A small block processing method for optimizing two-dimensional block recognition results

Abstract

The present invention relates to the field of two-dimensional rock block identification, and in particular to a small block processing method for optimizing two-dimensional block identification results. Based on the block geometry data obtained by two-dimensional block identification, all block areas are calculated, the blocks therein are searched and sorted, and then the small blocks are processed by setting a small block threshold and searching for small blocks, searching adjacent blocks, constructing a segmentation line of adjacent blocks, constructing a representative vertex domain, determining a representative vertex, processing small blocks, updating the vertex element data of the block system, updating the edge element data of the block system, and updating the surface element data of the block system. Compared with the prior art, the calculations of this application are all theoretical calculations and only use the small block area threshold as an input parameter. The method of the present invention is objective and simple to operate and easy to program. The method disclosed in this application is a block data optimization algorithm of a traditional two-dimensional block recognition algorithm. Based on the existing block data, all geometric elements related to small blocks in the block system are eliminated. While maintaining the consistency of the block shape in the block system as much as possible, all small block problems are solved, and the efficiency and robustness of subsequent numerical simulation calculations are improved.

Description

Technical Field

[0001] The invention relates to the field of two-dimensional rock block recognition, and in particular to a small block processing method for optimizing two-dimensional block recognition results. Background Art

[0002] Rock mass plays a vital role in human engineering construction. Accurately predicting its deformation and failure behavior constitutes the core concern of rock mechanics and engineering research. Rock mass has complex structural properties of heterogeneity, discontinuity and anisotropy. Discrete numerical methods show stronger ability in considering the discontinuity of jointed rock mass behavior, such as discrete element method (DEM) and discontinuous deformation analysis (DDA). These methods regard the rock mass as a block system composed of a large number of fractures cutting the calculation area. The process of converting a fracture system into a block system is often called block identification.

[0003] Rock mass can generally be regarded as a three-dimensional block system, but in order to improve the efficiency of numerical calculation, rock mass is often simplified into a two-dimensional object for analysis. At this time, two-dimensional block identification is a key pre-processing step for two-dimensional discrete element analysis. Researchers often optimize the two-dimensional block identification algorithm in terms of improving the computational efficiency and robustness of the algorithm itself, but ignore the impact of the identified block system expression on subsequent mechanical analysis. In practical applications, a two-dimensional block system is composed of many polygonal blocks with complex geometric shapes, including many small-sized blocks, and the area of ​​these blocks is generally more than 10 orders of magnitude different from the area of ​​the calculation area.

[0004] Although it is feasible to accurately identify these small blocks, DEM and DDA may encounter limitations in dealing with these small blocks, resulting in the following three major problems: (1) Setting of small time steps: In DEM or DDA, the time step must take into account the size of the smallest block to avoid missing the contact detection of the blocks. The size of the time step is positively correlated with the size of the smallest block, so smaller blocks may significantly reduce the computational efficiency of block contact detection. (2) Complex contact detection: The existence of small blocks leads to the complexity of the shapes of their adjacent blocks, increasing the difficulty of contact detection. (3) Generation of poor meshes: Small blocks share short edges with their adjacent blocks, which may lead to the generation of poor quality meshes inside these adjacent blocks. Poor quality meshes will significantly affect the efficiency and accuracy of numerical simulations.

[0005] There are currently two methods for processing small blocks: the deletion method and the pasting method. The deletion method directly deletes blocks in the block system that are smaller than the size threshold, while the pasting method pastes the small block with its adjacent blocks to form a new block to delete the small block. Although these two methods eliminate small blocks, the geometric elements related to the small blocks are not completely deleted. These geometric elements are often very short line segments, and their existence often means that the shapes of the adjacent blocks of the small blocks are complex, which not only increases the complexity of contact detection, but also produces extremely poor quality meshes during meshing, thereby affecting the effect of numerical simulation. Therefore, the existing small block processing methods cannot solve the three small block problems mentioned above at the same time.

[0006] In view of this, it is urgently needed in this field to propose a new small block processing method that optimizes the two-dimensional block recognition results and solves all small block problems at the same time. Summary of the invention

[0007] Based on the above purpose, the present application provides a small block processing method for optimizing two-dimensional block recognition results, comprising the following steps:

[0008] (1) Set the small block threshold and search for the smallest block in the block system (block i, i ≥ 1), with an area of ​​A s :If area A s If the area A is larger than the set small block threshold, no small block processing will be performed. s If it is smaller than the set small block threshold, proceed to step (2);

[0009] (2) Search for the adjacent block j (j ≥ 1) that shares vertices with the small block i;

[0010] (3) Construct a dividing line for adjacent block j, and determine whether all non-shared vertices of adjacent block j are distributed on the same side of the dividing line: the dividing line is formed by connecting two non-shared vertices D1 and D2 on adjacent block j, and the adjacent vertices of D1 and D2 on adjacent block j must contain shared vertices between small block i and adjacent block j. Construct a dividing line equation, and substitute the vertex coordinates of all non-shared vertices of adjacent block j into the dividing line equation, and determine the calculation result: if the calculation results are ≥0 or ≤0, it means that all non-shared vertices of adjacent block j are on the same side of the dividing line, and go to step (4); otherwise, go to step (6) and directly use the pasting method to process the small block;

[0011] (4) Constructing a representative vertex domain: Substitute a non-shared vertex of the adjacent block j into the dividing line equation to construct the representative vertex domain of the adjacent block j;

[0012] (5) Determine whether there is a representative vertex: Repeat steps (3)-(4) until all adjacent blocks j of small block i have constructed representative vertex domains, and determine whether these representative vertex domains have an intersection. If there is an intersection, it means that a representative vertex exists, and the representative vertex is determined. If there is no intersection, it means that there is no representative vertex.

[0013] (6) Processing small blocks: When a representative vertex exists, delete small block i and replace all vertices shared with small block i and edges composed entirely of shared vertices on adjacent block j with the representative vertex; if there is no representative vertex, directly use the pasting method to process the small block;

[0014] (7) Update the vertex element data of the block system: add representative vertex data and delete all vertex data of the original small block;

[0015] (8) Update edge element data of block system;

[0016] (9) updating the surface element data of the block system to obtain an updated block system;

[0017] (10) End the processing of small block i of the original block system;

[0018] (11) Extract the smallest block (block i) in the updated block system. If the area of ​​the block is less than the threshold, extract the information of small block i and continue steps (2)-(10). If the area of ​​the block is greater than or equal to the threshold, the small block processing is completed.

[0019] Set the small block threshold and search for small block i (i≥1);

[0020] Furthermore, the dividing line equations in steps (3) and (4) are:

[0021]

[0022] Where x1, y1, x2, y2 are the coordinates of two non-shared vertices, such as D1(x1, y1), D2(x2, y2), and a, b, c are the coefficients of the straight line expression.

[0023] Furthermore, the method for constructing the representative vertex domain in step (4) is: using an inequality to represent the semi-infinite plane area where the dividing line points to the non-shared vertex. Changing the inequality sign of the inequality equation, the new inequality represents the semi-infinite plane area pointing in another direction, and this area is called the representative vertex domain of block j.

[0024] Furthermore, the method for determining the representative vertex in step (5) is as follows: if there is an intersection in the representative vertex domain, and the intersection area includes the entire part of the small block i, then the center point of the small block is selected as the representative vertex; if the intersection area intersects part of the small block area, then any vertex of the small block in the intersection area is selected as the representative vertex.

[0025] Furthermore, in the edge element data of the updated block system in step (8), the three types of edges existing in the block system and the corresponding update methods are: (a) an edge composed of two shared vertices, in which case the edge does not exist in the new block system; (b) an edge composed of one shared vertex and one non-shared vertex, in which case the edge is updated in the new block system; (c) an edge composed of two non-shared vertices, in which case the edge does not need to be updated in the new block system.

[0026] Furthermore, the updating of the face element data of the block system in step (9) is as follows: judging all blocks whose geometric data have been changed: if the number of vertices of the new block is less than 3 or the number of directed edges is less than 3, the block should not be stored in the new block system.

[0027] The beneficial effects of the present application are as follows: 1. The present application proposes a new small block processing method for optimizing two-dimensional block recognition results. Compared with the prior art, the calculations of the present application are all theoretical calculations and only use the small block area threshold as an input parameter. The method of the present invention is objective and simple to operate and easy to program.

[0028] 2. The method disclosed in this application is a block data optimization algorithm for the traditional two-dimensional block recognition algorithm. Based on the existing block data, it eliminates all geometric elements related to small blocks in the block system, solves all small block problems while maintaining the consistency of block shapes in the block system as much as possible, and improves the efficiency and robustness of subsequent numerical simulation calculations. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Flow chart of the two-dimensional small block fusion method

[0030] Figure 2 Schematic diagram of a two-dimensional trace diagram

[0031] Figure 3 This is a schematic diagram of the two-dimensional block recognition results. The arc-shaped arrows in the figure indicate polygonal blocks.

[0032] Figure 4 Schematic diagram of constructing split lines for adjacent blocks

[0033] Figure 5 The intersection of the representative vertex domains of all adjacent blocks completely contains the small block

[0034] Figure 6The intersection of the representative vertex domains of all adjacent blocks completely contains the small block

[0035] Figure 7 Schematic diagram of the shared geometric elements of adjacent blocks being replaced by representative vertices

[0036] Figure 8 Schematic diagram of processing small blocks by pasting method

[0037] Fig. 9 Schematic diagram of UDEC slope model

[0038] Fig.10 Comparison diagram of model monitoring point displacement

[0039] Fig.11 Comparison chart of model cycle times and calculation time DETAILED DESCRIPTION

[0040] The present invention is further described in detail below in conjunction with the accompanying drawings. The following specific implementation steps can enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any form.

[0041] First of all, it should be noted that the use of computer simulation to generate two-dimensional trace diagrams and two-dimensional block recognition belongs to the prior art. The technical solution of the present invention is based on the small block processing of blocks obtained by two-dimensional block recognition, which will involve a large number of prior arts, and its definition or concept is already well known or mastered in the art. Therefore, unless there is a special explanation in the present invention, other meanings that are consistent with the existing known meanings will not be described one by one in the present invention.

[0042] It should be noted that this application proposes a new small block processing method for optimizing the two-dimensional block recognition results, which is optimized on the basis of the existing technology, so this application calls it a fusion method, and its flow chart is shown in Figure 1 shown.

[0043] Example 1 A small block processing method for optimizing two-dimensional block recognition results

[0044] (1) Set the small block threshold and search for the smallest block in the block system (block i, i ≥ 1), with an area of ​​A s :

[0045] Based on the existing two-dimensional trace graph, the vertex, line segment, and polygon data of all blocks in the block system are obtained through the two-dimensional block recognition algorithm. Figure 2 The figure shows a two-dimensional trace diagram, where the outer rectangle is the calculation area and the line segment is the trace. Figure 3The figure shows a schematic diagram of the two-dimensional block recognition result. The dots in the figure are the vertices of the polygonal blocks, the edges indicated by the arrows are the directed edges of the polygonal blocks, and the arc-shaped arrows indicate directed polygonal blocks. Based on the geometric data of the blocks, the area of ​​each block is calculated. These blocks are sorted in ascending order of area value. The area threshold of the small blocks is set to determine whether the block with the smallest area is a small block. If the smallest block is not a small block, the small block processing is not implemented. If the smallest block is a small block, the smallest block (small block i, i≥1) information is extracted;

[0046] (2) Search for the adjacent block j (j ≥ 1) that shares vertices with the small block i

[0047] Search for blocks that share vertices with small block i. These blocks are called neighboring blocks, and the vertices shared by neighboring blocks and small block i are called shared vertices.

[0048] (3) Construct the dividing line of adjacent blocks j and determine whether all non-shared vertices of adjacent blocks j are distributed on the same side of the dividing line

[0049] For a neighboring block j of a small block i, extract a line segment consisting of a shared vertex and a non-shared vertex. The neighboring block has only two line segments that meet this condition. Connect the two non-shared vertices of these two line segments to obtain the segmentation line of the neighboring block j. Figure 4 The figure shows a schematic diagram of constructing a dividing line. The small triangle above is a small block, and the large polygon below is the adjacent block of the small block. They share vertices s1 and s2. Line segments s1G and s2A are line segments on adjacent blocks consisting of a shared vertex and a non-shared vertex, so connect vertex G and vertex A to construct a dividing line. If the coordinates of these two vertices are: A(x1,y1), G(x2,y2), then the dividing line expression is:

[0050]

[0051] Where: a, b, c are the coefficients of the straight line expression.

[0052] All non-shared vertices of block j (such as Figure 4 Substitute the coordinates of the vertices A, B, C, D, E, F, G) in into the dividing line equation. If all the results are greater than or equal to zero or less than or equal to zero, it means that all non-shared vertices of adjacent blocks j are on the same side of the dividing line, then go to step (4). Otherwise, go directly to step (6) and use the pasting method to process the small blocks directly.

[0053] (4) Constructing a representative vertex domain

[0054] Substitute a non-shared vertex of block j into the splitter equation, and use an inequality to represent the semi-infinite plane area where the splitter points to the non-shared vertex. Change the inequality sign of the inequality equation, and the new inequality represents the semi-infinite plane area pointing in the other direction, which is called the representative vertex domain of block j.

[0055] (5) Determine whether there is a representative vertex:

[0056] Execute steps (3)-(4) for all adjacent blocks of small block i. After constructing the representative vertex domains of all adjacent blocks, determine whether these semi-infinite planes have an intersection. If there is no intersection, it means that there is no representative vertex, and go to step (6); when there is an intersection, it means that there is a representative vertex, and then select a suitable representative vertex in the intersection area: if the intersection area includes the entire part of the small block, select the center point of the small block as the representative vertex; if the intersection area partially intersects with the small block area, select any vertex of the small block in the intersection area as the representative vertex; if the intersection area does not intersect with the small block area, then go directly to step (6) and use the pasting method to process the small block directly. Figure 5 As shown in the figure, there are 7 blocks in a block system: small block 1 and its 6 adjacent blocks. The dotted lines and the arrows in the figure indicate the representative vertex domains indicated by the corresponding dividing lines. The representative vertex domain set area indicated by the six dividing lines includes small block 1, so the center point of small block 1 is selected as the representative vertex. Figure 6 As shown in the figure, there are 8 blocks in a block system: small block 1 and its 7 adjacent blocks. The dotted lines and the arrows in the figure indicate the representative vertex domains indicated by the corresponding dividing lines. The representative vertex domain set area indicated by the seven dividing lines partially includes small block 1, and the included part is shown by the black triangle. Then the vertex S2 of the small block is selected as the representative vertex.

[0057] (6) Processing small pieces

[0058] When a representative vertex exists, delete small block i and replace all vertices shared with small block i and edges composed entirely of shared vertices on adjacent blocks j with the representative vertex. Figure 5 For the original block system shown in the figure, we replace all shared vertices and shared edges of the adjacent blocks 2-7 with the center points of the small blocks to obtain new adjacent blocks 2'-7', as shown in Figure 7 If there is no representative vertex, the small block is processed by the pasting method: select any adjacent block with which the small block has a shared edge, and paste the small block onto the adjacent block. Figure 5 For the original block system shown in the figure, when we use the pasting method, we paste the small block onto the adjacent block 2 to obtain a new block 2', and the shapes of the other adjacent blocks remain unchanged, as shown in Figure 8 shown.

[0059] (7) Update the vertex element data of the block system

[0060] Add representative vertex data and delete all vertex data of the original small block.

[0061] (8) Update edge element data of block system

[0062] The three types of edges in the original block system and their corresponding update methods: (a) an edge consisting of two shared vertices, in which case the edge does not exist in the new block system; (b) an edge consisting of one shared vertex and one unshared vertex, in which case the edge is updated in the new block system; (c) an edge consisting of two unshared vertices, in which case the edge does not need to be updated in the new block system.

[0063] (9) Update the surface element data of the block system

[0064] All blocks whose geometric data have been changed are judged: if the number of vertices of the new block is less than 3 or the number of directed edges is less than 3, the block should not be stored in the new block system.

[0065] (10) End the processing of small block i of the original block system;

[0066] (11) Extract the smallest block (block i) in the updated block system. If the area of ​​the block is less than the threshold, extract the information of small block i and continue steps (2)-(10). If the area of ​​the block is greater than or equal to the threshold, the small block processing is completed.

[0067] Test Example 1

[0068] Use UDEC software to build the slope model, slope size, block distribution map and block area statistics map as shown in the figure. Fig. 9 As shown. The defined area is less than 0.003m 2 The block is a small block, and there is a small block in the block system of the slope model. 10 monitoring points are set to record the displacement of the blocks: 6 monitoring points are evenly arranged on the slope surface, namely monitoring points 1-6, and 4 monitoring points are arranged 5m below monitoring points 1-4, named monitoring points 7-10. The elastic isotropic model is used as the constitutive model for numerical analysis, and the physical and mechanical properties of rock blocks and joints are listed in Table 1 as input parameters. The left, right and bottom boundaries of the slope ( Fig. 9 The shaded border is fixed and does not allow displacement in the x and y directions.

[0069] The original model is processed by the method of Example 1 (fusion method), deletion method, and pasting method. The deletion method is to directly delete all the small block surface element data within the block system that is smaller than the size threshold. The pasting method is to paste the small block in the block system onto the adjacent block that shares the edge information with the small block, and adjust the block system geometric element data accordingly.

[0070] The model numbers and corresponding descriptions are shown in Table 2. Among them, the pasting method obtains three different new models by pasting small blocks to different adjacent blocks. The same parameters are input into the original model and the new model, the same boundary conditions are set, and only gravity is added as an external force. Numerical simulation is performed until the model is stable.

[0071] Compare the displacement of monitoring points of the new model with that of the original model in a stable state, such as Fig.10 As shown. According to the degree of consistency between the displacement of 10 monitoring points in different processing models and the original model, it can be seen that the consistency of the simulation results of the pasting method is affected by the selection of the adjacent blocks to be pasted, while the stable state of the model processed by the fusion method is highly consistent with the original model, indicating that the patented method can ensure the correctness of the numerical simulation calculation. The number of cycles and calculation time to reach the stable state of the new models (model 2, model 3, model 5, model 6) with good consistency with the original model and the original model (model 1) are compared, as shown in Fig.11 As shown. The fusion method proposed in this patent is the only method that can reduce the number of cycles and calculation time at the same time, reducing them by 57.1% and 38.4% respectively. The results show that the fusion method can not only maintain the correctness of the numerical simulation results, but also has the best effect in improving the calculation efficiency of the model, which can further enhance the convergence of the simulation. It can be seen that this method is effective and feasible.

[0072] Table 1. Input parameters of the UDEC slope model: physical and mechanical parameters of rock masses and cracks

[0073]

[0074] Table 2. Description and serial number of different models

[0075]

[0076] Note: The actual scope of the present invention includes not only the specific embodiments disclosed above, but also all equivalent solutions that implement or perform the present invention under the claims.

Claims

1. A small block processing method for optimizing two-dimensional rock block recognition results, characterized in that: The following steps are involved: (1) Set the small block threshold and search for the smallest block in the block system, that is, block i, i ≥ 1, with an area of A s :If the area A s If the area is larger than the set small block threshold, no small block processing will be performed. A s If it is smaller than the set small block threshold, proceed to step (2); (2) Search for the adjacent block j that shares vertices with the small block i, where j ≥ 1; (3) Construct a dividing line of adjacent block j, and determine whether all non-shared vertices of adjacent block j are distributed on the same side of the dividing line: the dividing line is formed by connecting two non-shared vertices D1 and D2 on adjacent block j, and the adjacent vertices of D1 and D2 on adjacent block j must contain shared vertices of small block i and adjacent block j; construct a dividing line equation, and substitute the vertex coordinates of all non-shared vertices of adjacent block j into the dividing line equation, and determine the calculation result: if the calculation results are all ≥0 or ≤0, it means that all non-shared vertices of adjacent block j are located on the same side of the dividing line, and go to step (4); otherwise, go to step (6) and directly use the pasting method to process the small block; (4) Constructing a representative vertex domain: Substitute a non-shared vertex of the adjacent block j into the dividing line equation to construct the representative vertex domain of the adjacent block j; (5) Determine whether there is a representative vertex: Repeat steps (3) to (4) until all adjacent blocks j of small block i have constructed representative vertex domains, and determine whether these representative vertex domains have an intersection. If there is an intersection, it means that there is a representative vertex, and the representative vertex is determined. If there is no intersection, it means that there is no representative vertex. (6) Processing small blocks: When a representative vertex exists, delete small block i and replace all vertices shared with small block i and edges composed entirely of shared vertices on adjacent block j with the representative vertex; if there is no representative vertex, directly use the pasting method to process the small block; (7) Update the vertex element data of the block system: add representative vertex data and delete all vertex data of the original small block; (8) Update edge element data of the block system; (9) Update the surface element data of the block system to obtain an updated block system; (10) End the processing of small block i of the original block system; (11) Extract the smallest block i in the updated block system. If the area of ​​the block is less than the threshold, extract the information of the small block i and continue steps (2)-(10). If the area of ​​the block is greater than or equal to the threshold, the processing of the small block is completed.

2. The processing method according to claim 1, characterized in that: The dividing line equations in steps (3) and (4) are: ； Where x1, y1, x2, y2 are the coordinates of two non-shared vertices, namely D1(x1, y1) and D2(x2, y2). a , b , c are the coefficients of the straight line expression.

3. The processing method according to claim 1, characterized in that: The method for constructing the representative vertex domain in step (4) is: using an inequality to represent the semi-infinite plane area where the dividing line points to the non-shared vertex; changing the inequality sign of the inequality equation, the new inequality represents the semi-infinite plane area pointing in another direction, and this area is called the representative vertex domain of block j.

4. The processing method according to claim 1, characterized in that: The method for determining the representative vertex in step (5) is as follows: if there is an intersection in the representative vertex domain, and the intersection area includes the entire part of the small block i, then the center point of the small block is selected as the representative vertex; if the intersection area intersects with part of the small block area, then any vertex of the small block in the intersection area is selected as the representative vertex.

5. The processing method according to claim 1, characterized in that: In the edge element data of the updated block system in step (8), the three types of edges existing in the block system and the corresponding update methods are: (a) an edge composed of two shared vertices, in which case the edge does not exist in the new block system; (b) an edge composed of one shared vertex and one non-shared vertex, in which case the edge is updated in the new block system; (c) an edge composed of two non-shared vertices, in which case the edge does not need to be updated in the new block system.

6. The processing method according to claim 1, characterized in that: Step (9) of updating the face element data of the block system is as follows: judging all blocks whose geometric data have been changed: if the number of vertices of the new block is less than 3 or the number of directed edges is less than 3, the block should not be stored in the new block system.

Images