A reliability evaluation method for block systems based on geometry-topology consistency

By adopting a geometric-topological consistency-based reliability evaluation method for block systems, the lack of reliability evaluation for block systems is addressed, and quantitative analysis of the structural consistency of block systems is achieved, ensuring the accuracy and reliability of rock mechanics simulation.

CN123069696APending Publication Date: 2026-11-10ZHEJIANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611308066.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-27
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing research lacks methods for evaluating the reliability of three-dimensional block systems, making it difficult to guarantee the reliability and repeatability of rock mechanics simulation results. In particular, when generating multiple statistically equivalent DFN models, the structural properties of the block system have a significant impact.

Method used

A reliability evaluation method for block systems based on geometric-topological consistency is proposed. By constructing block system pairs, geometric, topological, and comprehensive consistency indices are calculated, including size consistency, shape consistency, and topological consistency indices. Spectral clustering and optimal transmission models are used for comprehensive evaluation.

Benefits of technology

It enables quantitative evaluation of block systems, overcomes the limitations of judging by a single geometric attribute, comprehensively reflects the consistency of block systems, provides an objective basis for the reliability of DFN-generated block systems, and ensures the accuracy and repeatability of rock mechanics simulation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123069696A_ABST
    Figure CN123069696A_ABST
Patent Text Reader

Abstract

This invention discloses a reliability evaluation method for block systems based on geometric-topological consistency, comprising the following steps: S1, constructing several block systems based on the structural surface statistical parameters of the target rock mass, and combining all block systems in pairs to obtain several block system pairs; S2, sequentially calculating the geometric, topological, and comprehensive consistency indices of all block system pairs; S3, performing reliability analysis based on the comprehensive consistency indices of the block systems. This invention can quantitatively verify the reliability of block systems before conducting numerical simulations of discontinuous media mechanics, providing a basis for determining whether the generated block systems can serve as reasonable structural models for subsequent rock mass mechanics analysis.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of three-dimensional rock block recognition technology, specifically involving a reliability evaluation method for block systems based on geometric-topological consistency. Background Technology

[0002] Rock mass is a complex geological medium composed of structural planes and rock blocks. Predicting and evaluating its mechanical behavior is one of the core tasks in the field of rock mass engineering. Discrete fracture network-discrete element method (DFN-DEM) is a commonly used technical framework for analyzing the mechanical behavior of fractured rock mass. It mainly includes the following three steps: (1) constructing a discrete fracture network; (2) performing three-dimensional block identification; and (3) conducting mechanical simulation of the block system. Among them, the discrete fracture network is usually randomly generated by Monte Carlo method based on the statistical data of structural planes obtained from field investigation; three-dimensional block identification is used to calculate the polyhedral blocks formed by the mutual cutting of structural planes, and thereby construct the rock mass block system; the discrete element method, discontinuous deformation analysis method and other discontinuous medium numerical methods are based on the generated block system to analyze the deformation, failure and instability behavior of the rock mass. Therefore, the block system is the key intermediate model connecting discrete fracture network modeling and mechanical simulation, and its structural properties will directly affect the subsequent mechanical simulation results.

[0003] Existing research has extensively investigated the reliability of the DFN–DEM framework from two aspects: discrete fracture network construction and discrete element simulation (DEM). For discrete fracture network construction, the DFN model is typically validated by comparing the corresponding sections of the field-mapped profiles with those of the DFN model, or by comparing the geometric and statistical characteristics of their structural surfaces. For DEM, the accuracy of simulation results is improved primarily by refining basic theoretical assumptions, enhancing the convergence of numerical calculations, and increasing the precision of contact detection. In contrast, research on the reliability of block systems generated by 3D block identification is relatively limited. Existing research mainly focuses on improving the computational efficiency of block identification algorithms, handling small-volume blocks, and adapting to different types of structural surface models. For the same rock mass, multiple statistically equivalent DFN models can be generated based on the same field statistical data, and these can be transformed into different block systems through 3D block identification. The structural properties of these block systems include geometric and topological properties, both of which directly affect the subsequent DEM simulation results. However, there is currently a lack of evaluation methods to determine whether the block systems generated by statistically equivalent DFN models have consistent structural characteristics, making it difficult to guarantee the reliability and repeatability of subsequent mechanical prediction results. Therefore, it is necessary to quantify the geometric and topological consistency of different block systems and establish corresponding reliability evaluation methods for block systems based on this. Summary of the Invention

[0004] To address the above problems, this invention proposes a reliability evaluation method for block systems based on geometric-topological consistency.

[0005] The technical solution of this invention is: a reliability evaluation method for block systems based on geometric-topological consistency, comprising the following steps:

[0006] S1. Based on the statistical parameters of the structural surfaces of the target rock mass, construct several block systems and combine all block systems in pairs to obtain several block system pairs;

[0007] S2. Calculate the geometric, topological, and overall consistency indices for all block system pairs in sequence.

[0008] S3. Conduct reliability analysis based on the comprehensive consistency index of the block system.

[0009] Furthermore, S1 includes the following sub-steps:

[0010] S11. Based on the statistical parameters of the structural surfaces of the target rock mass, several discrete fracture network models are generated using the Monte Carlo method.

[0011] S12. The three-dimensional block recognition method is used to identify the blocks of each discrete fracture network model to obtain the corresponding block system data.

[0012] S13. Combine all the block systems in pairs to form several block system pairs.

[0013] Furthermore, in S12, the block system data includes the vertex, edge, face, and volume data of the block system, as well as the topological relationships between the geometric elements.

[0014] Furthermore, S2 includes the following sub-steps:

[0015] S21. Select a block system pair as the first block system and the second block system; calculate the volume of all blocks in each of the two block systems, and arrange the volumes in ascending order; for each block, calculate the cumulative volume of all blocks whose volume is not greater than that of the block, and use its proportion to the total volume of the system as the ordinate, and plot the common logarithm of the block volume log. 10 (V) is the abscissa, and cumulative block volume distribution curves are constructed respectively; where V represents volume;

[0016] S22. Calculate the area difference between the two cumulative block volume distribution curves in the same interval, and normalize the area difference as the size distribution difference index. Calculate the size consistency index of the first block system and the second block system.

[0017] S23. Construct a shape similarity matrix for the first block system and the second block system;

[0018] S24. Based on the shape similarity matrix, the first block system and the second block system are clustered and grouped using the spectral clustering method respectively;

[0019] S25. Based on the clustering results, calculate the proportion of the sum of the block volumes of each size category to the total volume of the corresponding shape group, and arrange each size category in descending order of volume contribution proportion. Determine the dominant size category of each block group according to the preset cumulative volume contribution threshold.

[0020] S26. Calculate the average shape similarity between all blocks in the dominant size category of each block group and all other blocks in the block group, and select the block with the largest average shape similarity as the representative shape of the group.

[0021] S27. Calculate the proportion of the number of blocks in each block group in the first block system and the second block system respectively, and use the proportion of each shape group as the weight of the corresponding representative shape to obtain the discrete shape distribution of the first block system and the second block system respectively.

[0022] S28. Calculate the shape similarity between the representative shapes in the first block system and the representative shapes in the second block system, and calculate the corresponding unit transportation cost. Obtain the unit transportation cost between all representative shapes in the first block system and the second block system, and construct a transportation cost matrix.

[0023] S29. Establish an optimal transmission model with the goal of minimizing the total transportation cost between the discrete shape distributions of the first block system and the second block system, solve the optimal transmission model to obtain the minimum total transportation cost, and calculate the shape consistency index of the first block system and the second block system.

[0024] S210, Construct a complex mesh for the first block system and the second block system;

[0025] S211. Extract the local connectivity, global connectivity, and structural partitioning features of complex meshes and use them as topology evaluation indicators.

[0026] S212. Calculate the relative symmetry error of the first block system and the second block system on each topological index respectively, and use it as the consistency index of different topological features between the two, and finally obtain the overall topological consistency evaluation result.

[0027] S213. Calculate the geometric consistency index by weighting the size consistency index and the shape consistency index with the same weight.

[0028] S214. Determine the geometric consistency weight and topological consistency weight, and calculate the comprehensive consistency index of the first block system and the second block system;

[0029] S215. Determine if there are still uncalculated block system pairs to be evaluated: If they exist, extract the next set of block system pairs and repeat S21-S214; otherwise, proceed to S3.

[0030] Furthermore, in S22, the area difference between the two cumulative block volume distribution curves in the interval [a,b] for:

[0031] ;

[0032] Among them, F A (x) represents the cumulative block volume distribution curve of the first block system, F B (x) represents the cumulative volume distribution curve of the second block system, a represents the lower limit of the integration interval, and b represents the upper limit of the integration interval;

[0033] In S22, the area difference normalization is specifically implemented as follows:

[0034] ;

[0035] in, This represents an index indicating the difference in size distribution after normalization.

[0036] In S22, the dimensional consistency index for:

[0037] .

[0038] Furthermore, in S23, any two blocks in the block system are scaled to the same volume, and the maximum volume overlap rate between the two normalized blocks is calculated as the shape similarity between the two blocks; all shape similarity indices are used as matrix elements to construct a shape similarity matrix;

[0039] The method for generating the maximum volume overlap rate between two normalized blocks is as follows: the point cloud registration method is used to spatially align the point clouds of the two normalized blocks so that the volume overlap of the two normalized blocks reaches the maximum. The proportion of the number of points in the point cloud of one block falling into the other block after point cloud registration is calculated as the proportion of the total number of points in the point cloud. This proportion is used as the maximum volume overlap rate between the two normalized blocks.

[0040] Furthermore, S24 includes the following sub-steps:

[0041] S241. Set the initial number of cluster groups to 2, and use spectral clustering to group the shape similarity matrix of the block system.

[0042] S242. In the clustered block groups, the groups with fewer than the minimum group size threshold are designated as random block groups, and the remaining groups are designated as grouped block groups.

[0043] S243. Calculate the average shape similarity among all blocks within each group of blocks. If the average shape similarity of all groups of blocks is greater than or equal to the minimum average shape similarity threshold, proceed to S25; otherwise, proceed to S244.

[0044] S244. Increment the initial number of clusters by 1, and repeat S241-S243 until the clustering results meet the preset conditions.

[0045] Further, in S25, determining the dominant size category for each block group is specifically as follows: the size categories are arranged in descending order of volume contribution ratio. If the volume contribution ratio of the first-ranked size category is greater than the preset cumulative volume contribution threshold, then the size category is determined as the dominant size category; otherwise, subsequent size categories are added in sequence according to the ranking result until the cumulative volume contribution ratio of the selected size category is greater than the preset cumulative volume contribution threshold, and the selected size categories are collectively determined as the dominant size categories.

[0046] In S27, the discrete shape distribution P of the first block system A The discrete shape distributions of the second block system are respectively P B for:

[0047] ;

[0048] Where, n A n represents the number of block groups in the first block system. B Indicates the number of block groups in the second block system; w i v represents the weight of the proportion of blocks in the first block system. j δ represents the weight of the proportion of the block group corresponding to the second block system. Ai δ represents the representative shape of the block group corresponding to the first block system. Bj This represents the representative shape of the block group corresponding to the second block system;

[0049] In S28, the unit transportation cost C ij for:

[0050] ;

[0051] Among them, S ij This represents the shape similarity between the i-th representative shape in the first block system and the j-th representative shape in the second block system;

[0052] In S29, the optimal transmission model is:

[0053] ;

[0054] Among them, T ij The weight for transporting the i-th shape group of the first block system to the j-th shape group of the second block system;

[0055] The constraints of the optimal transmission model are:

[0056] ;

[0057] In S29, the shape consistency index of the first block system and the second block system for:

[0058] ;

[0059] In S210, the construction of a complex mesh is specifically as follows: the blocks of the first block system and the second block system are converted into network nodes respectively; it is determined whether any two blocks have a shared surface. If the two blocks have at least one shared surface, an undirected edge is established between the corresponding nodes, and the number of shared surfaces between the two blocks is determined as the weight of the undirected edge; if the two blocks do not have a shared surface, no network edge is established between the corresponding nodes.

[0060] In S212, the symmetric relative error E between the k-th topological index of the first block system and the second block system k for:

[0061] ;

[0062] Where, x Ak Let x represent the k-th topological index of the first block system. Bk This represents the k-th topological index of the second block system.

[0063] In S212, the k-th topological consistency index C k for:

[0064] ;

[0065] In S212, the comprehensive topology consistency index C topo for:

[0066] ;

[0067] Where C1 represents the first topology consistency index, C2 represents the second topology consistency index, and C3 represents the third topology consistency index;

[0068] In S213, the geometric consistency index C geo for:

[0069] ;

[0070] In S214, the overall consistency index C system for:

[0071] ;

[0072] Among them, w g This represents the geometric consistency weight.

[0073] Furthermore, in S3, the average value and coefficient of variation of the overall consistency index for all block systems are calculated;

[0074] Based on the average value and coefficient of variation of the comprehensive consistency index, the structural consistency of several block systems is judged. When the average value of the comprehensive consistency index is greater than or equal to the set average value threshold and the coefficient of variation is less than or equal to the set coefficient of variation threshold, the reliability of the generated result of the block system is deemed qualified; otherwise, the reliability of the block system is deemed unqualified.

[0075] The beneficial effects of this invention are:

[0076] 1. This invention proposes a reliability evaluation method for DFN-generated block systems based on geometric-topological consistency. It comprehensively analyzes the block systems generated by statistically equivalent DFN from three dimensions: block size, block shape, and system topology. This method overcomes the limitations of judging the differences between block systems based solely on a single geometric attribute or qualitative observation, and can more comprehensively reflect the degree of structural consistency between block systems.

[0077] 2. This invention compares block systems generated by multiple statistically equivalent DFN models in pairs, and characterizes the overall level and dispersion of the block system's structural consistency based on the average value and coefficient of variation of each block system's comprehensive consistency index. This quantitatively determines the sensitivity of the block system to the randomness of DFN, providing a unified basis for the objective evaluation of the reliability of DFN-generated block systems.

[0078] 3. The size consistency, shape consistency and topology consistency calculations used in this invention are all based on the block system data obtained from block identification. The calculation process is clear, objective and easy to program. It can quantitatively verify the reliability of the block system before carrying out numerical simulation of discontinuous media mechanics, and provide a basis for judging whether the generated block system can be used as a reasonable structural model for subsequent rock mechanics analysis. Attached Figure Description

[0079] Figure 1This is a flowchart of a reliability evaluation method for block systems based on geometry-topology consistency.

[0080] Figure 2 This is a schematic diagram of the cumulative volume distribution curve of the block;

[0081] Figure 3 A schematic diagram of the area deviation of the cumulative volume distribution curve of the block;

[0082] Figure 4 A schematic diagram of the shape similarity matrix construction for a block system;

[0083] Figure 5 This is a schematic diagram illustrating the clustering and grouping of block systems based on shape similarity.

[0084] Figure 6 A schematic diagram for determining the dominant dimensions and representative shapes;

[0085] Figure 7 A schematic diagram illustrating the construction of an optimal transport model between shape distributions;

[0086] Figure 8 A schematic diagram of the network representation of a block system;

[0087] Figure 9 A schematic diagram of a block system pair used for comparing consistency;

[0088] Figure 10 A box plot showing the consistency of geometric properties;

[0089] Figure 11 A box plot showing the consistency of topological attributes;

[0090] Figure 12 For geometric, topological, and comprehensive consistency box plots. Detailed Implementation

[0091] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0092] like Figure 1 As shown, this invention provides a reliability evaluation method for block systems based on geometry-topology consistency, comprising the following steps:

[0093] S1. Based on the statistical parameters of the structural surfaces of the target rock mass, construct several block systems and combine all block systems in pairs to obtain several block system pairs;

[0094] S2. Calculate the geometric, topological, and overall consistency indices for all block system pairs in sequence.

[0095] S3. Conduct reliability analysis based on the comprehensive consistency index of the block system.

[0096] In this embodiment of the invention, S1 includes the following sub-steps:

[0097] S11. Based on the statistical parameters of the structural surfaces of the target rock mass, several discrete fracture network models are generated using the Monte Carlo method.

[0098] S12. The three-dimensional block recognition method is used to identify the blocks of each discrete fracture network model to obtain the corresponding block system data.

[0099] S13. Combine all the block systems in pairs to form several block system pairs.

[0100] Based on the statistical parameters of the structural surfaces of the target rock mass, N statistically equivalent discrete fracture network models are generated using the Monte Carlo method. A three-dimensional block identification method is then used to identify the blocks in each discrete fracture network model, resulting in N corresponding block system data, including the vertex, edge, face, and volume data of each block system, as well as the topological relationships between the geometric elements.

[0101] By combining all the block systems in pairs, a total of N(N−1) / 2 unordered block system pairs are formed.

[0102] In this embodiment of the invention, in S12, the block system data includes the vertex, edge, face, and volume data of the block system, as well as the topological relationships between the geometric elements.

[0103] In this embodiment of the invention, S2 includes the following sub-steps:

[0104] S21. Select a block system pair as the first block system and the second block system; calculate the volume of all blocks in each of the two block systems, and arrange the volumes in ascending order; for each block, calculate the cumulative volume of all blocks whose volume is not greater than that of the block, and use its proportion to the total volume of the system as the ordinate, and plot the common logarithm of the block volume log. 10 (V) is the abscissa, and cumulative block volume distribution curves are constructed respectively; where V represents volume;

[0105] Figure 2 This is a schematic diagram of the cumulative volume distribution curve of a block.

[0106] S22. Calculate the area difference between the two cumulative block volume distribution curves in the same interval, and normalize the area difference as the size distribution difference index. Calculate the size consistency index of the first block system and the second block system.

[0107] S23. Construct a shape similarity matrix for the first block system and the second block system;

[0108] S24. Based on the shape similarity matrix, the first block system and the second block system are clustered and grouped using the spectral clustering method respectively;

[0109] S25. Based on the clustering results, calculate the proportion of the sum of the block volumes of each size category to the total volume of the corresponding shape group, and arrange each size category in descending order of volume contribution proportion. Determine the dominant size category of each block group according to the preset cumulative volume contribution threshold.

[0110] S26. Calculate the average shape similarity between all blocks in the dominant size category of each block group and all other blocks in the block group, and select the block with the largest average shape similarity as the representative shape of the group.

[0111] S27. Calculate the proportion of the number of blocks in each block group in the first block system and the second block system respectively, and use the proportion of each shape group as the weight of the corresponding representative shape to obtain the discrete shape distribution of the first block system and the second block system respectively.

[0112] S28. Calculate the shape similarity between the representative shapes in the first block system and the representative shapes in the second block system, and calculate the corresponding unit transportation cost. Obtain the unit transportation cost between all representative shapes in the first block system and the second block system, and construct a transportation cost matrix.

[0113] S29. Establish an optimal transmission model with the goal of minimizing the total transportation cost between the discrete shape distributions of the first block system and the second block system, solve the optimal transmission model to obtain the minimum total transportation cost, and calculate the shape consistency index of the first block system and the second block system.

[0114] S210, Construct a complex mesh for the first block system and the second block system;

[0115] S211. Extract the local connectivity, global connectivity, and structural partitioning features of complex meshes and use them as topology evaluation indicators.

[0116] S212. Calculate the relative symmetry error of the first block system and the second block system on each topological index respectively, and use it as the consistency index of different topological features between the two, and finally obtain the overall topological consistency evaluation result.

[0117] S213. Calculate the geometric consistency index by weighting the size consistency index and the shape consistency index with the same weight.

[0118] S214. Determine the geometric consistency weight and topological consistency weight, and calculate the comprehensive consistency index of the first block system and the second block system;

[0119] S215. Determine if there are still uncalculated block system pairs to be evaluated: If they exist, extract the next set of block system pairs and repeat S21-S214; otherwise, proceed to S3.

[0120] The calculation process for the dimensional consistency index between the first and second block systems is as follows: Figure 3 As shown, the area difference between the two curves corresponds to the shaded area shown in the figure.

[0121] Figure 4 A schematic diagram of the shape similarity matrix construction for a block system is shown. The maximum volume overlap rate between blocks i and j within the system is their shape similarity index, i.e., SSI. ij The block system contains m blocks, which ultimately form an m×m shape similarity matrix.

[0122] Set a minimum average shape similarity threshold t m Minimum group size threshold t s And the initial number of cluster groups k. Based on the shape similarity matrix, the blocks in the first block system and the second block system are clustered and grouped respectively, such as Figure 5 The diagram shows a clustering and grouping process for a block system. The blocks in this system are divided into two main categories based on shape similarity: grouped blocks and random blocks. The grouped blocks are further divided into three groups.

[0123] Based on the preset 3D block size classification standard and the standards in Table 1, the blocks within each shape group are classified by size. The proportion of the sum of the block volumes of each size category to the total volume of the blocks in the corresponding shape group is calculated, and the size categories are arranged in descending order of volume contribution proportion. If the volume contribution proportion of the first-ranked size category is greater than the preset cumulative volume contribution threshold t... v If the size category is determined to be the dominant size category, then the volume contribution ratio is not greater than t. v Then, according to the sorting results, the selected size categories are added sequentially until the cumulative volume contribution ratio of the selected size category is greater than t. v The selected size categories are then collectively identified as the dominant size categories. Figure 6 The left side shows Figure 5 Calculation of the dominant size category for block group 3. Based on size, this group of blocks is divided into small blocks, medium blocks, large blocks, and very large blocks. The cumulative volume and cumulative volume percentage of each type of block are calculated, and a cumulative size distribution map is created. The cumulative volume contribution threshold t is then applied. v The percentage was set at 68%. However, since the cumulative volume of large blocks did not exceed 68%, but the cumulative volume of large and very large blocks exceeded 68%, these two size categories were designated as the dominant size categories.

[0124] Table 1

[0125]

[0126] Calculate the average shape similarity between all blocks in the dominant size category of each block group and all other blocks within the same block group. Select the block with the highest average shape similarity and define its shape as the representative shape of that group. Figure 6 As shown.

[0127] The proportion of blocks in each block group in the two block systems is counted separately. The proportion of each shape group is used as the weight of the corresponding representative shape to obtain the discrete shape distribution of the block system. Figure 7 A schematic diagram of the construction of the optimal transport model between shape distributions. Figure 7 The left and right sides are schematic diagrams of the shape distribution of the first block system and the second block system, respectively: the first block system is clustered into 3 groups of blocks, and the second block system is clustered into 5 groups of blocks.

[0128] Each block in the block system is abstracted as a network node; it is determined whether any two blocks share a surface. If the two blocks share at least one surface, an undirected edge is established between the corresponding nodes, and the number of shared surfaces between the two blocks is determined as the weight of the undirected edge; if the two blocks do not share a surface, no network edge is established between the corresponding nodes. For example... Figure 8 As shown, a simple block system consisting of four blocks is used to illustrate the process of building a network. These four blocks are represented by different colors and mapped to nodes of the corresponding colors in the network representation. If two blocks share a face, an edge is created between their corresponding nodes in the network, as shown by the blocks in the figure. In the example shown, all edges have a weight of 1. It is important to emphasize that the resulting complex network is considered an undirected network, meaning that the edges in the network have no directionality.

[0129] In this embodiment of the invention, in S22, the area difference between the two cumulative block volume distribution curves within the interval [a,b] is... for:

[0130] ;

[0131] Among them, F A (x) represents the cumulative block volume distribution curve of the first block system, F B (x) represents the cumulative volume distribution curve of the second block system, a represents the lower limit of the integration interval, and b represents the upper limit of the integration interval;

[0132] The lower limit 'a' of the integration interval is determined by the logarithm of the smallest block volume in the two block systems to the base 10, and the upper limit 'b' of the integration interval is determined by the logarithm of the largest block volume in the two block systems to the base 10.

[0133] In S22, the area difference normalization is specifically implemented as follows:

[0134] ;

[0135] in, This represents an index indicating the difference in size distribution after normalization.

[0136] In S22, the dimensional consistency index for:

[0137] .

[0138] In this embodiment of the invention, in S23, any two blocks in the block system are scaled to the same volume, and the maximum volume overlap rate between the two normalized blocks is calculated as the shape similarity between the two blocks; all shape similarity indices are used as matrix elements to construct a shape similarity matrix;

[0139] The method for generating the maximum volume overlap rate between two normalized blocks is as follows: the point cloud registration method is used to spatially align the point clouds of the two normalized blocks so that the volume overlap of the two normalized blocks reaches the maximum. The proportion of the number of points in the point cloud of one block falling into the other block after point cloud registration is calculated as the proportion of the total number of points in the point cloud. This proportion is used as the maximum volume overlap rate between the two normalized blocks.

[0140] In this embodiment of the invention, S24 includes the following sub-steps:

[0141] S241. Set the initial number of cluster groups to 2, and use spectral clustering to group the shape similarity matrix of the block system.

[0142] S242. In the clustered block groups, the groups with fewer than the minimum group size threshold are designated as random block groups, and the remaining groups are designated as grouped block groups.

[0143] S243. Calculate the average shape similarity among all blocks within each group of blocks. If the average shape similarity of all groups of blocks is greater than or equal to the minimum average shape similarity threshold, proceed to S25; otherwise, proceed to S244.

[0144] S244. Increment the initial number of clusters by 1, and repeat S241-S243 until the clustering results meet the preset conditions.

[0145] In this embodiment of the invention, in S25, determining the dominant size category of each block group specifically involves: arranging each size category in descending order of volume contribution ratio; if the volume contribution ratio of the first-ranked size category is greater than a preset cumulative volume contribution threshold, then that size category is determined as the dominant size category; otherwise, subsequent size categories are added sequentially according to the ranking result until the cumulative volume contribution ratio of the selected size category is greater than the preset cumulative volume contribution threshold, and the selected size categories are collectively determined as the dominant size categories.

[0146] In S27, the discrete shape distribution P of the first block system A The discrete shape distributions of the second block system are respectively P B for:

[0147] ;

[0148] Where, n A n represents the number of block groups in the first block system. B Indicates the number of block groups in the second block system; w i v represents the weight of the proportion of blocks in the first block system. j δ represents the weight of the proportion of the block group corresponding to the second block system. Ai δ represents the representative shape of the block group corresponding to the first block system. Bj This represents the representative shape of the block group corresponding to the second block system;

[0149] In S28, the unit transportation cost C ij for:

[0150] ;

[0151] Among them, S ij This represents the shape similarity between the i-th representative shape in the first block system and the j-th representative shape in the second block system;

[0152] In S29, the optimal transmission model is:

[0153] ;

[0154] Among them, T ij The weight for transporting the i-th shape group of the first block system to the j-th shape group of the second block system;

[0155] The constraints of the optimal transmission model are:

[0156] ;

[0157] In S29, the shape consistency index of the first block system and the second block system for:

[0158] ;

[0159] In S210, the construction of a complex mesh is specifically as follows: the blocks of the first block system and the second block system are converted into network nodes respectively; it is determined whether any two blocks have a shared surface. If the two blocks have at least one shared surface, an undirected edge is established between the corresponding nodes, and the number of shared surfaces between the two blocks is determined as the weight of the undirected edge; if the two blocks do not have a shared surface, no network edge is established between the corresponding nodes.

[0160] In S212, the symmetric relative error E between the k-th topological index of the first block system and the second block system k for:

[0161] ;

[0162] Where, x Ak Let x represent the k-th topological index of the first block system. Bk This represents the k-th topological index of the second block system.

[0163] In S212, the k-th topological consistency index C k for:

[0164] ;

[0165] In S212, the comprehensive topology consistency index C topo for:

[0166] ;

[0167] Where C1 represents the first topology consistency index, C2 represents the second topology consistency index, and C3 represents the third topology consistency index;

[0168] In S213, the geometric consistency index C geo for:

[0169] .

[0170] In S214, the geometric consistency weight w is taken. g The topology consistency weight is set to 0.5, resulting in a comprehensive consistency index C. system for:

[0171] ;

[0172] In this embodiment of the invention, in step S3, the average value and coefficient of variation of the overall consistency index for all block systems are calculated;

[0173] Based on the average value and coefficient of variation of the comprehensive consistency index, the structural consistency of several block systems is judged. When the average value of the comprehensive consistency index is greater than or equal to the set average value threshold and the coefficient of variation is less than or equal to the set coefficient of variation threshold, the reliability of the generated result of the block system is deemed qualified; otherwise, the reliability of the block system is deemed unqualified.

[0174] In this experimental example, a virtual rock mass model was constructed using a cubic region with a side length of 10m. Three sets of structural surfaces were set up using a disk model, and the distribution and corresponding statistical values ​​of their characterization parameters are shown in Table 2.

[0175] Table 2

[0176]

[0177] Considering that structural surfaces located outside the computational region may still intersect with the boundary or internal structural surfaces of the computational region and participate in the formation of blocks, in order to reduce the truncation effect of structural surfaces, the generation range of structural surface centers is expanded from (0,0,0)-(10,10,10) to (-3,-3,-3)-(13,13,13) with an average structural surface radius of 3m as the expansion distance, and structural surface centers are generated uniformly within this region.

[0178] Based on the aforementioned statistical parameters, 30 statistically equivalent but geometrically different DFN models were generated using the Monte Carlo method. A general-purpose block recognition program developed based on MATLAB was then used to perform 3D block recognition on each model, resulting in 30 block systems. The obtained block system data includes the vertex, edge, face, and volume data of each block, as well as the topological relationships between the geometric elements. Figure 9 Two representative block systems from the controlled virtual rock mass example are shown. In the figure, each finite block is displayed in a different color for easy distinction; the double-headed arrows indicate that these two block systems will be compared as a pair of block systems in the subsequent consistency evaluation.

[0179] The size, shape, and topological attributes of each block system are further calculated. Size attributes are represented by the cumulative block volume distribution curve; shape attributes are represented by shape similarity calculation, spectral clustering, and representative shape extraction; blocks are abstracted as network nodes, and when two blocks share at least one face, an undirected edge is established between the corresponding nodes, with the number of shared faces used as the edge weight, thus constructing a weighted complex network for the block system. The average weighted degree, network diameter, and number of communities are further selected to represent the local connectivity, global connectivity, and structural partitioning characteristics of the complex network, respectively.

[0180] The 30 block systems can form 435 unordered block system pairs. To balance statistical representativeness and computational efficiency, 60 unique block system pairs are randomly selected for consistency calculation. For each block system pair, the dimensional consistency index C is calculated. size And shape consistency index C shape Box plots of these geometric property consistency indices are as follows: Figure 10 As shown. Further calculations were performed on the consistency indices of various topological attributes for each block system pair, and the corresponding box plots are shown below. Figure 11 As shown. Finally, the geometric consistency index C is calculated. geo Topology consistency index C topo The overall consistency index C is calculated using the same weights for geometric topology. system The corresponding box plot is as follows Figure 12 As shown in Table 3, the average values ​​and coefficients of variation of the main consistency indicators for these 60 block system pairs are also shown.

[0181] As shown in Table 3, the average value of the size consistency index is 90.5%, indicating that the block size distribution of different DFN models is generally consistent. The average value of the shape consistency index is 69.5%, indicating that the differences in the distribution of representative shapes are relatively large. The average values ​​of the consistency indices of average weighted degree, network diameter, and number of communities all exceed 90%, indicating that the local connectivity, global connectivity, and structural partitioning characteristics of different block systems are highly consistent.

[0182] Table 3

[0183]

[0184] The average topological consistency index was 95.0%, higher than the geometric consistency index of 80.0%, indicating that the topological structure is less affected by the randomness of the DFN compared to the geometric properties of the block system. The average comprehensive consistency index was 87.5%, with a coefficient of variation of 2.4%, indicating that the block system generated under the same structural surface statistical parameters has high and stable geometric-topological consistency. Therefore, the DFN-generated block system in this experimental example has high reliability.

[0185] In summary, the reliability evaluation method for block systems based on geometric-topological consistency described in this invention can quantitatively analyze the structural consistency between block systems generated by statistically equivalent DFN models from aspects such as block size, block shape, and system topology. The overall consistency level and stability of the block system are reflected by the average value and coefficient of variation of the comprehensive consistency index. Experimental results show that the average value of the comprehensive consistency index in this experimental example is 87.5%, and the coefficient of variation is 2.4%, indicating that the block systems generated under the same structural surface statistical parameters have high and stable geometric-topological consistency. Therefore, this invention can effectively verify the reliability of DFN-generated block systems, providing a quantitative basis for determining whether the generated block systems can serve as reasonable structural models for subsequent numerical simulations of discontinuous media mechanics. The method is effective and feasible.

[0186] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A reliability evaluation method for block systems based on geometric-topological consistency, characterized in that, Includes the following steps: S1. Based on the statistical parameters of the structural surfaces of the target rock mass, construct several block systems and combine all block systems in pairs to obtain several block system pairs; S2. Calculate the geometric, topological, and overall consistency indices for all block system pairs in sequence. S3. Conduct reliability analysis based on the comprehensive consistency index of the block system.

2. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 1, characterized in that, S1 includes the following sub-steps: S11. Based on the statistical parameters of the structural surfaces of the target rock mass, several discrete fracture network models are generated using the Monte Carlo method. S12. The three-dimensional block recognition method is used to identify the blocks of each discrete fracture network model to obtain the corresponding block system data. S13. Combine all the block systems in pairs to form several block system pairs.

3. The method for reliability evaluation of block systems based on geometric-topological consistency according to claim 2, characterized in that, In S12, the block system data includes the vertex, edge, face, and volume data of the block system, as well as the topological relationships between the geometric elements.

4. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 1, characterized in that, S2 includes the following sub-steps: S21. Select a block system pair as the first block system and the second block system; calculate the volume of all blocks in each of the two block systems, and arrange the volumes in ascending order; for each block, calculate the cumulative volume of all blocks whose volume is not greater than that of the block, and use its proportion to the total volume of the system as the ordinate, and plot the common logarithm of the block volume log. 10 (V) is the abscissa, and cumulative block volume distribution curves are constructed respectively; where V represents volume; S22. Calculate the area difference between the two cumulative block volume distribution curves in the same interval, and normalize the area difference as the size distribution difference index. Calculate the size consistency index of the first block system and the second block system. S23. Construct a shape similarity matrix for the first block system and the second block system; S24. Based on the shape similarity matrix, the spectral clustering method is used to cluster and group the first block system and the second block system respectively; S25. Based on the clustering results, calculate the proportion of the sum of the block volumes of each size category to the total volume of the corresponding shape group, and arrange each size category in descending order of volume contribution proportion. Determine the dominant size category of each block group according to the preset cumulative volume contribution threshold. S26. Calculate the average shape similarity between all blocks in the dominant size category of each block group and all other blocks in the block group, and select the block with the largest average shape similarity as the representative shape of the group. S27. Calculate the proportion of the number of blocks in each block group in the first block system and the second block system respectively, and use the proportion of each shape group as the weight of the corresponding representative shape to obtain the discrete shape distribution of the first block system and the second block system respectively. S28. Calculate the shape similarity between the representative shapes in the first block system and the representative shapes in the second block system, and calculate the corresponding unit transportation cost. Obtain the unit transportation cost between all representative shapes in the first block system and the second block system, and construct a transportation cost matrix. S29. Establish an optimal transmission model with the goal of minimizing the total transportation cost between the discrete shape distributions of the first block system and the second block system, solve the optimal transmission model to obtain the minimum total transportation cost, and calculate the shape consistency index of the first block system and the second block system. S210, Construct a complex mesh for the first block system and the second block system; S211. Extract the local connectivity, global connectivity, and structural partitioning features of complex meshes and use them as topology evaluation indicators; S212. Calculate the relative symmetry error of the first block system and the second block system on each topological index respectively, and use it as the consistency index of different topological features between the two, and finally obtain the overall topological consistency evaluation result. S213. Calculate the geometric consistency index by weighting the size consistency index and the shape consistency index with the same weight. S214. Determine the geometric consistency weight and topological consistency weight, and calculate the comprehensive consistency index of the first block system and the second block system; S215. Determine if there are still uncalculated block system pairs to be evaluated: If they exist, extract the next set of block system pairs and repeat S21-S214; otherwise, proceed to S3.

5. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 4, characterized in that, In S22, the area difference between the two cumulative block volume distribution curves in the interval [a,b] for: ; Among them, F A (x) represents the cumulative block volume distribution curve of the first block system, F B (x) represents the cumulative volume distribution curve of the second block system, a represents the lower limit of the integration interval, and b represents the upper limit of the integration interval; In step S22, normalizing the area difference specifically involves: ; in, Indicator representing the difference in size distribution after normalization; In S22, the size consistency index for: 。 6. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 4, characterized in that, In step S23, any two blocks in the block system are scaled to the same volume, and the maximum volume overlap rate between the two normalized blocks is calculated as the shape similarity between the two blocks. Construct a shape similarity matrix by using all shape similarity indices as matrix elements; The method for generating the maximum volume overlap rate between the two normalized blocks is as follows: the point cloud registration method is used to spatially align the point clouds of the two normalized blocks so that the volume overlap of the two normalized blocks reaches the maximum. The proportion of the number of points in the point cloud of one block falling into the other block after point cloud registration is calculated as the proportion of the total number of points in the point cloud. This proportion is used as the maximum volume overlap rate between the two normalized blocks.

7. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 4, characterized in that, S24 includes the following sub-steps: S241. Set the initial number of cluster groups to 2, and use spectral clustering to group the shape similarity matrix of the block system. S242. In the clustered block groups, the groups with fewer than the minimum group size threshold are designated as random block groups, and the remaining groups are designated as grouped block groups. S243. Calculate the average shape similarity among all blocks within each group of blocks. If the average shape similarity of all groups of blocks is greater than or equal to the minimum average shape similarity threshold, proceed to S25; otherwise, proceed to S244. S244. Increment the initial number of clusters by 1, and repeat S241-S243 until the clustering results meet the preset conditions.

8. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 4, characterized in that, In step S25, determining the dominant size category for each block group specifically involves: arranging each size category in descending order of volume contribution ratio; if the volume contribution ratio of the first-ranked size category is greater than a preset cumulative volume contribution threshold, then that size category is determined as the dominant size category; otherwise, subsequent size categories are added sequentially according to the ranking result until the cumulative volume contribution ratio of the selected size category is greater than the preset cumulative volume contribution threshold, and the selected size categories are collectively determined as the dominant size categories. In S27, the discrete shape distribution P of the first block system A The discrete shape distributions of the second block system are respectively P B for: ; Where, n A n represents the number of block groups in the first block system. B Indicates the number of block groups in the second block system; w i v represents the weight of the proportion of blocks in the first block system. j δ represents the weight of the proportion of the block group corresponding to the second block system. Ai δ represents the representative shape of the block group corresponding to the first block system. Bj This represents the representative shape of the block group corresponding to the second block system; In S28, the unit transportation cost C ij for: ; Among them, S ij This represents the shape similarity between the i-th representative shape in the first block system and the j-th representative shape in the second block system; In S29, the optimal transmission model is: ; Among them, T ij The weight for transporting the i-th shape group of the first block system to the j-th shape group of the second block system; The constraints of the optimal transmission model are: ; In S29, the shape consistency index of the first block system and the second block system for: ; In S210, constructing a complex mesh specifically involves: converting the blocks of the first block system and the second block system into network nodes respectively; determining whether any two blocks have a shared surface; if the two blocks have at least one shared surface, then establishing an undirected edge between the corresponding nodes, and determining the number of shared surfaces between the two blocks as the weight of the undirected edge; if the two blocks do not have a shared surface, then not establishing a network edge between the corresponding nodes. In S212, the symmetric relative error E between the k-th topological index of the first block system and the second block system k for: ; Where, x Ak Let x represent the k-th topological index of the first block system. Bk This represents the k-th topological index of the second block system; In S212, the k-th topology consistency index C k for: ; In S212, the comprehensive topology consistency index C topo for: ; Where C1 represents the first topology consistency index, C2 represents the second topology consistency index, and C3 represents the third topology consistency index; In S213, the geometric consistency index C geo for: ; In S214, the comprehensive consistency index C system for: ; Among them, w g This represents the geometric consistency weight.

9. The reliability evaluation method for block systems based on geometric-topological consistency according to claim 1, characterized in that, In step S3, the average value and coefficient of variation of the overall consistency index for all block systems are calculated. The structural consistency of several block systems is judged based on the average value and coefficient of variation of the comprehensive consistency index. When the average value of the comprehensive consistency index is greater than or equal to the set average value threshold and the coefficient of variation is less than or equal to the set coefficient of variation threshold, the reliability of the generated block system is deemed to be qualified. Otherwise, the reliability of the block system is deemed unqualified.