Rock mass fracture parameter estimation method and estimation system

Abstract

The invention relates to the technical field of rock mass structure prediction methods, in particular to a rock mass fracture parameter estimation method and system. The method includes: acquiring a rock fracture image; preprocessing the rock mass fracture image to obtain a preprocessed image; the preprocessed image is analyzed and calculated to obtain a dominant fracture density range and a fracture density range in any direction, and Monte Carlo simulation is adopted to analyze the dominant fracture density range and the fracture density range in any direction so as to estimate the density of rock mass fractures; normal distribution and logarithmic normal distribution fitting are carried out on the preprocessed image, and the inclination angle of the rock mass fracture is estimated based on the projection angle; and counting the fracture length of each group of fractures in the preprocessed image, and estimating the length of the rock mass fractures in the group based on the fractures with the fracture length greater than a set threshold value. The rock mass fracture image is analyzed based on the geometrical characteristics and the distribution rule of the fractures, and the density, the inclination angle and the length of the rock mass fractures can be accurately estimated.

Description

Technical Field

[0001] The invention relates to the technical field of rock mass structure prediction methods, and in particular to a rock mass fracture parameter estimation method and estimation system. Background Art

[0002] In rock mass engineering, fracture density, fracture inclination, and trace length are important parameters for describing rock mass structural characteristics. They directly affect the mechanical properties, permeability, and stability of the rock mass.

[0003] The crack density reflects the degree of fracture of the rock mass and is an important indicator for measuring the integrity of the rock mass. The higher the crack density, the worse the integrity of the rock mass structure, its mechanical properties (such as strength, stiffness, etc.) will usually weaken, and the permeability of the rock mass will also increase. On the one hand, in the design and construction of engineering projects such as mining and tunnel construction, the higher the crack density, the easier it is for the rock mass to collapse or deform, so estimating the crack density helps to formulate reasonable support measures. On the other hand, in the assessment of the impact of crack density on the permeability of water or gas in the rock mass, estimating the crack density can be used to predict problems such as water flow channels and groundwater pollution.

[0004] The crack inclination is an important parameter that describes the geometric characteristics of the crack space. It determines the directionality of the crack and affects the potential instability modes of the rock mass structure surface, such as sliding and shear failure. The crack inclination not only determines the possibility of the sliding surface in the rock mass stability analysis of slope or tunnel engineering. In addition, in the study of fault activity, the crack inclination can be used to analyze the history of fault activity to predict the possibility of future geological disasters.

[0005] The length of the crack trace refers to the length of the crack that appears on the exposed surface, which reflects the scale and extension of the crack. Longer traces usually indicate that the crack is more permeable, which may have a more significant impact on the mechanical properties and permeability of the rock mass. The length of the crack trace is mainly used in the analysis of crack permeability and the identification of rock mass blocks. The length of the crack trace can be used to analyze the extension of the crack and estimate whether the crack runs through the rock mass, which is crucial for the prediction of the seepage path and the stability assessment of underground structures. By estimating the length of the trace, it is possible to determine whether the rock mass is easy to break into pieces and further assess its risk of instability under load.

[0006] These three parameters are generally measured directly by hand during the measurement process. Although the measurement process is intuitive and simple, the workload on site is large, and on-site technicians are required to use measuring instruments such as measuring tapes and curve plates for measurement, which consumes a lot of manpower and material resources. Moreover, there are difficult-to-reach measurement areas on the outcropping slopes with higher terrain or the surrounding rocks of underground tunnels, and rash measurements will also face safety risks. Therefore, in engineering design, construction, monitoring and other links, it is very important to accurately estimate these parameters. Different from traditional measurement methods, the more effective method now is non-contact measurement based on photography technology, image processing technology, and 3D laser scanners.

[0007] For example, the prior art entitled "A method and system for predicting ground pressure risk classification" provides a non-contact measurement estimation method, which obtains the crack image on the surface of the rock mass, extracts the crack contour in the crack image on the surface of the rock mass according to the hybrid algorithm of the improved K-means algorithm and the Canny edge detection operator, obtains the crack morphology information, ground stress information and geological structure information based on the crack feature points on the crack contour, inputs the crack morphology information, ground stress information and geological structure information into the trained prediction model, and outputs the ground pressure risk classification result. This method is to infer the crack morphology information, ground stress information and geological structure information through image processing. This inference method is simple and convenient, and can accurately infer the crack morphology information, ground stress information and geological structure information of the current rock mass, but this method involves very few levels, and only judges and analyzes the crack morphology information, ground stress information and geological structure information, and does not conduct specific analysis on other characteristic parameters of rock mass cracks. Summary of the invention

[0008] The purpose of this application is to solve the deficiencies of the above-mentioned background technology and to provide a rock mass fracture parameter estimation method and estimation system.

[0009] The technical solution of the present application is: a method for estimating rock mass fracture parameters, comprising:

[0010] Obtain rock mass fracture images;

[0011] Preprocessing the rock mass fracture image to obtain a preprocessed image;

[0012] The pre-processed images are analyzed and calculated to obtain the dominant fracture density range and the fracture density range in any direction. The dominant fracture density range and the fracture density range in any direction are analyzed by Monte Carlo simulation to estimate the density of rock mass fractures.

[0013] Normal distribution and log-normal distribution fitting are performed on the pre-processed images, and the inclination angle of the rock mass fracture is estimated based on the projection angle;

[0014] The length of each group of cracks in the preprocessed image is counted, and the length of the rock mass cracks in this group is estimated based on the cracks whose length is greater than a set threshold.

[0015] According to a rock mass fracture parameter estimation method provided by the present application, the method for preprocessing a rock mass fracture image to obtain a preprocessed image includes: using a Deeplabv3+ model to process the rock mass fracture image to obtain a rock mass fracture segmentation image; using an OpenCV library to grayscale the rock mass fracture segmentation image to obtain a grayscale image; using an Otsu method to binarize the grayscale image to obtain a binary image; using a Zhang-Suen algorithm to extract a fracture skeleton of the binary image; and grouping the fracture skeleton based on angle and node information while retaining the geometric morphology of the fracture structure to form a preprocessed image.

[0016] According to a method for estimating rock mass fracture parameters provided by the present application, the method for estimating the density of rock mass fractures is as follows: performing pixel-level feature extraction on a preprocessed image to obtain a dominant fracture density range and a fracture density range in any direction; based on the statistical distribution of the dominant fracture density range and the fracture density range in any direction, a Monte Carlo simulation method is used to obtain fracture density data through multiple random sampling and experiments, and the statistical characteristics of the fracture density data, including the expected value, standard deviation and its confidence interval, are calculated to estimate the rock mass fracture density.

[0017] According to a rock mass fracture parameter estimation method provided by the present application, the method for obtaining the dominant fracture density range of fractures includes: calculating the two-dimensional coordinate covariance matrix of each group of fractures; calculating the mean of each group of fractures; drawing the dominant survey line of the group of fractures with the mean as the starting point and the two-dimensional coordinate covariance matrix of each group of fractures as the direction, if the dominant survey line intersects with the i-th group of fractures, recording the coordinates of the intersection of the dominant survey line and the i-th group of fractures, calculating the dominant fracture density of the group of fractures based on the spacing between the intersections of the same group within the range of the dominant survey line and the sum of the spacing between the intersections of the same group; forming the dominant fracture density range of the rock mass based on the dominant fracture density of each group of fractures.

[0018] According to a rock mass fracture parameter estimation method provided by the present application, the method for obtaining the fracture density range in any direction includes: arranging set survey lines on a rock mass fracture segmentation image, arranging horizontal set survey lines at a set percentage of the image height; marking the intersection points of the set survey lines and the structural surfaces; counting the intersection points of the fractures and the survey lines on each group of set survey lines, recording the distance between the two intersection points, and calculating the arbitrary fracture density corresponding to the group of set survey lines based on the arbitrary intersection distance and arbitrary node distance in each group of set survey lines; and forming the arbitrary fracture density range of the rock mass based on the arbitrary fracture density of each group of set survey lines.

[0019] According to a rock fracture parameter estimation method provided by the present application, the method for estimating the inclination angle of rock fractures based on projection angle includes: traversing each group of fractures on a preprocessed image one by one based on the connectivity between pixels, and when traversing to an intersection area or an intersection point, judging the angle and connectivity to determine whether the fracture line segments to be continuously iterated can be approximately fitted to the same fracture, and if the judgment result is that they can be fitted to the same fracture, storing the relevant series of coordinate points in a dictionary; performing straight line fitting on the coordinate points stored in the dictionary by the least squares method, and calculating the projection angle of the fitting straight line; and obtaining the actual inclination angle of the rock fracture based on the relationship between the projection angle and the actual inclination angle.

[0020] According to a method for estimating rock fracture parameters provided in the present application, the method for estimating the length of the group of rock fractures includes: performing statistical analysis on each group of pixels on the preprocessed image whose length is greater than a set threshold, and calculating the average value of the group of pixels as the trace length of the fracture represented by the group of pixels.

[0021] The present application also provides a rock mass fracture parameter estimation system, comprising:

[0022] An image preprocessing module, which is used to segment, grayscale, binarize, and extract skeletons of the original image to obtain a preprocessed image;

[0023] A fracture density estimation module, wherein the fracture density estimation module is used to analyze and calculate the preprocessed image to obtain the dominant fracture density range and the fracture density range in any direction, and to analyze the dominant fracture density range and the fracture density range in any direction by using Monte Carlo simulation to estimate the density of rock mass fractures;

[0024] A fracture inclination estimation module, which is used to perform normal distribution and log-normal distribution fitting on the pre-processed image and estimate the inclination of the rock fracture based on the projection angle;

[0025] The crack length estimation module is used to count the crack length of each group of cracks in the preprocessed image, and estimate the length of the group of rock cracks based on the cracks whose length is greater than a set threshold.

[0026] According to a rock mass fracture parameter estimation system provided by the present application, the image preprocessing module includes:

[0027] An image segmentation processing module, wherein the image segmentation processing module processes the rock mass fracture image using a Deeplabv3+ model to obtain a rock mass fracture segmentation image;

[0028] An image grayscale processing module, wherein the image grayscale processing module grayscales the rock mass fracture segmentation image using the OpenCV library to obtain a grayscale image;

[0029] An image binarization processing module, wherein the image binarization processing module uses the Otsu method to binarize the grayscale image to obtain a binarized image;

[0030] An image skeleton extraction module, wherein the image skeleton extraction module uses a Zhang-Suen algorithm to extract a crack skeleton of a binary image;

[0031] The image crack skeleton grouping module groups the crack skeletons based on angle and node information while retaining the geometric form of the crack structure, so as to form a preprocessed image.

[0032] According to a rock mass fracture parameter estimation system provided by the present application, the fracture density estimation module includes:

[0033] A crack density range acquisition module, which is used to perform pixel-level feature extraction on the preprocessed image to obtain a dominant crack density range and a crack density range in any direction;

[0034] The fracture density statistics module is based on the statistical distribution of the dominant fracture density range and the fracture density range in any direction. It adopts the Monte Carlo simulation method, obtains fracture density data through multiple random sampling and experiments, and calculates the statistical characteristics of the fracture density data, including the expected value, standard deviation and its confidence interval, so as to estimate the rock fracture density.

[0035] According to a rock mass fracture parameter estimation system provided by the present application, the fracture density range acquisition module includes:

[0036] A dominant fracture density range acquisition module, the dominant fracture density range acquisition module is used to calculate the two-dimensional coordinate covariance matrix of each group of fractures, calculate the mean of each group of fractures, take the mean as the starting point, and draw the dominant survey line of the group of fractures with the two-dimensional coordinate covariance matrix of each group of fractures as the direction. If the dominant survey line intersects with the i-th group of fractures, record the coordinates of the intersection of the dominant survey line and the i-th group of fractures, calculate the dominant fracture density of the group of fractures based on the spacing of the same group of intersections within the range of the group of dominant survey lines and the sum of the spacing of the same group of intersections, and form the dominant fracture density range of the rock mass based on the dominant fracture density of each group of fractures;

[0037] An arbitrary fracture density range acquisition module is used to arrange set survey lines on a rock mass fracture segmentation image, arrange horizontal set survey lines at a set percentage of the image height, mark the intersections of the set survey lines and the structural surfaces, count the intersections of the fractures and the survey lines on each set survey line, record the distance between the two intersections, calculate the arbitrary fracture density corresponding to the set survey line based on the arbitrary intersection distance and arbitrary node distance in each set survey line area, and form an arbitrary fracture density range of the rock mass based on the arbitrary fracture density of each set survey line.

[0038] According to a rock mass fracture parameter estimation system provided by the present application, the fracture inclination estimation module includes:

[0039] An effective coordinate point acquisition module, which traverses each group of cracks on the preprocessed image one by one based on the connectivity between pixels, and when traversing to the intersection area or intersection point, judges the angle and connectivity to determine whether the crack line segments to be iterated can be approximately fitted to the same crack. If the judgment result is that they can be fitted to the same crack, the relevant series of coordinate points are stored in the dictionary;

[0040] The actual inclination acquisition module performs straight line fitting on the coordinate points stored in the dictionary by the least square method, calculates the projection angle of the fitting straight line, and obtains the actual inclination angle of the rock mass fracture based on the relationship between the projection angle and the actual inclination angle.

[0041] According to a rock fracture parameter estimation system provided by the present application, the fracture inclination estimation module is used to perform statistical analysis on each group of pixels on the preprocessed image whose length is greater than a set threshold, and take the average value of the group of pixels as the trace length of the fracture represented by the group of pixels.

[0042] The advantages of this application are as follows: 1. This application analyzes the characteristics of rock cracks. Based on the geometric characteristics of the cracks and their distribution laws, the parameters of the rock cracks can be obtained by image processing, and the density, inclination and length of the rock cracks can be accurately estimated, achieving the effect of efficiently, quickly and accurately estimating the geometric parameters of rock cracks. The geometric parameters of the shelter cracks involved are rich, the estimation method is simple, and the operation is convenient, which has great promotion value;

[0043] 2. The image preprocessing method of this application is very simple. By segmenting, graying, binarizing and extracting the skeleton of the original image, it is convenient to analyze the cracks on the image later, greatly reducing the amount of data processing and improving the estimation accuracy of rock mass geometric parameters;

[0044] 3. The method for estimating rock mass fracture density in this application is completely based on the geometric characteristics of rock mass fractures and their distribution law. The fracture tendency is mainly normally distributed or uniformly distributed. The azimuths of most fractures are concentrated in a specific direction, or in some cases, the direction is evenly distributed. This distribution law combined with the Monte Carlo simulation method can accurately estimate the rock mass fracture density. The idea is simple and reasonable, and the operation is convenient.

[0045] 4. The method for determining the dominant crack density range and the arbitrary crack density range in the present application is very simple. By performing targeted calculation and analysis on the characteristics and distribution rules of the cracks, the desired dominant crack density range and the arbitrary crack density range can be accurately obtained. The operation method is simple and the acquisition efficiency is extremely high.

[0046] 5. This application determines the inclination of the crack by fitting a straight line, and analyzes the preprocessed image based on the connectivity between pixels. The estimated result is accurate, and the model can be used for automatic analysis, which greatly improves the efficiency of estimation;

[0047] 6. The method for obtaining the length of rock mass cracks in this application is very simple. According to the characteristics and distribution law of the cracks, the trace length of the cracks can be accurately obtained by analyzing the crack pixels on the preprocessed image. The overall operation is simple and fast;

[0048] 7. The present application also relates to a rock mass fracture parameter estimation system, which can be integrated in a control system to automatically analyze and process rock mass images, quickly obtain geometric parameters of rock mass fractures, greatly improve the efficiency of rock mass fracture analysis, and accurately estimate the results.

[0049] This application analyzes rock fracture images based on the geometric characteristics of the fractures and their distribution patterns, and can accurately estimate the density, inclination and length of rock fractures. The overall estimation method is simple, the estimation results are accurate, and the efficiency of obtaining the geometric parameters of rock fractures is extremely high, which has great promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 : Schematic diagram of the rock mass fracture parameter estimation method flow chart of the present application. DETAILED DESCRIPTION

[0051] The embodiments of the present application are described in detail below, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0052] In the description of the present application, it should be understood that the terms "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0053] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0054] The present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0055] The present application relates to a method for estimating rock mass fracture parameters, which obtains the geometric parameters of rock mass fractures, including the density, inclination and length of rock mass fractures, by analyzing the image information of rock mass fractures. The density, inclination, structural plane spacing and trace length of fractures have regularity in rock mass engineering and can be described and estimated by different statistical distributions. Usually, the fracture density is mainly normal distribution or uniform distribution, indicating that the azimuth of most fractures is concentrated in a specific direction, or in some cases the direction is evenly distributed. The inclination of the fracture often obeys the normal distribution or lognormal distribution, which means that most inclinations are concentrated in a certain range, but there may be larger or smaller deviations. The structural plane spacing usually obeys the negative exponential distribution, indicating that small-spacing fractures are more common, and large-spacing fractures are less common. This memoryless distribution reflects the randomness of the fracture spacing. The trace length often obeys the lognormal distribution, reflecting the positive skewness of the fracture scale on the logarithmic scale. Most fractures are short, but there are also a few longer fractures. By fitting these distributions, the geometric characteristics of the fracture can be inferred, which helps in strength assessment and stability analysis in rock mass engineering.

[0056] Specifically, Figure 1 As shown, a rock mass fracture parameter estimation method of the present application includes the following steps:

[0057] S1, obtaining rock mass fracture images;

[0058] There are many methods to obtain the original image of the cracks on the rock surface, as long as the image information of the rock cracks can be accurately obtained;

[0059] S2, preprocessing the rock mass fracture image to obtain a preprocessed image;

[0060] The purpose of preprocessing the rock mass fracture image is to eliminate useless information in the original image information, remove most of the useless information and interference information, reduce the amount of data processing for subsequent analysis and calculation, and facilitate subsequent processing and analysis;

[0061] S3, analyzing the preprocessed image to obtain geometric parameters of rock mass fractures;

[0062] S31, analyzing and calculating the preprocessed image to obtain the dominant fracture density range and the fracture density range in any direction, and using Monte Carlo simulation to analyze the dominant fracture density range and the fracture density range in any direction to estimate the density of rock mass fractures;

[0063] S32, performing normal distribution and log-normal distribution fitting on the pre-processed image, and estimating the inclination angle of the rock mass fracture based on the projection angle;

[0064] S33, counting the length of each group of cracks in the preprocessed image, and estimating the length of the group of rock mass cracks based on the cracks whose lengths are greater than a set threshold.

[0065] In some embodiments of the present application, this embodiment optimizes the above-mentioned step S2, and the specific method for preprocessing the rock fracture image to obtain the preprocessed image is: using the Deeplabv3+ model to process the rock fracture image to obtain a rock fracture segmentation image; using the OpenCV library to grayscale the rock fracture segmentation image to obtain a grayscale image; using the Otsu method to binarize the grayscale image to obtain a binary image; using the Zhang-Suen algorithm to extract the fracture skeleton of the binary image; grouping the fracture skeleton based on the angle and node information while retaining the geometric morphology of the fracture structure to form a preprocessed image.

[0066] The DeepLabv3+ model is an extension of the Convolutional Neural Networks (CNN) in the field of semantic segmentation. It is specifically used to handle pixel-level classification tasks. It effectively captures multi-scale contextual information by introducing dilated convolutions and spatial pyramid pooling, and has good recognition capabilities for target details and edges. Compared with traditional full convolutional networks, the DeepLabv3+ model performs better in capturing target features at different scales. The DeepLabv3+ model processes the original image by performing image segmentation, segmenting the parts of the original image that involve important details and eliminating a large amount of useless information.

[0067] The OpenCV library is used to grayscale the segmented image and the Otsu method is used to binarize it. The purpose is to further highlight the rock cracks and make them easier to identify and obtain in the image.

[0068] The purpose of skeleton extraction by Zhang-Suen algorithm is to obtain the characteristic information of rock mass fracture skeleton in graphic information, so as to facilitate the subsequent further analysis and processing of rock mass fracture.

[0069] In a further embodiment of the present application, this embodiment optimizes the above-mentioned step S31. Specifically, to estimate the density of rock cracks, first, pixel-level feature extraction is performed on the preprocessed image to obtain the dominant crack density and crack density in any direction of the cracks; then, based on the statistical distribution of the crack density, the Monte Carlo simulation method is used to calculate the expected value of the crack density through multiple random sampling and experiments to ensure that the results are statistically significant and accurate.

[0070] For the estimation of the density range of dominant fractures, based on the existing fracture grouping, the two-dimensional coordinate covariance matrix of each group of fractures is first calculated. The corresponding covariance matrix formula is as follows:

[0071]

[0072] C i v i =λv i

[0073] i=1,2,…group_num

[0074] Where: C i ——The two-dimensional coordinate covariance matrix of the i-th group of cracks;

[0075] v i ——Eigenvector;

[0076] λ——eigenvalue;

[0077] From the existing grouping, extract each group, containing a list of multiple coordinate points; use the PCA module of sklearn.decomposition, fit the data with the PCA model, calculate the covariance matrix of the data, and extract the eigenvector v from it i and eigenvalue λ. The eigenvector v i indicates the main direction of the data distribution, and the eigenvalue λ represents the variance along these directions; where the eigenvector v i(eigenvectors): Each row represents the direction of a principal component (unit vector), which is the main direction of change of the data; eigenvalue λ (eigenvalues): corresponds to the variance of the eigenvector. The larger the value, the greater the degree of variation along that direction.

[0078] Further calculate the mean point μ of each group of cracks i , the expression is as follows:

[0079]

[0080] Where: μ i ——mean point of each group of cracks;

[0081] n – the number of cracks in each group;

[0082] x j ——the abscissa of the jth point in each group of cracks;

[0083] y j ——The vertical coordinate of the jth point in each group of cracks.

[0084] Then use Python's matplotlib library to plot i As the starting point, with v i Draw the dominant line for the direction. When the dominant line intersects with the i-th group of fractures, record the intersection coordinates, and calculate the intersection distance and L.

[0085] Calculate the average crack spacing d of the i-th group of cracks i and dominant crack density D i , and its calculation formula is:

[0086]

[0087] Where: d i ——the average crack spacing of the i-th group of cracks;

[0088] D i ——the dominant crack density of the i-th crack group;

[0089] L——the sum of the distances between the same group of intersections within the dominant survey line range;

[0090] N is the number of the i-th group of fractures within the range of the dominant survey line.

[0091] The dominant fracture density range D of the rock mass is formed according to the dominant fracture density of each group, which is expressed as:

[0092] D={DminD i ≤D≤maxD i}

[0093] The principle of obtaining the range of dominant fracture density in rock mass is: in the direction of all eigenvectors, the larger the eigenvalue, the larger the variance of sample projection in that direction, and the higher the degree of dispersion of fractures in that direction. Therefore, a set of ideal coordinate systems is constructed by calculating eigenvalues ​​and eigenvectors. Among them, one direction (i.e., the axis with a larger eigenvalue) allows fracture pixels to be projected on it as much as possible, representing the main axis direction of the fracture group; the other direction (the axis with a smaller eigenvalue) is orthogonal to the fracture group and serves as the optimal dominant survey line layout direction. Through this method, the geometric characteristics of the fracture group can be accurately calculated, and the dominant survey line layout in fracture density estimation can be effectively optimized.

[0094] For the estimation of the density of cracks in any direction, firstly, the survey lines are re-arranged on the segmentation results of the crack image; the horizontal survey lines are automatically arranged at 10%, 30%, 50% and 70% of the image height (not necessarily limited to the values ​​mentioned, and can be set accordingly according to actual needs) to ensure the uniformity and representativeness of the distribution of the set survey lines; the intersection points of the set survey lines and the structural surface (i.e., the rock mass structural surface) are marked to facilitate the subsequent spacing calculation; there is no need to consider the specific grouping of the cracks, only the spacing between the two points where the cracks on each survey line intersect is counted to calculate the crack density R in any direction. In this way, the crack density in any direction can be accurately estimated to ensure a comprehensive analysis of the crack distribution.

[0095] The corresponding formula is:

[0096]

[0097] Where: r i ——the average crack spacing of the i-th group of cracks;

[0098] D i ——the density of cracks in any direction of the i-th group of cracks;

[0099] l——the sum of the distances between any intersections within the range of the i-th set of survey lines;

[0100] N - This group sets the number of any cracks within the survey line.

[0101] According to the density of cracks in any direction of each group, the density range R of cracks in any direction of the rock mass is formed, which can be expressed as:

[0102] R = {RminR i ≤R≤maxR i}

[0103] The spatial distribution of cracks is defined. Since the spacing between structural planes usually obeys a negative exponential distribution, the dominant crack density range D and the crack density range R in any direction are set to obey a negative exponential distribution, thereby better reflecting the random distribution characteristics of cracks.

[0104] Next, we use the numpy.random.exponential function in Python to generate random samples and simulate the distribution of crack spacing. The spacing data generated by this function can better match the spatial distribution of actual rock mass cracks.

[0105] For each set of generated samples D and R, the weighted average of the comprehensive fracture linear density J is calculated using the following formula:

[0106] J=ω 1 D+ω 2 R

[0107] Where: J——comprehensive crack linear density;

[0108] D——dominant crack density range;

[0109] R——Crack density range in any direction;

[0110] ω 1 ——the first weight coefficient;

[0111] ω 2 ——the second weight coefficient;

[0112] Considering that the dominant fracture usually determines the dominant failure mode of the rock mass, the higher the fracture density, the greater the possibility of rock mass failure in this direction and the worse the stability. Therefore, the weight ω is set 1 is 0.6, and the density of cracks in any direction reflects the overall degree of fragmentation of the rock mass, affecting the uniformity and integrity of the rock mass. 2 is 0.4.

[0113] Repeat the above steps to perform Monte Carlo simulation. The core of Monte Carlo simulation is to estimate the expected value through a large number of random samples and generate a large amount of crack density data; finally, calculate the statistical characteristics of these data, including the expected value, standard deviation and its confidence interval, and estimate the crack density, that is, the rock crack density.

[0114] In a preferred embodiment of the present application, the present embodiment optimizes the above step S32, specifically, using normal distribution and lognormal distribution for fitting. Based on the connectivity between pixels, each group of cracks on the preprocessed image is traversed one by one, and when traversing to the intersection area or intersection point, the angle and connectivity are judged to determine whether the crack line segments to be continuously iterated can be approximately fitted to the same crack. If the judgment result is that they can be fitted to the same crack, the relevant series of coordinate points are stored in the dictionary.

[0115] Then, the polyfit function of the NumPy library is used to fit the stored coordinate points by the least squares method, and the projection angle of the fitted line is calculated. Because the projection angle obtained in this way is based on the two-dimensional image, it is difficult to accurately reflect the actual three-dimensional inclination. Therefore, it is necessary to calculate the actual inclination based on the relationship between the projection inclination and the actual inclination. The relationship between the projection angle and the actual inclination of the crack in the two-dimensional image can be constructed through a large number of actual statistical analyses.

[0116] The inclination data of rock mass fractures usually obeys normal distribution or lognormal distribution. Therefore, this embodiment performs fitting from the two-dimensional inclination calculated from the preprocessed image to obtain the expected value of each group of inclinations, thereby more accurately estimating the actual inclination, that is, the inclination of the rock mass fracture. This method not only improves the estimation accuracy of the inclination of the rock mass fracture, but also provides important data support for further rock mass mechanics analysis.

[0117] In other embodiments of the present application, this embodiment optimizes the above-mentioned step S33. Specifically, first, based on the preprocessing of the crack image, a statistical analysis is performed on each group of pixels whose length is greater than the set threshold value T, and the arithmetic mean is calculated, which is used as the trace length of the group of cracks. The engineering rock mass classification standard GB / T50218-2014 describes the statistics of joints as follows: scattered joints with a trace length greater than 1m should be counted, and joints that are already cemented by siliceous, ferrous, and calcareous materials should not be included in the statistics. Therefore, the threshold value T is set here as the number of pixels in the crack image when the actual length is one meter.

[0118] This method can effectively quantify the size of the cracks and extract the crack trace characteristics at the pixel level. It provides important geometric characteristic data and lays the foundation for subsequent rock mass analysis and engineering applications. The trace length obtained by this method will help to more accurately evaluate the structural characteristics and stability of the rock mass.

[0119] In addition, the present application also provides a rock mass fracture parameter estimation system, including an image preprocessing module, a fracture density estimation module, a fracture inclination estimation module and a fracture length estimation module, wherein the image preprocessing module is used to segment, grayscale, binarize and skeleton-extract the original image to obtain a preprocessed image; the fracture density estimation module is used to analyze and calculate the preprocessed image to obtain the dominant fracture density range and the fracture density range in any direction, and use Monte Carlo simulation to analyze the dominant fracture density range and the fracture density range in any direction to estimate the density of rock mass fractures; the fracture inclination estimation module is used to fit the preprocessed image with normal distribution and log-normal distribution, and estimate the inclination of rock mass fractures based on the projection angle; the fracture length estimation module is used to count the fracture length of each group of fractures in the preprocessed image, and estimate the length of the group of rock mass fractures based on the fractures whose length is greater than a set threshold.

[0120] Among them, the image preprocessing module includes an image segmentation processing module, an image grayscale processing module, an image binarization processing module, an image skeleton extraction module and an image crack skeleton grouping module. The image segmentation processing module uses the Deeplabv3+ model to process the rock crack image to obtain the rock crack segmentation image; the image grayscale processing module uses the OpenCV library to grayscale the rock crack segmentation image to obtain a grayscale image; the image binarization processing module uses the Otsu method to binarize the grayscale image to obtain a binary image; the image skeleton extraction module uses the Zhang-Suen algorithm to extract the crack skeleton of the binary image; the image crack skeleton grouping module groups the crack skeleton based on the angle and node information while retaining the geometric morphology of the crack structure, thereby forming a preprocessed image.

[0121] The fracture density estimation module includes a fracture density range acquisition module and a fracture density statistics module. The fracture density range acquisition module is used to perform pixel-level feature extraction on the preprocessed image to obtain the dominant fracture density range of the fracture and the fracture density range in any direction. The fracture density statistics module is based on the statistical distribution of the dominant fracture density range and the fracture density range in any direction, adopts the Monte Carlo simulation method, obtains fracture density data through multiple random sampling and experiments, and calculates the statistical characteristics of the fracture density data, including the expected value, standard deviation and its confidence interval, so as to estimate the rock fracture density.

[0122] The fracture density range acquisition module includes a dominant fracture density range acquisition module and an arbitrary fracture density range acquisition module. The dominant fracture density range acquisition module is used to calculate the two-dimensional coordinate covariance matrix of each group of fractures, calculate the mean of each group of fractures, take the mean as the starting point, and draw the dominant survey line of the group of fractures with the two-dimensional coordinate covariance matrix of each group of fractures as the direction. If the dominant survey line intersects with the i-th group of fractures, record the coordinates of the intersection of the dominant survey line and the i-th group of fractures, calculate the dominant fracture density of the group of fractures based on the spacing of the same group of intersections within the range of the dominant survey line and the sum of the spacing of the same group of intersections, and calculate the dominant fracture density of the group of fractures based on each group of fractures. The dominant fracture density of each group of fractures forms the dominant fracture density range of the rock mass; the arbitrary fracture density range acquisition module is used to arrange the set survey lines on the rock mass fracture segmentation image, arrange the horizontal set survey lines at the set percentage of the image height, mark the intersection points of the set survey lines and the structural surfaces, count the intersection points of the fractures and the survey lines on each group of set survey lines, record the distance between the two intersection points, calculate the arbitrary fracture density corresponding to the group of set survey lines based on the arbitrary intersection distance and the arbitrary node distance in each group of set survey lines, and form the arbitrary fracture density range of the rock mass based on the arbitrary fracture density of each group of set survey lines.

[0123] The fracture inclination estimation module includes an effective coordinate point acquisition module and an actual inclination acquisition module. The effective coordinate point acquisition module traverses each group of fractures on the preprocessed image one by one based on the connectivity between pixels. When traversing to the intersection area or intersection point, the angle and connectivity are judged to determine whether the fracture line segments to be continuously iterated can be approximately fitted to the same fracture. If the judgment result is that they can be fitted to the same fracture, the relevant series of coordinate points are stored in a dictionary; the actual inclination acquisition module performs straight line fitting on the coordinate points stored in the dictionary through the least squares method, calculates the projection angle of the fitting straight line, and obtains the actual inclination of the rock fracture based on the relationship between the projection angle and the actual inclination.

[0124] The crack inclination estimation module is used to perform statistical analysis on each group of pixels whose length is greater than a set threshold on the preprocessed image, and take the average value of the group of pixels as the trace length of the crack represented by the group of pixels.

[0125] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited by the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present application. The present application may have various changes and improvements without departing from the spirit and scope of the present application, and these changes and improvements fall within the scope of the present application to be protected. The scope of protection claimed in the present application is defined by the attached claims and their equivalents.

Claims

1. A method for estimating rock mass fracture parameters, characterized in that: include, Obtain rock mass fracture images; Preprocessing the rock mass fracture image to obtain a preprocessed image; The pre-processed images are analyzed and calculated to obtain the dominant fracture density range and the fracture density range in any direction. The dominant fracture density range and the fracture density range in any direction are analyzed by Monte Carlo simulation to estimate the density of rock mass fractures. Normal distribution and log-normal distribution fitting are performed on the pre-processed images, and the inclination angle of the rock mass fracture is estimated based on the projection angle; The length of each group of cracks in the preprocessed image is counted, and the length of the rock mass cracks in this group is estimated based on the cracks whose length is greater than a set threshold.

2. A method for estimating rock mass fracture parameters according to claim 1, characterized in that: The method for preprocessing a rock mass fracture image to obtain a preprocessed image comprises: processing the rock mass fracture image using a Deeplabv3+ model to obtain a rock mass fracture segmentation image; graying the rock mass fracture segmentation image using an OpenCV library to obtain a grayscale image; binarizing the grayscale image using an Otsu method to obtain a binary image; extracting a fracture skeleton of the binary image using a Zhang-Suen algorithm; and grouping the fracture skeleton based on angle and node information while retaining the geometric morphology of the fracture structure to form a preprocessed image.

3. A method for estimating rock mass fracture parameters according to claim 1, characterized in that: The method for estimating the density of rock mass cracks is as follows: performing pixel-level feature extraction on a preprocessed image to obtain a dominant crack density range and a crack density range in any direction; based on the statistical distribution of the dominant crack density range and the crack density range in any direction, a Monte Carlo simulation method is used to obtain crack density data through multiple random sampling and experiments, and the statistical characteristics of the crack density data, including the expected value, standard deviation and its confidence interval, are calculated to estimate the rock mass crack density.

4. A method for estimating rock mass fracture parameters according to claim 3, characterized in that: The method for obtaining the dominant fracture density range of fractures comprises: calculating the two-dimensional coordinate covariance matrix of each group of fractures; calculating the mean of each group of fractures; drawing the dominant survey line of the group of fractures with the mean as the starting point and the two-dimensional coordinate covariance matrix of each group of fractures as the direction, if the dominant survey line intersects with the i-th group of fractures, recording the coordinates of the intersection of the dominant survey line and the i-th group of fractures, calculating the dominant fracture density of the group of fractures based on the spacing of the same group of intersections within the range of the group of dominant survey lines and the sum of the spacing of the same group of intersections; and forming the dominant fracture density range of the rock mass based on the dominant fracture density of each group of fractures.

5. A method for estimating rock mass fracture parameters according to claim 3, characterized in that: The method for obtaining the range of fracture density in any direction comprises: arranging set survey lines on a rock mass fracture segmentation image, arranging horizontal set survey lines at a set percentage of the image height; marking intersections of the set survey lines and the structural surface; counting intersections of fractures and survey lines on each set survey line, recording the distance between two intersections, and calculating the arbitrary fracture density corresponding to the set survey line based on the arbitrary intersection distance and arbitrary node distance in each set survey line area; and forming the arbitrary fracture density range of the rock mass based on the arbitrary fracture density of each set survey line.

6. A method for estimating rock mass fracture parameters according to claim 1, characterized in that: The method for estimating the inclination angle of rock fractures based on projection angles comprises: traversing each group of fractures on a preprocessed image one by one based on the connectivity between pixels, and when traversing to an intersection area or an intersection point, judging the angle and connectivity to determine whether the fracture line segments to be continuously iterated can be approximately fitted to the same fracture, and if the judgment result is that they can be fitted to the same fracture, storing the relevant series of coordinate points in a dictionary; performing straight line fitting on the coordinate points stored in the dictionary by the least squares method to calculate the projection angle of the fitting straight line; and obtaining the actual inclination angle of the rock fracture based on the relationship between the projection angle and the actual inclination angle.

7. A method for estimating rock mass fracture parameters according to claim 1, characterized in that: The method for estimating the length of the group of rock mass cracks includes: performing statistical analysis on each group of pixels whose length is greater than a set threshold on the preprocessed image, and calculating the average value of the group of pixels as the trace length of the crack represented by the group of pixels.

8. A rock mass fracture parameter estimation system, characterized in that: The estimation system is operated according to any one of the rock mass fracture parameter estimation methods according to claims 1 to 7, including: An image preprocessing module, which is used to segment, grayscale, binarize, and extract skeletons of the original image to obtain a preprocessed image; A fracture density estimation module, wherein the fracture density estimation module is used to analyze and calculate the preprocessed image to obtain the dominant fracture density range and the fracture density range in any direction, and to analyze the dominant fracture density range and the fracture density range in any direction by using Monte Carlo simulation to estimate the density of rock mass fractures; A fracture inclination estimation module, which is used to perform normal distribution and log-normal distribution fitting on the pre-processed image and estimate the inclination of the rock fracture based on the projection angle; The crack length estimation module is used to count the crack length of each group of cracks in the preprocessed image, and estimate the length of the group of rock cracks based on the cracks whose length is greater than a set threshold.

9. A rock mass fracture parameter estimation system according to claim 8, characterized in that: The image preprocessing module includes: An image segmentation processing module, wherein the image segmentation processing module processes the rock mass fracture image using a Deeplabv3+ model to obtain a rock mass fracture segmentation image; An image grayscale processing module, wherein the image grayscale processing module grayscales the rock mass fracture segmentation image using the OpenCV library to obtain a grayscale image; An image binarization processing module, wherein the image binarization processing module uses the Otsu method to binarize the grayscale image to obtain a binarized image; An image skeleton extraction module, wherein the image skeleton extraction module uses a Zhang-Suen algorithm to extract a crack skeleton of a binary image; The image crack skeleton grouping module groups the crack skeletons based on angle and node information while retaining the geometric form of the crack structure, so as to form a preprocessed image.

10. A rock mass fracture parameter estimation system according to claim 8, characterized in that: The crack density estimation module includes: A crack density range acquisition module, which is used to perform pixel-level feature extraction on the preprocessed image to obtain a dominant crack density range and a crack density range in any direction; A fracture density statistics module, which is based on the statistical distribution of the dominant fracture density range and the fracture density range in any direction, adopts a Monte Carlo simulation method, obtains fracture density data through multiple random sampling and experiments, and calculates the statistical characteristics of the fracture density data, including the expected value, standard deviation and its confidence interval, so as to estimate the rock mass fracture density; The crack density range acquisition module includes: A dominant fracture density range acquisition module, the dominant fracture density range acquisition module is used to calculate the two-dimensional coordinate covariance matrix of each group of fractures, calculate the mean of each group of fractures, take the mean as the starting point, and draw the dominant survey line of the group of fractures with the two-dimensional coordinate covariance matrix of each group of fractures as the direction. If the dominant survey line intersects with the i-th group of fractures, record the coordinates of the intersection of the dominant survey line and the i-th group of fractures, calculate the dominant fracture density of the group of fractures based on the spacing of the same group of intersections within the range of the group of dominant survey lines and the sum of the spacing of the same group of intersections, and form the dominant fracture density range of the rock mass based on the dominant fracture density of each group of fractures; An arbitrary fracture density range acquisition module is used to arrange set survey lines on a rock mass fracture segmentation image, arrange horizontal set survey lines at a set percentage of the image height, mark the intersections of the set survey lines and the structural surfaces, count the intersections of the fractures and the survey lines on each set survey line, record the distance between the two intersections, calculate the arbitrary fracture density corresponding to the set survey line based on the arbitrary intersection distance and arbitrary node distance in each set survey line area, and form an arbitrary fracture density range of the rock mass based on the arbitrary fracture density of each set survey line.

Images