A rock mass random fracture network generation method, system, device and medium

Abstract

The application discloses a rock mass random fissure network generation method, system, device and medium, relates to the technical field of rock mass random fissure network generation, and comprises the following steps: obtaining measured data of a rock mass in a to-be-measured area; selecting a fissure network geometric parameter probability distribution model according to the measured data; determining the number of fissure groups according to the occurrence of fissure development; inputting the mean value and standard deviation of the occurrence, trace length and gap width of each fissure group into the distribution model, obtaining pseudo-random numbers of standard uniform distribution of each group, and generating an initial random fissure network through specific transformation of the pseudo-random numbers; calculating node information of the random fissure network by adopting a plane scanning algorithm, and judging all intersection points of the fissure network; searching for isolated single fissures and fissures with only one intersection point thereon by adopting a block unit search method, searching for isolated closed loops, and generating a random fissure network without isolated single fissures and closed loops.

Description

Technical Field

[0001] This invention relates to the field of random fracture network generation technology in rock mass, specifically to a method, system, equipment, and medium for generating random fracture networks in rock mass. Background Technology

[0002] In their natural state, fractured rock masses often exhibit fracture networks of varying scales, unlike intact rocks. The stability of rock slopes and surrounding rock of underground caverns is frequently influenced by these fracture networks. During slope or underground cavern excavation, the independent rock blocks cut out by the dominant connected fracture network often present significant challenges to stability analysis and design due to their random locations. It is often necessary to clarify the fracture distribution and predict the displacement of random rock blocks within the dominant connected network during the engineering survey and design phases. Therefore, the generation of random fracture networks in rock masses based on survey results and the search for their dominant connected surfaces have considerable practical value.

[0003] Numerical simulation of fracture networks is based on field surveys and probabilistic distribution models, and can also fully consider the numerous microfractures distributed within the rock mass. It has high theoretical and practical value, thus attracting widespread attention from researchers, leading to the development of various simulation algorithms. Currently, numerical simulation of fracture networks mainly employs the Monte Carlo method, graph theory methods, and image analysis methods. The Monte Carlo method can reflect the distribution patterns of different geometric parameters of fractures, and its program is easy to implement, thus gaining considerable attention and becoming a widely used algorithm. Existing techniques rarely analyze the acquisition of parameters for stochastic fracture networks, while the nodes, length, and connectivity of the fracture network have a significant impact on the calculation of rock mass stress and seepage. Some studies use mathematical matrix methods, establishing connection matrices and loop matrices respectively. The former describes the connection relationship between fracture nodes and fracture line elements, as well as the fracture direction, while the latter describes the basic loops in the fracture network, with each basic loop corresponding to a polygonal rock block. However, this method requires manually setting the directions of each characteristic during the calculation process, and it leads to repeated calculations of nodes or fracture line elements, increasing both the complexity of the numerical calculation and the computational cost.

[0004] In summary, existing technologies still lack a simple method for searching the dominant connected network of fracture networks obtained from numerical simulations. It is difficult to calculate the node (intersection) information on each fracture, and it is difficult to obtain parameters such as the number and location of intersections on each fracture in a random fracture network. Moreover, random fracture networks often contain special fractures such as isolated fractures, which do not constitute the dominant structural surface for rock mass stability analysis, increasing the complexity of numerical calculation of the model. Summary of the Invention

[0005] To address the problem that existing technologies struggle to obtain information on the intersection points of various fractures and thus fail to identify specific fractures, this invention proposes a method, system, device, and medium for generating a random fracture network in rock mass. By generating a random fracture network in rock mass and a search method for the master connected network, the problems existing in the prior art are solved.

[0006] A method for generating a random fracture network in rock mass includes the following steps:

[0007] Obtain measured data of the rock mass in the area to be tested, and determine the probability distribution model of the geometric parameters of the fracture network, the number of rock fracture groups, and the attitude, trace length and width of each fracture group based on the measured data of the rock mass;

[0008] The mean and standard deviation of the fracture orientation, trace length and fracture width of each group are input into the probability distribution model to obtain pseudo-random numbers of each group with standard uniform distribution, and an initial random fracture network is generated by performing a specific transformation on the pseudo-random numbers.

[0009] Determine all intersection points of each fracture segment in the initial random fracture network. Specifically, this includes: inputting the endpoint information of each fracture segment in the initial random fracture network into the planar scanning algorithm and sorting it by coordinate size; establishing the node sequence sequence intersecting the scan line and the fracture segment sequence number; determining whether the node sequence number of each fracture segment is empty; if empty, outputting the node information and fracture segment number; if not empty, extracting the smallest node, dividing it into the left endpoint, right endpoint, and intersection point of the fracture segment based on the smallest node; performing addition, deletion, or position swapping operations on the fracture segment sequence number, and then sorting the fracture segment endpoints again by coordinate size; updating the fracture segment node sequence sequence to obtain all intersection points of the initial random fracture network.

[0010] Based on all the intersections of the initial random fracture network, isolated single fractures, fractures with only one intersection point, and isolated closed loop fractures in the initial random fracture network are removed to generate a two-dimensional random fracture network.

[0011] Furthermore, the probability distribution model of the geometric parameters of the fracture network includes: uniform distribution, negative exponential distribution, normal distribution, or log-normal distribution model.

[0012] Furthermore, the Monte Carlo sampling method is used to obtain pseudo-random numbers from a standard uniform distribution for each group, specifically including the following steps:

[0013] A set of pseudo-random numbers is obtained using the linear congruential method. The process of obtaining this set is as follows:

[0014] x i+1 =(ax i +c)mod(m)

[0015]

[0016] Where, ε i It is a pseudo-random number; x i Let be a pseudo-random variable; m be the modulus, i.e., the maximum period of the production sequence; a be the multiplier, and 0 < a < m; c be the increment, and 0 ≤ c < m; mod is the modulo function; when m is 2k, the parameters a and c are expressed as: The parameter k reflects the period of the random number generator;

[0017] The distribution function F(ε) is obtained from the pseudo-random number sequence. i If we calculate the inverse function F of the distribution function, we obtain the inverse function F. -1 (ε i Let i be the random number sequence number, and let:

[0018] R i =F -1 (ε i )

[0019] Obtain a sample R that conforms to uniform random numbers in the interval [0,1];

[0020] By transforming the sample R accordingly, so that the sample can be applied to different distribution functions in any interval, random numbers x in the interval [s,t] that conform to a uniform distribution, a negative exponential distribution, a normal distribution, or a log-normal distribution are obtained respectively; where the sample transformation expression is:

[0021] x=s+(ts)R

[0022] x = -ln(1-R) / M

[0023] Where M is a probability parameter, referring to the number of times the event occurs per unit time;

[0024] By the Central Limit Theorem, given n independent and identically distributed random variables, and that they follow a uniform distribution in the range [0,1], Let S be its mean and S be its standard deviation, then x asymptotically follows a normal distribution as follows:

[0025]

[0026] Given x that asymptotically follows a normal distribution, the log-normal random numbers, mean, and standard deviation are obtained as follows:

[0027]

[0028] Where x' is a random number that follows a log-normal distribution. S and S' are its mean and standard deviation, respectively.

[0029] Furthermore, the endpoint information of each fracture segment in the initial random fracture network includes the inclination angle, trace length, and coordinates of the random fracture network.

[0030] Furthermore, the block element search method is used to find isolated single cracks and cracks with only one intersection point, specifically including the following steps:

[0031] Based on the output node information and fracture segment numbers, establish sequences Q, P, and R respectively; sequence Q is used to store the order of intersection points on each fracture line; sequence P is used to store the fracture segment number and the number of nodes on that fracture; sequence R is used to store the coordinates of all nodes;

[0032] The search process iteratively identifies the fracture segment number in sequence P where the number of nodes equals 1. Each occurrence of this event triggers sequence Q, which extracts the node information stored in the corresponding fracture and then deletes that node information. The search process then finds and clears the coordinates of the corresponding node information in sequence R. If the number of nodes on all fractures in P is greater than 1 or equal to 0, the search ends, and the fracture segment number, the corresponding node order, and the coordinates are output. If there are fracture segments in P that are not greater than 1 or equal to 0, the search process returns to the initial position and continues to search, eventually finding isolated single fractures and fractures with only one intersection point.

[0033] Furthermore, the method of searching for isolated closed loop fractures using the correlation element search method includes the following steps:

[0034] Read the boundary information and the node information of the fracture network; where the boundary is selected from the leftmost or rightmost fracture in the fracture network.

[0035] The line segments that intersect the boundary are marked as associated line segments and stored in the sequence M;

[0036] Screen the clusters of intersecting line segments of related line segments and remove the marked line segments and duplicate line segments;

[0037] The number of intersecting segments stored in the cluster of intersecting segments is counted. If the number of segments is not 0, the intersecting segment is used as the boundary and the associated segments are re-marked. If the number of intersecting segments is 0, the search ends and all marked segments and node information are output, thus obtaining the isolated closed loop fracture.

[0038] The present invention also provides a system for generating random fracture networks in rock masses, comprising:

[0039] The data acquisition module is used to acquire measured data of the rock mass in the area to be tested, and to determine the probability distribution model of the geometric parameters of the fracture network, the number of rock fracture groups, and the attitude, trace length and width of each fracture group based on the measured data of the rock mass.

[0040] The initial random fracture network generation module is used to input the mean and standard deviation of the fracture orientation, trace length and fracture width of each group into the probability distribution model to obtain pseudo-random numbers of each group with standard uniform distribution, and generate the initial random fracture network by performing a specific transformation on the pseudo-random numbers.

[0041] The judgment module is used to determine all intersection points of each fracture segment in the initial random fracture network. Specifically, it includes: inputting the endpoint information of each fracture segment in the initial random fracture network into the planar scanning algorithm and sorting it by coordinate size to establish the node sequence sequence intersecting with the scan line and the fracture segment sorting sequence; determining whether the node sequence number of each fracture segment is empty. If it is empty, it outputs the node information and fracture segment number. If it is not empty, it takes out the smallest node and divides it into the left endpoint, right endpoint, and intersection point of the fracture segment according to the smallest node. After adding, deleting, or swapping the positions of the fracture segment sorting sequence, it sorts the endpoints of the fracture segment again by coordinate size, updates the fracture segment node sequence sequence, and obtains all intersection points of the initial random fracture network.

[0042] The generation module is used to remove isolated single fractures, fractures with only one intersection point, and isolated closed loop fractures from the initial random fracture network based on all the intersection points of the initial random fracture network, thereby generating a two-dimensional random fracture network.

[0043] The present invention also provides a computer device for generating random fracture networks in rock mass, comprising: a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method for generating random fracture networks in rock mass.

[0044] The present invention also provides a readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, are used to perform the steps of the rock mass random fracture network generation method.

[0045] This invention provides a method, system, device, and medium for generating random fracture networks in rock masses, which has the following beneficial effects:

[0046] This invention rationally groups fractures according to their orientation, determines the number of fracture groups to be simulated, selects a probability distribution model to generate an initial random fracture network, and can accurately simulate the orientation of rock mass fracture networks. By inputting the endpoint information of each fracture segment of the initial random fracture network into a plane scanning algorithm to judge all intersections of each fracture segment, the calculation method is simple and efficient, and can easily obtain the master connected network. It can simulate multiple fracture networks at different scales without restriction. At the same time, it removes special fractures such as isolated fractures that often appear in the initial random fracture network, reducing the complexity of the model's numerical calculation and facilitating the solution of the physical model under different working conditions. Attached Figure Description

[0047] Figure 1 This is a flowchart of the method for generating a random fracture network in rock mass and searching the master connected network in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of two sets of fracture network models in an embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of three sets of fracture network models in an embodiment of the present invention;

[0050] Figure 4 These are three sets of fracture network distribution rose diagrams obtained by a random algorithm in this embodiment of the invention.

[0051] Figure 5 This is a schematic diagram of the planar scanning method calculation process in an embodiment of the present invention;

[0052] Figure 6 This is a schematic diagram of the initial fracture network for planar scanning in an embodiment of the present invention;

[0053] Figure 7 This is a schematic diagram of the block unit calculation process in an embodiment of the present invention;

[0054] Figure 8 This is a schematic diagram illustrating the block unit algorithm for deleting invalid cracks in an embodiment of the present invention;

[0055] Figure 9 This is a schematic diagram of the calculation process of the associated unit in an embodiment of the present invention;

[0056] Figure 10 This is a schematic diagram of the calculation of isolated closed loops by the associated unit in an embodiment of the present invention;

[0057] Figure 11 This is a schematic diagram of the displacement calculation profile change of the rock slope of the main control through-fracture network in an embodiment of the present invention. Detailed Implementation

[0058] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0059] This invention proposes a method for generating a random fracture network in rock mass, specifically including the following steps:

[0060] S1. Using field surveys, borehole data, geophysical data, and other measured data, select a probability distribution model for the geometric parameters of the fracture network to determine the length and width of the simulated region. Probability distribution models for the geometric parameters of the fracture network include: uniform distribution, negative exponential distribution, normal distribution, or log-normal distribution.

[0061] S2. Based on the orientation of fracture development, rationally group the fractures to determine the number of fracture groups to be simulated.

[0062] S3. For each group of fractures, the mean value of their orientation should be reasonably specified. The standard deviation (S), where the dip angle is calculated as the angle between each fracture surface and the longitudinal axis (i.e., the y-axis), is used as an initial value and substituted into the selected distribution model. The mean and standard deviation of the fracture trace length for each group are defined and substituted into the selected distribution model as initial values. The mean and standard deviation of the fracture width for each group are also defined and substituted into the selected distribution model as initial values. After substituting the initial values ​​into each model, a series of random numbers can be obtained using the following calculation method, and their distribution conforms to the selected distribution model, ultimately generating a random fracture network.

[0063] The Monte Carlo sampling method is used to obtain pseudo-random numbers with a standard uniform distribution. The density function of these pseudo-random numbers remains constant within the sampling interval. Specific transformations of these pseudo-random numbers are then applied to obtain random values ​​of geometric parameters that conform to different distribution characteristics, thus completing the random sampling process. The specific steps include:

[0064] (1) Generating pseudo-random numbers using the linear congruential method.

[0065] The linear congruential method obtains a set of pseudo-random numbers using the following recurrence relation:

[0066] x i+1 =(ax i +c)mod(m) (1)

[0067]

[0068] In the formula, ε i It is a pseudo-random number; x i is a pseudo-random variable; m is the modulus, which is the maximum period of the generated sequence; a is the multiplier and 0 < a < m; c is the increment and 0 ≤ c < m; mod is the modulo function. All parameters are positive integers, and i is the pseudo-random number index.

[0069] Statistically, when m is 2k, the formulas for selecting parameters a and c are as follows:

[0070] a = 2 c +1 (3)

[0071]

[0072] Where c is a number near k / 2, which can be approximated as... The parameter k reflects the period of the random number generator; for example, it is 31 for a 32-bit computer system.

[0073] (2) Direct sampling of pseudo-random numbers (Monte-Carlo sampling).

[0074] Based on the pseudo-random number sequence calculated above, a distribution function F(ε) is obtained. i If we have , then we can obtain its inverse function F. -1 (ε i ),make:

[0075] R i =F -1 (ε i (5)

[0076] We can obtain samples that conform to uniform random numbers in the interval [0,1], F(ε) i ) is the distribution function.

[0077] (3) Calculate random values ​​under different distributions using pseudo-random numbers.

[0078] The uniform random number sample R obtained by Monte Carlo sampling is only distributed in the interval [0,1]. In order to make the sample applicable to different distribution functions in any interval, it should be transformed by equations (6) to (9) to finally obtain random numbers x in the interval [s,t] that conform to uniform distribution, negative exponential distribution, normal distribution or log-normal distribution. The transformation expressions for each sample are as follows:

[0079] x=s+(ts)R (6)

[0080] x=-ln(1-R) / M (7)

[0081] In the formula, M is a probability parameter, which refers to the number of times the event occurs per unit time.

[0082] By the Central Limit Theorem, for n independent and identically distributed random variables that follow a uniform distribution in the range [0,1], they asymptotically follow a normal distribution. Therefore, from the above analysis, we can derive the following expression for x that asymptotically follows a normal distribution:

[0083]

[0084] Substituting the result of equation (8) into equation (9) yields the random number, mean, and standard deviation of the log-normal distribution.

[0085]

[0086] In the formula, x' is a random number that follows a log-normal distribution, while S and S' are its mean and standard deviation, respectively.

[0087] S4. Import parameters such as dip angle, trace length, and coordinates of the random fracture network; use the planar scanning algorithm to calculate the node information of the random fracture network and complete the determination of all intersections of the fracture network; specifically including the following steps:

[0088] (1) Import the information of each fracture endpoint in the fracture network.

[0089] (2) Sort the endpoints of the fracture coordinates according to the coordinate size (such as x coordinate value), and establish the node sequence E of the comparable line segments that intersect with the scan line and the fracture line segment sorting sequence T.

[0090] (3) Determine whether the sequence number of each line segment node is empty: if yes, output the node information and line segment number and end the calculation.

[0091] (4) If the number of node sequences of each line segment is not empty, take out the smallest node P, divide the line segment into left endpoint, right endpoint and intersection according to P, and add, delete or swap the line segment sorting sequence respectively, and then return to step (2) to update the line segment node sequence.

[0092] To achieve an ordered arrangement of fracture segment positions at each scan location, a fracture segment sequence T needs to be established during the solution process. This sequence stores the sorting of comparable segments in the fracture network, and is updated once a segment is added, deleted, or its position is swapped. Additionally, to ensure the scan line can sequentially scan the upper endpoints and intersections of each fracture segment, a node sequence sequence E should also be established. Sequence E must arrange the node positions in a specific way (e.g., the size of the x-coordinate). When the scan line encounters the left endpoint, right endpoint, or intersection of a segment, sequence E will activate sequence T to update the segment number sorting. Simultaneously, sequence T itself undergoes a node position reordering. This process continues until no new comparable segments are added or deleted when the scan line moves to the next position, i.e., when sequence E is empty, the scan ends.

[0093] In addition, this method can also calculate whether comparable line segments have a real intersection point, as follows:

[0094] Assume that for any two intersecting line segments V1V2 and V3V4, their intersection point is p. The coordinates of the endpoints of each line segment are represented as follows:

[0095] V i =(x i ,y i (i = 1, 2, 3, 4) (10)

[0096] Therefore, the equation of line segment V1V2 can be expressed as:

[0097] x = x1 + (x2 - x1)t

[0098] y = y1 + (y2 - y1)t (11)

[0099] Similarly, the equation for line segment V3V4 is expressed as follows:

[0100] x = x³ + (x⁴ - x³)T

[0101] y = y3 + (y4 - y3)T (12)

[0102] The parameters t and T both range from [0,1], and represent the position of point (x,y) at a certain proportion of the line segment. When t = 0 or t = 1, they represent the two endpoints of line segment V1V2, respectively; while when t is in (0,1), it represents the position of any point between the two endpoints.

[0103] If point (x, y) is the intersection of line segments V1V2 and V3V4, and both points are located inside the endpoints of the two line segments, then from equations (11) and (12), we know that:

[0104]

[0105] The above system of equations is linear, with only two unknowns, t and T, and can be represented in matrix form as follows:

[0106]

[0107] If the equation has a solution and satisfies:

[0108] 0≤t≤1,0≤T≤1 (15)

[0109] Then, the point (x,y) lies between line segments V1V2 and V3V4, and the values ​​of x and y can be obtained from equation (11) or (12). If a solution exists but does not satisfy equation (15), then the point (x,y) lies on the extension of one of the two line segments. In this case, there is no actual intersection between the two line segments, and their intersection is the intersection of their extensions.

[0110] S5. Use the block element search method to find isolated single cracks (i.e., cracks that do not intersect with other line segments) and crack line segments that have only one intersection point; specifically, this includes the following steps:

[0111] (1) Based on the results of the planar scanning algorithm, sequences Q, P, and R can be established respectively. Sequence Q is used to store the order of intersection points on each fracture line, such as arranging the nodes in ascending order of x position. Sequence P is used to store the fracture number and the number of nodes on the fracture. Sequence R is used to store the coordinates of all nodes.

[0112] (2) Find the crack segment number in sequence P where the number of nodes is equal to 1. Each time this event occurs, sequence Q is triggered, the corresponding crack storage node number in the sequence is extracted, and then the node number is deleted;

[0113] (3) Find the coordinates of the corresponding node number in the sequence R and clear them, thus completing a search for dead end cells;

[0114] (4) If the number of nodes on all fractures in P is greater than 1 or equal to 0, then the search ends and the fracture number, the corresponding node order and coordinates are output.

[0115] (5) If there is a crack segment in P that is no greater than 1 or equal to 0, then return to the initial search position and search cyclically.

[0116] S6. In seepage analysis of fractured rock masses, the connected main fracture network contributes significantly to the seepage flow. Isolated fractures, single-intersection fractures, and closed-loop fractures (i.e., multiple intersecting fractures not connected to the main fracture network) have smaller seepage flows due to their disconnected fracture networks. To simplify the analysis, they can be excluded from the seepage calculation. Therefore, they need to be removed from the initial random fracture network. After finding isolated fractures and single-intersection fractures in S5, the associated element search method is used to find isolated closed-loop fractures. Finally, these three types of fractures are deleted to simplify the calculation. The specific steps include the following:

[0117] (1) Read the boundary conditions and fracture network node information. The boundary can be selected as the leftmost or rightmost fracture, or the vertical line passing through the leftmost or rightmost fracture can be specified as the sweep boundary. The network nodes are swept sequentially from the boundary.

[0118] (2) Mark the line segments that intersect with the boundary as associated line segments and store them in the sequence M;

[0119] (3) Screen the clusters of intersecting line segments of related line segments and remove the marked line segments and duplicate line segments;

[0120] (4) Count the number of intersecting line segments stored in the intersecting line cluster. If the number of line segments is not 0, take the intersecting line segment as the boundary and return to step (2) to calculate repeatedly. If the number of intersecting line segments is 0, the search ends and all marked line segments and node information are output.

[0121] S7. Generate a two-dimensional random fracture network model that does not consider the master connectivity of isolated single fractures and closed loops.

[0122] Example 1:

[0123] In one or more embodiments, a method for generating a random fracture network in rock mass is disclosed, specifically including the following process:

[0124] 1. Based on field surveys, borehole data, geophysical data, and other measured data, a probability distribution model of the geometric parameters of the fracture network was obtained to determine the length and width of the simulated region. See Table 1.

[0125] Table 1. Grouping of fissure orientation distribution in natural rock slopes

[0126]

[0127]

[0128] In this embodiment, the model size is 20m×20m, the fracture dip angle follows a uniform distribution, and the trace length is assumed to follow a log-normal distribution.

[0129] 2. Based on the orientation of fracture development, the fractures are reasonably grouped to determine the number of fracture groups to be simulated.

[0130] In this embodiment, the natural fissures are mainly divided into 3 groups, and the average dip angles of each group are 0 degrees, 35 degrees and 55 degrees respectively.

[0131] 3. For each group of fractures, the mean value of their orientation should be reasonably specified. The standard deviation (S), where the dip angle is calculated as the angle between each fracture surface and the longitudinal axis (i.e., the y-axis), is used as the initial value and substituted into the distribution model selected in step 1.

[0132] 4. Specify the mean and standard deviation of the crack trace length for each group, and substitute them as initial values ​​into the distribution model selected in step 1.

[0133] 5. Specify the mean and standard deviation of the crack width for each group, and substitute them as initial values ​​into the distribution model selected in step 1.

[0134] In this embodiment, the initial geometric parameters of each group of cracks are shown in Table 2:

[0135] Table 2. Values ​​of geometric parameters of each group of fractures

[0136]

[0137] 6. Output the simulation results and perform post-processing on the results through secondary development of the plotting software.

[0138] In this embodiment, Figure 3 This demonstrates the relationship between the above parameters and the three sets of fracture network models obtained from the probability distribution model simulation. Figure 4 This indicates the distribution range of each group of cracks in the simulated model.

[0139] This embodiment performs post-processing of the results, including using plotting software (CAD, Tecplot, etc.) to plot the simulation data, and combining the model size selected in step 1 to trim the model to obtain the final fracture network diagram.

[0140] Example 2

[0141] In one or more embodiments, a two-dimensional random fracture network planar scanning algorithm is disclosed, combined with Figure 5 Specifically, it includes the following processes:

[0142] 1. Import the information of each fracture endpoint in the fracture network.

[0143] In this embodiment, Figure 6 This demonstrates the initial fracture network obtained using a two-dimensional stochastic simulation method, which is used for fracture network intersection determination and intersection point calculation in this embodiment. The model size is 9cm × 9cm, and two sets of fractures with a total of 16 fracture segments are randomly generated.

[0144] 2. Sort the endpoints of the fracture coordinates according to the coordinate size (such as the x-coordinate value), and establish the node sequence E of the comparable line segments that intersect with the scan line and the fracture line segment sorting sequence T.

[0145] 3. Determine if the sequence number of each line segment node is empty: if yes, output the node information and line segment number and end the calculation.

[0146] 4. If the number of node sequences of each line segment is not empty, take out the smallest node P, divide the line segment into left endpoint, right endpoint and intersection according to P, and add, delete or swap the line segment sorting sequence respectively, and then return to step (2) to update the line segment node sequence.

[0147] In this embodiment, a calculation program was written using the Fortran language. The scan was performed from the left boundary of the model to the right boundary, resulting in a total of 23 intersection points. All intersection points were arranged in ascending order of their x-coordinates. Detailed information such as the coordinates of each node and the number of its intersecting line segments is listed in Table 3.

[0148] Table 3. Coordinates of intersection points and intersection segment numbers of the fracture network

[0149]

[0150]

[0151] Example 3

[0152] In one or more embodiments, a block element calculation method is disclosed, combined with Figure 7 Specifically, it includes the following processes:

[0153] 1. Based on the results of the planar scanning algorithm, sequences Q, P, and R can be established respectively. Sequence Q is used to store the order of intersection points on each fracture line, such as arranging the nodes in ascending order of x position. Sequence P is used to store the fracture number and the number of nodes on that fracture, while sequence R is used to store the coordinates of all nodes.

[0154] In this embodiment, as Figure 8 The diagram shows two sets of initial fractures, with 5 and 8 fractures in each set, respectively. The fracture network has 17 intersection points, arranged in ascending order of x-coordinate. Write a Fortran program to find isolated single fractures and fracture segments with only one intersection point.

[0155] 2. Successively find the crack segment number in sequence P where the number of nodes is equal to 1. Each time this event occurs, it will trigger sequence Q, extract the corresponding crack storage node number in the sequence, and then delete the node number.

[0156] In this embodiment, as Figure 8 The diagram shows three fractures, l2, l3, and l5, with a total of 1 intersection point, and another fracture, l1, with a total of 0 intersection points. The calculation program then reads the sequence Q and sets the corresponding node number to null.

[0157] 3. Find the coordinates of the corresponding node number in the sequence R and clear them, thus completing one search for dead-end cells.

[0158] 4. If the number of nodes on all fractures in P is greater than 1 or equal to 0, then the search ends, and the fracture number, the corresponding node order, and the coordinates are output.

[0159] 5. If there is a crack segment in P that is no greater than 1 or equal to 0, then return to the initial search position and continue searching.

[0160] In this embodiment, a calculation program is used to traverse the fracture number sequence P and save the coordinates and nodes of all fractures corresponding to nodes that are not equal to 1, so as to facilitate the later drawing of the fracture network diagram.

[0161] Example 4

[0162] In one or more embodiments, a method for calculating associated units is disclosed, combined with Figure 9 Specifically, it includes the following processes:

[0163] 1. Read boundary conditions and fracture network node information.

[0164] In this embodiment, as Figure 10 As shown, the left boundary line is first assumed to be the boundary condition, and the crack network nodes are calculated by the block elements and then imported.

[0165] 2. Mark the line segments that intersect the boundary as associated line segments and store them in sequence M.

[0166] 3. Screen the clusters of intersecting line segments of related line segments and remove the marked line segments and duplicate line segments.

[0167] In this embodiment, the associated line segments in sequence M are used as boundaries to search for intersecting line segments and stored in sequence N. To prevent duplicate counting of intersecting line segments, the intersecting line segments in sequence N need to be screened, removing the marked line segments in sequence M and merging the line segment numbers that appear twice or more in N.

[0168] 4. Count the number of intersecting line segments stored in the intersecting line cluster. If the number of line segments is not 0, take the intersecting line segment as the boundary and return to step (2) to calculate repeatedly. If the number of intersecting line segments is 0, the search ends and all marked line segments and node information are output.

[0169] Example 5:

[0170] In one or more embodiments, a process for jointly applying a method for generating a random fracture network in rock mass and a method for searching a master connected network is disclosed, combined with... Figure 11 Specifically, it includes the following processes:

[0171] 1. Select the probability distribution model of the geometric parameters of the fracture network based on the measured data.

[0172] 2. Group the fractures appropriately according to their orientation.

[0173] In this embodiment, the model has a slope angle of 60°, a slope height of 6.95m, a horizontal length of 3.97m, and a slope surface length of 8m. There are two sets of internal fissures in the rock slope. To simplify the modeling process, it should also be noted that the rock slope boundary line is also considered as a third set of fissures in this example.

[0174] 3. Using the Monte-Carlo sampling method, randomized split networks are generated using the Fortran language.

[0175] In this embodiment, a program written in Fortran is used to generate two sets of fracture networks inside the rock slope. The rock slope boundary line is a known condition and is directly read into the program. Figure 11 The initial fracture network profile of the rock slope needs to have some fractures extending beyond the rock slope boundary deleted.

[0176] 4. Use the planar scanning algorithm to calculate the node information of the random fracture network and determine all intersections of the fracture network.

[0177] 5. Use the block element search method to find isolated single cracks and cracks with only one intersection point.

[0178] In this embodiment, as Figure 11 As shown, the calculation program obtained a total of 3 special fractures, including isolated fractures and dead-end fractures. Figure 11 In section a (1-3), after deleting these types of fractures, the dominant connected fracture network is obtained as follows: Figure 11 b.

[0179] 6. Use the associated cell search method to find isolated closed loops.

[0180] 7. Generate a master-controlled, interconnected two-dimensional random gap network model.

[0181] Based on the same inventive concept, this invention also proposes a system for generating random fracture networks in rock masses, comprising:

[0182] The data acquisition module is used to acquire measured data of the rock mass in the area to be tested, and to determine the probability distribution model of the geometric parameters of the fracture network, the number of rock fracture groups, and the attitude, trace length and width of each fracture group based on the measured data of the rock mass.

[0183] The initial random fracture network generation module is used to input the mean and standard deviation of the fracture orientation, trace length and fracture width of each group into the probability distribution model to obtain pseudo-random numbers of each group with standard uniform distribution, and generate the initial random fracture network by performing a specific transformation on the pseudo-random numbers.

[0184] The judgment module is used to determine all intersection points of each fracture segment in the initial random fracture network. Specifically, it includes: inputting the endpoint information of each fracture segment in the initial random fracture network into the planar scanning algorithm and sorting it by coordinate size to establish the node sequence sequence intersecting with the scan line and the fracture segment sorting sequence; determining whether the node sequence number of each fracture segment is empty. If it is empty, it outputs the node information and fracture segment number. If it is not empty, it takes the smallest node and divides it into the left endpoint, right endpoint, and intersection point of the fracture segment according to the smallest node. After adding, deleting, or swapping the positions of the fracture segment sorting sequence, it sorts the endpoints of the fracture segment again by coordinate size, updates the fracture segment node sequence sequence, and obtains all intersection points of the initial random fracture network.

[0185] The generation module is used to remove isolated single fractures, fractures with only one intersection point, and isolated closed loop fractures from the initial random fracture network based on all the intersection points of the initial random fracture network, thereby generating a two-dimensional random fracture network.

[0186] The present invention also proposes a computer device for generating random fracture networks in rock mass, comprising: a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method for generating random fracture networks in rock mass.

[0187] The present invention also proposes a readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, are used to perform the steps of the method for generating a random fracture network in rock mass.

[0188] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for generating a random fracture network in rock mass, characterized in that, Includes the following steps: Obtain measured data of the rock mass in the area to be tested, and determine the probability distribution model of the geometric parameters of the fracture network, the number of rock fracture groups, and the attitude, trace length and width of each fracture group based on the measured data of the rock mass; The mean and standard deviation of the fracture orientation, trace length and fracture width of each group are input into the probability distribution model to obtain pseudo-random numbers of each group with standard uniform distribution, and an initial random fracture network is generated by performing a specific transformation on the pseudo-random numbers. Determine all intersections of each fracture segment in the initial random fracture network; Specifically, this includes: inputting the endpoint information of each fracture segment of the initial random fracture network into the planar scanning algorithm and sorting them by coordinate size; establishing the node sequence sequence intersecting the scan line and the fracture segment sorting sequence; determining whether the node sequence number of each fracture segment is empty; if empty, outputting the node information and fracture segment number; if not empty, taking out the smallest node and dividing it into the left endpoint, right endpoint, and intersection point of the fracture segment according to the smallest node; performing addition, deletion, or position swapping operations on the fracture segment sorting sequence and then sorting the endpoints of the fracture segment again by coordinate size; updating the fracture segment node sequence sequence; and obtaining all the intersection points of the initial random fracture network. Based on all the intersections of the initial random fracture network, isolated single fractures, fractures with only one intersection point, and isolated closed loop fractures in the initial random fracture network are removed to generate a two-dimensional random fracture network. The method employs a block element search to find isolated single fractures and fractures with only one intersection point. Specifically, it includes the following steps: Establishing sequences Q, P, and R based on the output node information and fracture segment numbers; sequence Q stores the intersection order of each fracture line; sequence P stores the fracture segment number and the number of nodes on each fracture; sequence R stores the coordinates of all nodes; iteratively identifies fracture segment numbers in sequence P where the number of nodes is not greater than 1; each occurrence of this event triggers sequence Q, extracts the corresponding fracture's stored node information, and then deletes the node information; it finds and clears the coordinates of the corresponding node information in sequence R; if the number of nodes on all fractures in P is greater than 1, the search ends, and the fracture segment number, corresponding node order, and coordinates are output; if there are fracture segments in P with a number not greater than 1, it returns to the initial search position and loops through the search to find isolated single fractures and fractures with only one intersection point. The method of searching for isolated closed-loop fractures using associated unit search includes the following steps: reading boundary information and node information of the fracture network; wherein, the boundary is selected from the leftmost or rightmost fracture in the fracture network; marking the line segments intersecting the boundary as associated line segments and storing them in a sequence; screening the clusters of intersecting line segments of associated line segments, removing the marked line segments and duplicate line segments; counting the number of intersecting line segments stored in the clusters of intersecting line segments, if the number of line segments is not 0, then the intersecting line segment is used as the boundary and the associated line segment is re-marked; if the number of intersecting line segments is 0, the search ends, and all marked line segments and node information are output, thus obtaining the isolated closed-loop fractures.

2. The method for generating a random fracture network in rock mass according to claim 1, characterized in that, The probability distribution model of the geometric parameters of the fracture network includes: uniform distribution, negative exponential distribution, normal distribution or log-normal distribution model.

3. The method for generating a random fracture network in rock mass according to claim 1, characterized in that, The Monte Carlo sampling method is used to obtain pseudo-random numbers from a standard uniform distribution for each group. The specific steps include: A set of pseudo-random numbers is obtained using the linear congruential method. The process of obtaining this set is as follows: in, ε i It is a pseudo-random number; x i It is a pseudo-random variable; m The modulus is the maximum period of the production sequence. a It is a multiplier, and 0 < a < m ; c It is an increment, and 0 ≤ c < m mod is the modulo function; when m 2 k At that time, parameters a and c Represented as: , ,parameter k It reflects the period of the random number generator; A distribution function is obtained from the pseudo-random number sequence. Then we obtain the inverse function of the distribution function. , i Let be the random number sequence number, and let: Obtain samples that conform to uniform random numbers in the interval [0,1]. R ; By analyzing the samples R By performing the corresponding transformations, the samples are applied to different distribution functions in any interval, resulting in […]. s , t Random numbers that conform to a uniform distribution, negative exponential distribution, normal distribution, or log-normal distribution on an interval x The sample transformation expression is: in, M This is a probability parameter, referring to the number of times an event occurs per unit of time; By the central limit theorem, n There are n independent and identically distributed random variables, which follow a uniform distribution in the range [0,1]. Its mean, S If its standard deviation is given, then it asymptotically follows a normal distribution. x for: Based on the fact that it asymptotically follows a normal distribution x The random numbers, mean, and standard deviation obtained from a log-normal distribution are expressed as follows: in, x 'A random number that follows a log-normal distribution,' and S 'These are its mean and standard deviation, respectively.

4. The method for generating a random fracture network in rock mass according to claim 1, characterized in that, The endpoint information of each fracture segment in the initial random fracture network includes the inclination angle, trace length, and coordinates of the random fracture network.

5. A system for generating random fracture networks in rock mass, characterized in that, include: The data acquisition module is used to acquire measured data of the rock mass in the area to be tested, and to determine the probability distribution model of the geometric parameters of the fracture network, the number of rock fracture groups, and the attitude, trace length and width of each fracture group based on the measured data of the rock mass. The initial random fracture network generation module is used to input the mean and standard deviation of the fracture orientation, trace length and fracture width of each group into the probability distribution model to obtain pseudo-random numbers of each group with standard uniform distribution, and generate the initial random fracture network by performing a specific transformation on the pseudo-random numbers. The judgment module is used to determine all intersections of each fracture segment in the initial random fracture network. Specifically, this includes: inputting the endpoint information of each fracture segment of the initial random fracture network into the planar scanning algorithm and sorting them by coordinate size; establishing the node sequence sequence intersecting the scan line and the fracture segment sorting sequence; determining whether the node sequence number of each fracture segment is empty; if empty, outputting the node information and fracture segment number; if not empty, taking out the smallest node and dividing it into the left endpoint, right endpoint, and intersection point of the fracture segment according to the smallest node; performing addition, deletion, or position swapping operations on the fracture segment sorting sequence and then sorting the endpoints of the fracture segment again by coordinate size; updating the fracture segment node sequence sequence; and obtaining all the intersection points of the initial random fracture network. The generation module is used to remove isolated single fractures, fractures with only one intersection point, and isolated closed loop fractures from the initial random fracture network based on all intersection points, generating a two-dimensional random fracture network. The module uses a block element search method to find isolated single fractures and fractures with only one intersection point, specifically including the following steps: Establishing sequences Q, P, and R based on the output node information and fracture segment numbers; sequence Q stores the order of intersection points on each fracture line; sequence P stores the fracture segment number and the number of nodes on that fracture; sequence R stores the coordinates of all nodes; sequentially finding fracture segment numbers in sequence P where the number of nodes is no greater than 1; each time this event occurs, sequence Q is triggered, the corresponding fracture's stored node information is extracted from the sequence, and then the node information is deleted; the coordinates of the corresponding node information are found in sequence R and cleared. If the number of nodes on all fractures in P is greater than 1, the search ends, and the fracture segment number, corresponding node order, and coordinates are output. If there is a fracture segment in P that is not greater than 1, the search returns to the initial position and loops to find isolated single fractures and fractures with only one intersection point. The method of searching for isolated closed loop fractures using the associated unit search method includes the following steps: reading boundary information and node information of the fracture network; wherein, the boundary is selected as the fracture at the leftmost or rightmost end of the fracture network; marking the line segments that intersect with the boundary as associated line segments and storing them in a sequence; screening the clusters of intersecting line segments of associated line segments and removing the marked line segments and duplicate line segments; counting the number of intersecting line segments stored in the clusters of intersecting line segments, if the number of line segments is not 0, then the intersecting line segment is used as the boundary and the associated line segment is re-marked; if the number of intersecting line segments is 0, the search ends, and all marked line segments and node information are output, thus obtaining isolated closed loop fractures.

6. A computer device for generating random fracture networks in rock mass, characterized in that, include: The memory, the processor, and the computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method for generating a random fracture network in rock mass according to any one of claims 1-4.

7. A readable storage medium, characterized in that, The readable storage medium stores a computer program, which includes program instructions that, when executed by a processor, are used to perform the steps of the method for generating a random fracture network in rock mass according to any one of claims 1-4.

Images