Grid optimization method for tunnel DC resistivity inversion based on dual-guide indicator combination

Abstract

The present invention proposes a tunnel DC resistivity inversion grid optimization method based on a dual-guide indicator combination, comprising: preliminarily locking anomaly regions; performing regional optimization calculations on the preliminarily locked anomaly regions to determine the final anomaly regions; after determining the anomaly regions, locking the cells in the inner regions of each anomaly, then estimating the number of cells occupied by the outer boundaries of the anomaly based on the range of each anomaly region, and determining the selection ratio of the boundary cells in each anomaly region; calculating the gradient indicator factor of the inversion region model, sorting the cells in the inversion region, and locking the boundary regions of each anomaly according to the boundary selection ratio; after locking the anomaly and boundary regions, then sequentially performing single and multiple encryption on the grid in the region to optimize the inversion grid. The method can more accurately delineate the cells in the anomaly and its boundary region, and adaptively encrypt the grid in the region, thereby more efficiently optimizing the inversion grid and improving the inversion grid quality.

Description

Technical Field

[0001] The invention belongs to the technical field of tunnel DC resistivity advance detection, and in particular relates to a tunnel DC resistivity inversion grid optimization method based on a dual-guide indication combination. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The gridding method and density are not only closely related to the forward modeling results of tunnel DC resistivity, but also directly affect the accuracy and resolution of the inversion results. Due to the unknown geometry of underground structures and the uncertainty of the distribution of geological anomalies, it is difficult to establish a high-quality tunnel DC resistivity inversion grid in one go simply by relying on prior information obtained through geological surveys and drilling.

[0004] Global refinement can improve the fit between the model grid boundary and the boundary of the real anomaly area, but it is difficult to control the number of cells in the inversion area, which will increase the multi-solution of the inversion and thus affect the reliability of the inversion results.

[0005] Some scholars have proposed the use of different grid systems for forward and inverse calculations, where a fine grid is used for forward calculations and a coarse grid is used for inversion. Such a grid system improves the accuracy of forward modeling while limiting the degrees of freedom of inversion, and to a certain extent reasonably solves the above problems.

[0006] However, the use of a relatively sparse grid in the inversion area cannot accurately reflect the changes in electrical properties between strata, which will reduce the accuracy of the inversion. Summary of the Invention

[0007] To overcome the above-mentioned deficiencies of the prior art, the present invention provides a tunnel DC resistivity inversion grid optimization method based on a dual-guide indicator combination, which appropriately refines the inversion grid during the inversion process and improves the resolution of the inversion image.

[0008] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0009] In a first aspect, a tunnel DC resistivity inversion grid optimization method based on a dual-guide indicator combination is disclosed, comprising:

[0010] Calculate the model change of each unit in the inversion area according to the indicator factor of the model change, and preliminarily locate the abnormal body area;

[0011] Perform regional optimization calculation on the initially locked anomaly area to determine the final anomaly area, avoiding the presence of non-anomalous area units in each anomaly area;

[0012] After the abnormal body area is determined, the cells of the internal area of each abnormal body are locked, and then the number of cells occupied by the external boundary of the abnormal body is estimated according to the range of each abnormal body area, and the selection ratio of the boundary cells of each abnormal body area is determined accordingly;

[0013] Calculate the gradient indicator factor of the inversion area model, sort the units in the inversion area, and lock the boundary area of each anomaly according to the boundary selection ratio;

[0014] After locking the abnormal body and boundary area, the grid of the area is then encrypted once and multiple times in sequence to optimize the inversion grid.

[0015] As a further technical solution, the abnormal body area is initially locked down, specifically:

[0016] The unit numbers in the inversion area are sorted in descending order according to the model change amount, and about 8% of the units in the previous inversion area are selected to form the anomaly area, and the anomaly area is roughly determined.

[0017] As a further technical solution, it also includes:

[0018] Calculate the center coordinates of each unit in the Ω1 area respectively, and randomly select K coordinate points as the center of the anomaly. Calculate the minimum distance between the center point of each unit in the Ω1 area and the centers of the K anomalies, so that each unit in the Ω1 area is assigned to its nearest anomaly center.

[0019] As a further technical solution, the method further includes: calculating the distance d between the center point of each abnormal body region unit and the center point of the abnormal body. i , the d of each unit in each abnormal body area i Composition distance factor matrix;

[0020] Based on the distance factor matrix, regional optimization calculation of Ω1 is performed to avoid the presence of units in non-abnormal areas in each abnormal area.

[0021] As a further technical solution, when determining the final abnormal body area, the unit numbers of the Ω1 area are sorted in descending order according to the elements of the standard matrix C of each abnormal body area, and the Ω1 area units are selected as the final abnormal body area Ω2 according to a certain optimization ratio.

[0022] As a further technical solution, the method further includes: determining common units in the boundary area and the final abnormal body area, and removing these common units in the boundary area.

[0023] As a further technical solution, one-time encryption refers to dividing the hexahedron in the locked area into multiple tetrahedrons by adding nodes. The division method is divided into two categories depending on whether nodes are added:

[0024] (1) Divide the irregular hexahedral element into 6 tetrahedral elements without adding nodes;

[0025] (2) Add a node inside the irregular hexahedron unit to divide it into 12 tetrahedron units, that is, generate two new tetrahedron units for the node on each face of the irregular hexahedron and the newly added node, and adjust the position of the newly added node to control the unit quality of the generated tetrahedron unit.

[0026] In the second aspect, a tunnel DC resistivity inversion grid optimization system based on a dual-guide indicator combination is disclosed, comprising:

[0027] The module for initially locking the abnormal body region is configured to calculate the model change amount of each unit in the inversion region according to the indicator factor of the model change amount, and initially lock the abnormal body region;

[0028] The abnormal body area final determination module is configured to: perform regional optimization calculation on the initially locked abnormal body area to determine the final abnormal body area, and avoid the presence of non-abnormal body area units in each abnormal body area;

[0029] The anomaly and boundary region locking module is configured to: after determining the anomaly region, lock the cells of each anomaly region, then estimate the number of cells occupied by the outer boundary of the anomaly based on the range of each anomaly region, and use this to determine the selection ratio of the boundary cells of each anomaly region;

[0030] Calculate the gradient indicator factor of the inversion area model, sort the units in the inversion area, and lock the boundary area of each anomaly according to the boundary selection ratio;

[0031] The encryption module is configured to: after locking the abnormal body and the boundary area, encrypt the grid of the area once and multiple times in sequence to optimize the inversion grid.

[0032] One or more of the above technical solutions have the following beneficial effects:

[0033] The present invention initially locates the anomaly region by selecting a grid with a relatively large model change. Then, by calculating the distance between each unit in the region and the region's center, combined with the model change parameter, an optimization matrix for the anomaly region is constructed. Regional optimization calculations are performed to further locate the anomaly region. Finally, the number of units occupying the anomaly's outer boundary is estimated based on the range of each anomaly region. This is used to determine the proportion of units selected for the anomaly region's boundary. By calculating an indicator factor based on the model gradient, the anomaly boundary units are located. This establishes a dual-guide indicator combined anomaly location method that can more accurately delineate the anomaly and its boundary region units, efficiently optimizing the inversion grid and improving the inversion grid quality.

[0034] The dual-guide combined anomaly location method of the present invention takes into account both physical property changes and physical property gradient changes, enabling rapid localization of anomalies and their boundary regions. This method allows for relatively accurate delineation of anomalies and their boundary regions. As inversion iterations progress, adaptive mesh refinement of the anomaly and its boundary regions is performed to varying degrees, enabling more efficient optimization of the inversion mesh and improving its quality.

[0035] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0037] Figure 1 Schematic diagram of the grid optimization scheme with dual-guided combined encryption;

[0038] Figure 2 Schematic diagram of the anomaly area optimization process;

[0039] Figure 3 One-time encryption diagram;

[0040] Figure 4 Schematic diagram of multiple encryption. DETAILED DESCRIPTION

[0041] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0042] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.

[0043] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0044] The overall idea proposed by the present invention is:

[0045] By selecting a grid with a relatively large model change, the anomaly region can be initially identified. Then, by calculating the distance between each unit in the region and the region center and combining it with the model change parameter, an optimization matrix for the anomaly region is constructed. Regional optimization calculations are performed to further identify the anomaly region. Finally, based on the range of each anomaly region, the number of units occupying the outer boundary of the anomaly is estimated. This is used to determine the proportion of units selected at the boundary of the anomaly region. By calculating the indicator factor based on the model gradient, the anomaly boundary units are identified.

[0046] Example 1

[0047] This embodiment discloses a tunnel DC resistivity inversion grid optimization method based on a dual-guide indicator combination. After triggering the grid encryption condition, the low-resistance anomaly and its boundary area are calculated based on the dual-guide indicator combination anomaly positioning method, and then the anomaly and its boundary area units are encrypted and optimized.

[0048] Specific steps include:

[0049] (1) The model change in the abnormal body area is often larger. Based on the geological survey, the scale of the abnormal body is estimated. By selecting a certain proportion of grids with relatively larger model change, the abnormal body area can be initially locked. Based on the basic situation of geological analysis and advance drilling, the number K of abnormal body areas in the tunnel face ahead can be preliminarily determined.

[0050] (2) In each inversion iteration, the model change of each unit in the inversion area is calculated according to the indicator factor of the model change. The unit numbers in the inversion area are sorted in descending order according to the model change, and about 8% of the number of units in the previous inversion area are selected to form the anomaly area. The anomaly area is roughly determined using a scheme based on the model change. Figure 1 As shown in (a).

[0051] The model change of each unit in the inversion area can be calculated by the indicator factor of the model change. The indicator factor of the model change can be expressed as:

[0052] Smt i =(m i -m0) 2

[0053] where m i is the i-th model parameter, and m0 is the i-th model parameter in the last inversion iteration.

[0054] (3) Calculate the center coordinates (x i ,y i ,z i ), and randomly select K coordinate points as u k (μ x ,μy ,μ z )u k (μ x ,μ y ,μ z ), calculate the minimum distance between the center point of each unit in the Ω1 area and the center of the K abnormal bodies, so that each unit in the Ω1 area is assigned to the center of its nearest abnormal body. At this time, the units in the Ω1 area are divided into K areas. Figure 1 As shown in (b), Figure 1 The middle hexagon represents the center of the anomaly.

[0055] (4) Repeat the previous step to calculate the coordinates of the center point of each area, and update the center of each abnormal body. Set the maximum number of repetitions, and finally divide the Ω1 area into K areas. Figure 1 The circled areas shown in (d) represent K regions respectively.

[0056] The coordinates of the center points of each area can be expressed as:

[0057]

[0058] N k is the number of unit grids in each area, u k (μ x ,μ y ,μ z ) is the center coordinate of the abnormal body, (x i ,y i ,z i ) are the center coordinates of each unit.

[0059] (5) After the abnormal body area is identified, the distance d between the center point of each abnormal body area unit and the center point of the abnormal body is calculated. i .

[0060] Calculate the distance d between the center point of each abnormal body area unit and the center point of the abnormal body i It can be expressed as:

[0061]

[0062] In order to more accurately select the abnormal body area and avoid the presence of non-abnormal body units in each abnormal body area, the regional optimization calculation of Ω1 is performed. Combined with the model change matrix Smt, the abnormal body area optimization standard matrix C is constructed to perform regional optimization calculation and further lock the abnormal body area. The element is c i .

[0063] Element c i It can be expressed as

[0064]

[0065] In this formula, α, β, and λ are the weight coefficients for adjusting the distance of each unit in Ω1 from the regional center point and the change in model parameters, respectively. Based on the geological survey results, the scale of the anomaly area is preliminarily estimated, and the values of the above weight coefficients are further determined. Their values are directly related to the volume of the anomaly, the roughness of the initial model grid, and the number of inversion iterations.

[0066] (6) Sort the unit numbers of the Ω1 region in descending order according to the elements of the standard matrix C of each abnormal body region, and select the Ω1 region units as the final abnormal body region Ω2 according to a certain (60%-80%) optimization ratio. The optimization process is as follows: Figure 2 As shown, the abnormal body region Ω2 is as follows Figure 2 (b) shows the circle unit composition, which largely eliminates non-abnormal units. Figure 2 The middle squares represent the outlier units that were removed.

[0067] (7) After determining the abnormal body region Ω2, the cells of the internal region of each abnormal body are locked, and then the number of cells of the external boundary of the abnormal body is estimated according to the range of each abnormal body region, and the selection ratio of the boundary cells of each abnormal body region is determined accordingly.

[0068] (8) Calculate the gradient indicator factor of the inversion region model. Similarly, sort the cells in the inversion region in descending order. According to the boundary selection ratio, lock the boundary region Ω3 of each anomaly. Find the common cells in the boundary region Ω3 and the anomaly region Ω2, and remove these cells from the boundary region Ω3.

[0069] While the anomaly region is encrypted based on the model change, the inner boundary of the anomaly is also encrypted. The cells at the inner boundary of the anomaly are likely to be doubly encrypted. This will result in significant differences in cell volume within this region, directly impacting model quality and reducing forward modeling accuracy and inversion performance. Therefore, it is necessary to identify common cells in the boundary region Ω3 and the anomaly region Ω2 and remove these cells from the boundary region Ω3.

[0070] The model gradient of each unit in the inversion area is calculated by the indicator factor of the model gradient. The indicator factor of the model gradient has the following expression:

[0071]

[0072] where m i is the i-th model parameter.

[0073] (9) After locking the abnormal body and the boundary area, the grid of the area is encrypted once and multiple times in sequence.

[0074] Specifically, one-time encryption means dividing the hexahedron in the locked area into multiple tetrahedrons by adding nodes. The division method is divided into two categories depending on whether nodes are added:

[0075] (1) The irregular hexahedral unit is divided into 6 tetrahedral units without adding nodes. Compared with the method of dividing the hexahedron into 5 tetrahedral units, the volume of the tetrahedral units generated by this method is more uniform.

[0076] (2) Adding a node inside the irregular hexahedron unit divides it into 12 tetrahedron units. That is, two new tetrahedron units can be generated for each node on each face of the irregular hexahedron and the newly added node. The location of the newly added node can be adjusted to control the unit quality of the generated tetrahedron unit.

[0077] Furthermore, the first division form is first adopted. When the unit quality of the generated tetrahedron does not meet the requirements, the second division method is adopted to control the quality of the tetrahedron unit by adding nodes and adjusting the positions of the nodes until the requirements are met.

[0078] Multiple encryption refers to the process of dividing a tetrahedron into multiple tetrahedrons. A node is added inside the original tetrahedron unit to divide it into 4 tetrahedron units.

[0079] By setting the minimum cell volume, the generation of too small cells during the mesh refinement process is limited, over-refinement in a certain area is avoided, and the uncertainty of the inversion is reduced.

[0080] (10) In this embodiment, the hexahedron unit is first divided into 6 tetrahedron units without adding nodes. Compared with the method of dividing the hexahedron into 5 tetrahedron units, the volume of the tetrahedron units generated by this method is more uniform. Assuming that a hexahedron is (123456), it can be divided into six tetrahedrons: (2678), (2378), (2348), (2568), (1258), and (1248). Figure 3 shown.

[0081] (11) When the quality of the generated tetrahedral unit does not meet the requirements, a node is added inside the hexahedral unit to divide it into 12 tetrahedral units. That is, two new tetrahedral units are generated for each node on each face of the hexahedron and the newly added node. The position of the newly added node can be adjusted to control the unit quality of the generated tetrahedral unit until the requirements are met.

[0082] (12) Through multiple encryption such as Figure 4 As shown in the figure, the minimum cell volume is set to limit the generation of too small cells during the mesh refinement process, avoid over-refinement in a certain area, and reduce the uncertainty of the inversion.

[0083] In an embodiment of the present invention, a dual-guide indicator combined with an adaptively refined inversion grid optimization method can more effectively identify anomalies and their boundary regions. First, an indicator factor based on model change is calculated. Model change in the anomaly region is often greater. By selecting a grid with a relatively larger model change, the anomaly region can be initially located. Then, by calculating the distance between each unit in the region and the region center and combining it with the model change parameter, an optimization matrix for the anomaly region is constructed. Regional optimization calculations are then performed to further locate the anomaly region. Finally, based on the range of each anomaly region, the number of units occupying the anomaly's outer boundary is estimated. This ratio is then used to determine the selection ratio of the anomaly region boundary units. By calculating an indicator factor based on the model gradient, the anomaly boundary units are located. A dual-guide indicator combined anomaly location method has been established that can more accurately delineate the anomaly and its boundary region units, and adaptively refine the grid in that region, enabling more efficient optimization of the inversion grid and improving the inversion grid quality.

[0084] Example 2

[0085] The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the program.

[0086] Example 3

[0087] The purpose of this embodiment is to provide a computer-readable storage medium.

[0088] A computer-readable storage medium stores a computer program, which, when executed by a processor, performs the steps of the above method.

[0089] Example 4

[0090] The purpose of this embodiment is to provide a tunnel DC resistivity inversion grid optimization system based on a dual-guide indicator combination, including:

[0091] The module for initially locking the abnormal body region is configured to calculate the model change amount of each unit in the inversion region according to the indicator factor of the model change amount, and initially lock the abnormal body region;

[0092] The abnormal body area final determination module is configured to: perform regional optimization calculation on the initially locked abnormal body area to determine the final abnormal body area, and avoid the presence of non-abnormal body area units in each abnormal body area;

[0093] The anomaly and boundary region locking module is configured to: after determining the anomaly region, lock the cells of each anomaly region, then estimate the number of cells occupied by the outer boundary of the anomaly based on the range of each anomaly region, and use this to determine the selection ratio of the boundary cells of each anomaly region;

[0094] Calculate the gradient indicator factor of the inversion area model, sort the units in the inversion area, and lock the boundary area of each anomaly according to the boundary selection ratio;

[0095] The encryption module is configured to: after locking the abnormal body and the boundary area, encrypt the grid of the area once and multiple times in sequence to optimize the inversion grid.

[0096] The steps involved in the apparatuses of Examples 2, 3, and 4 above correspond to those of Method Example 1. For detailed implementations, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media containing one or more instruction sets; it should also be understood to include any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and causing the processor to perform any method of the present invention.

[0097] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0098] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A tunnel DC resistivity inversion grid optimization method based on a dual-guide indicator combination is characterized by: include: According to the geological survey, the scale of the anomaly is estimated, and the anomaly area can be initially located by selecting a grid with a relatively large model change at a certain scale; Calculate the model change of each unit in the inversion area according to the indicator factor of the model change, and preliminarily locate the abnormal body area; Perform regional optimization calculation on the initially locked anomaly area to determine the final anomaly area, avoiding the presence of non-anomalous area units in each anomaly area; After the abnormal body area is determined, the cells of the internal area of each abnormal body are locked, and then the number of cells occupied by the external boundary of the abnormal body is estimated according to the range of each abnormal body area, and the selection ratio of the boundary cells of each abnormal body area is determined accordingly; Calculate the gradient indicator factor of the inversion area model, sort the units in the inversion area, and lock the boundary area of each anomaly according to the boundary selection ratio; After locking the abnormal body and boundary area, the grid of the area is then encrypted once and multiple times in sequence to optimize the inversion grid.

2. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 1 is characterized in that: Lock the abnormal area, specifically: The unit numbers of the inversion area are sorted in descending order according to the model change amount, and eight percent of the number of units in the previous inversion area are selected to form the anomaly area to determine the anomaly area.

3. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 1 is characterized in that: include: Calculate separately The center coordinates of each unit in the region, and randomly select K coordinate points as the center of the abnormal body, calculate The minimum distance between the center point of each unit in the region and the center of K abnormal bodies, so that Each cell in the region is assigned to the center of its nearest anomaly.

4. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 3 is characterized in that: include: Calculate the distance between the center point of each abnormal body area unit and the center point of the abnormal body , by each unit of each abnormal body area Composition distance factor matrix; Based on the distance factor matrix Perform regional optimization calculations to avoid the appearance of units in non-abnormal areas in each abnormal area.

5. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 1 is characterized in that: When determining the final abnormal body area, The unit number of the area is based on the standard matrix of each anomaly area Elements are sorted in descending order and selected according to a certain optimization ratio The regional unit serves as the final anomaly region .

6. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 1 is characterized in that: include: The common cells in the boundary region and the final outlier region are determined, and these common cells are removed from the boundary region.

7. The tunnel DC resistivity inversion grid optimization method based on dual-guide indicator combination according to claim 1 is characterized in that: One-time encryption means dividing the hexahedron in the locked area into multiple tetrahedrons by adding nodes. The division is divided into two categories depending on whether nodes are added: (1) Divide the irregular hexahedral element into 6 tetrahedral elements without adding nodes; (2) Add a node inside the irregular hexahedron unit to divide it into 12 tetrahedron units. That is, two new tetrahedron units are generated for the node on each face of the irregular hexahedron and the newly added node. The position of the newly added node is adjusted to control the unit quality of the generated tetrahedron unit.

8. A tunnel DC resistivity inversion grid optimization system based on a dual-guide indicator combination is characterized by: include: The module for preliminary locking of the abnormal body area is configured to: estimate the scale of the abnormal body according to the geological survey, and preliminary lock the abnormal body area by selecting a grid with a relatively large model change amount at a certain ratio; Calculate the model change of each unit in the inversion area according to the indicator factor of the model change, and preliminarily locate the abnormal body area; The abnormal body area final determination module is configured to: perform regional optimization calculation on the initially locked abnormal body area to determine the final abnormal body area, and avoid the presence of non-abnormal body area units in each abnormal body area; The anomaly and boundary region locking module is configured to: after determining the anomaly region, lock the cells of each anomaly region, then estimate the number of cells occupied by the outer boundary of the anomaly based on the range of each anomaly region, and use this to determine the selection ratio of the boundary cells of each anomaly region; Calculate the gradient indicator factor of the inversion area model, sort the units in the inversion area, and lock the boundary area of each anomaly according to the boundary selection ratio; The encryption module is configured to: after locking the abnormal body and the boundary area, encrypt the grid of the area once and multiple times in sequence to optimize the inversion grid.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 7 are performed.

Images