracted in step 3.1, the three-dimensional fracture network model is constructed by controlling the COMSOL software using Matlab scripts. Specifically, N points with uniform spatial distribution are randomly generated within the volume of the coal sample as the centers of the fracture disks, and these points are numbered in the order of 1 to N; then, the corresponding radius size is generated for each fracture disk according to the fracture radius in the fracture parameters, and the disks are rotated and adjusted in turn according to the inclination and tendency data in the fracture parameters to match their directional properties; finally, all adjusted disks are integrated into the coal sample volume through geometric Boolean operations to form a complete three-dimensional fracture network model, which provides a basis for subsequent numerical calculations.

[0052] 3.3. Optimization of the three-dimensional fracture network model: After constructing the three-dimensional fracture network model, if some disk parts exceed the volume range of the coal sample, delete the exceeding parts; after completing the adjustment of the three-dimensional fracture network model, calculate the fracture geometric parameters of the three-dimensional fracture network model, including the mean fracture radius μS, the standard deviation of the fracture radius σS, the mean fracture inclination μα, the standard deviation of the fracture inclination σα, the mean fracture inclination μβ, the standard deviation of the fracture inclination σβ and the fracture density λ, and compare them with the geometric characteristic parameters of the fracture obtained in step 2.2 (mean fracture inclination μa, standard deviation of fracture inclination σa, mean fracture inclination μb, standard deviation of fracture inclination σb, mean fracture radius μL, standard deviation of fracture radius σL and fracture body density ρ) and perform error calculation; if the current three-dimensional fracture If the errors of the geometric parameters of the network model after comparison do not exceed 5%, the three-dimensional fracture network model is determined to be a three-dimensional fracture network model of the coal body, which is used to subsequently determine the distribution of coal body fractures; if the error of at least one geometric parameter in the current three-dimensional fracture network model after comparison exceeds 5%, the model parameters in step 3.1 and the fracture network construction data in step 3.2 are deleted, and the model parameter generation in step 3.1 and the fracture network construction in step 3.2 are repeated; during the construction process, if some disks exceed the volume range of the coal sample, the exceeding parts are still deleted; then, the geometric parameters of the three-dimensional fracture network model are recalculated, and the error calculation is performed again with the geometric characteristic parameters of the fractures obtained in step 2.2 until a three-dimensional fracture network model that meets the error requirements is generated.

[0053] The specific process of this embodiment is as follows: after the fracture network is constructed, some fractures may exceed the volume range of the coal sample. This part of the fractures that exceed the volume does not conform to the actual situation of the coal seam, so it needs to be removed from the network. Through the MATLAB script and COMSOL interface, first check the position of each fracture disk to determine whether it exceeds the boundary of the coal sample. If it exceeds, use Boolean operations to delete or trim the exceeding part to ensure that the final fracture network only contains fractures within the range of the coal sample; if the generated fractures overlap, it may affect the permeability or other properties. In this case, Boolean operations are used to merge fractures. Through Boolean operations in COMSOL (such as union or subtraction operations), the overlapping disk parts are processed, the overlapping parts are merged or the redundant parts are deleted, so as to obtain a more reasonable fracture network. Next, it is necessary to calculate the relevant parameters of the optimized fracture network. These parameters include: mean fracture radius μ S and the radius standard deviation σ S , mean fracture inclination μ α and the standard deviation of the inclination angle σ α , mean crack tendency μ β and the tendency standard deviation σ β, fracture density λ. These parameters can be calculated using statistical functions in MATLAB. Using the geometric properties of the generated fractures (such as radius, dip, inclination, etc.), the statistical characteristics of the fracture network are calculated to ensure that the fracture network meets the preset model requirements. The calculated fracture network parameters are calculated for error, and the optimized fracture network is compared with the preset model parameters. If the error does not exceed 5%, it is considered that the fracture network meets the preset requirements, and the model can be enabled for subsequent calculation and analysis. If the error exceeds 5%, it is necessary to regenerate model parameters and reconstruct the fracture network. This process can be automated by writing scripts in MATLAB to ensure model accuracy. That is, if the error does not exceed 5%, the optimized fracture network model is applied to subsequent COMSOL simulation analysis to simulate permeability, stress or other physical properties; if the error exceeds 5%, it is necessary to regenerate the fracture parameters and reconstruct the three-dimensional fracture network model.

[0054] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for constructing a three-dimensional fracture network model of coal based on CT scanning, characterized in that: The following steps are involved: Step 1, obtaining a CT scan image of a coal sample: selecting a coal sample of a certain volume, and performing CT scanning on the coal sample to obtain a two-dimensional CT image, and then preprocessing the image to obtain the required CT scan image of the coal sample; Step 2: Extracting geometric features of coal sample cracks: extracting crack regions from the coal sample CT scan image obtained in step 1, and performing morphological analysis on the crack regions, and then calculating geometric feature parameters of the cracks in each crack region; Step 3: Construct a three-dimensional fracture network model: Based on the volume of the coal sample in step 1 and combined with the geometric characteristic parameters of each fracture in step 2, the Monte-Carle algorithm is used to generate model parameters for each fracture; then, a three-dimensional fracture network model is constructed based on the generated model parameters, and after optimizing the model, the required three-dimensional fracture network model of the coal body is obtained.

2. The method for constructing a three-dimensional fracture network model of coal based on CT scanning according to claim 1, characterized in that: The step 1 is specifically as follows: 1.

1. Sampling the target coal body to obtain the required volume of coal sample, and physically treating the coal sample including drying, decontamination and calibration to ensure the representativeness of the coal sample and the consistency of experimental conditions; 1.

2. Perform CT scanning on the coal sample to obtain a two-dimensional CT image. The scanning data is reasonably adjusted according to the physical properties of the coal sample. The spatial resolution should meet the resolution requirements of the minimum crack geometric features. The scanning voltage and current are adjusted according to the density of the coal sample. High-density coal samples use higher voltages, and low-density coal samples use lower voltages. The current range is controlled at 100-200μA to optimize the image contrast. The scanning step size needs to balance time efficiency and data accuracy to obtain a two-dimensional CT image. 1.

3. Perform pre-processing of denoising, image smoothing and enhancement on the scanned two-dimensional CT images to ensure that the data quality meets the requirements of subsequent analysis. If errors occur in the scanned images, perform the CT scan again.

3. The method for constructing a three-dimensional fracture network model of coal body based on CT scanning according to claim 1, characterized in that: The step 2 is specifically as follows: 2.

1. Image segmentation: Perform fracture image segmentation on the two-dimensional CT image processed in step 1, extract the fracture area in the coal sample, and process the image using Avizo 3D; 2.

2. Crack morphology analysis: Perform morphological analysis on the extracted crack area and calculate the geometric characteristic parameters of the crack, including the mean value of the crack inclination μ a , standard deviation of crack inclination σ a , mean crack tendency μ b , standard deviation of crack tendency σ b , mean crack radius μ L , standard deviation of crack radius σ L and the fracture volume density ρ.

4. The method for constructing a three-dimensional fracture network model of coal body based on CT scanning according to claim 3, characterized in that: The step three is specifically as follows: 3.

1. Model parameter generation: According to step 2, the geometric characteristic parameters of the fractures are extracted. In the coal sample with the volume determined in step 1, the number of fractures is calculated by the formula N=ρV, where N is the number of fractures and V is the volume of the coal sample. The Monte-Carle algorithm is used to generate model parameters for each fracture, as follows: Crack radius Li, i = 1, 2, 3, ..., N; according to the mean crack radius μ L and the standard deviation of crack radius σ L Generate and assign values, which conform to normal distribution; Its probability density function is: Crack inclination angle a i , i = 1, 2, 3, ..., N; according to the mean value of the crack inclination μ a and the standard deviation of fracture inclination σ a Generate and assign values, which conform to normal distribution; Its probability density function is: Crack tendency b i , i = 1, 2, 3, ..., N; according to the mean value of crack tendency μ b and the standard deviation of crack tendency σ b Generate and assign values, which conform to normal distribution; Its probability density function is: 3.

2. Construction of three-dimensional fracture network model: According to the fracture parameters extracted in step 3.1, the three-dimensional fracture network model is constructed by controlling the COMSOL software using Matlab scripts. Specifically, N points with uniform spatial distribution are randomly generated within the volume of the coal sample as the centers of the fracture disks, and these points are numbered in the order of 1 to N; then, the corresponding radius size is generated for each fracture disk according to the fracture radius in the fracture parameters, and the disks are rotated and adjusted in turn according to the inclination and dip data in the fracture parameters to match their directional attributes; finally, all adjusted disks are integrated into the coal sample volume through geometric Boolean operations to form a complete three-dimensional fracture network model, which provides a basis for subsequent numerical calculations; 3.

3. Optimization of the three-dimensional fracture network model: After constructing the three-dimensional fracture network model, if some disk parts exceed the volume range of the coal sample, delete the excess part; after completing the adjustment of the three-dimensional fracture network model, calculate the fracture geometric parameters of the three-dimensional fracture network model, including the mean fracture radius μS, the standard deviation of the fracture radius σS, the mean fracture inclination μα, the standard deviation of the fracture inclination σα, the mean fracture inclination μβ, the standard deviation of the fracture inclination σβ and the fracture density λ, and compare them one by one with the geometric characteristic parameters of the fractures obtained in step 2.2 and perform error calculation; if the errors of the various geometric parameters after comparison of the current three-dimensional fracture network model do not exceed 5%, the three-dimensional fracture network model is determined. The model is a three-dimensional fracture network model of the coal body, which is used to subsequently determine the distribution of fractures in the coal body; if the error of at least one geometric parameter in the current three-dimensional fracture network model after comparison exceeds 5%, the model parameters in step 3.1 and the fracture network construction data in step 3.2 are deleted, and the model parameter generation in step 3.1 and the fracture network construction in step 3.2 are re-performed; during the construction process, if some disks exceed the volume range of the coal sample, the excess part is still deleted; then, the geometric parameters of the three-dimensional fracture network model are recalculated, and the error calculation is performed again with the geometric characteristic parameters of the fractures obtained in step 2.2 until a three-dimensional fracture network model that meets the error requirements is generated.

Images