A tunnel integral waterproofing construction method

The tunnel data is obtained through three-dimensional laser scanning and image acquisition equipment, combined with finite element analysis to identify risk areas, and targeted pre-processing construction is carried out, which solves the problem of insufficient targeted waterproofing construction methods in the existing tunnel, and improves the level of refinement of construction and Waterproof effect.

CN118499029BActive Publication Date: 2024-11-29THE GUANGDONG NO 3 WATER CONSERVANCY & HYDRO ELECTRIC ENG BOARD CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410583298.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-11
Publication Date
2024-11-29
Estimated Expiration
2044-05-11

AI Technical Summary

Technical Problem

The existing tunnel waterproof construction methods are difficult to fully consider the on-site conditions, working surface flatness and seepage conditions, resulting in insufficient construction targeting and prone to seepage problems.

Method used

Three-dimensional laser scanning and image acquisition equipment are used to obtain the three-dimensional model and surface images of the tunnel, identify the seepage area through image processing, and construct risk area data in combination with finite element analysis, and carry out targeted pre-processing construction.

Benefits of technology

It improves the refinement level of tunnel waterproofing construction, enhances the effectiveness of waterproofing projects, reduces the risk of water seepage, and extends the service life of the tunnel.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118499029B_ABST
    Figure CN118499029B_ABST
Patent Text Reader

Abstract

The present invention belongs to the technical field of construction simulation, and provides a method for overall waterproof construction of a tunnel, which uses three-dimensional laser scanning to obtain a three-dimensional model of the tunnel, and obtains data of a pre-processed area of ​​the working face by comparing it with design data; uses an image acquisition device to obtain an image of the tunnel surface, and obtains data of key waterproof areas through image recognition; uses field survey data, three-dimensional models, and data of key waterproof areas to construct a finite element analysis model to obtain data of risk areas; performs pre-processing operations based on the data of pre-processed areas of the working face and data of key waterproof areas; and then performs conventional waterproof operations. The present invention relies on three-dimensional laser scanning and image acquisition equipment to quickly determine defects existing on the working face of tunnel waterproof construction, identify water seepage areas, and then obtain risk area data through finite element analysis, and then performs targeted pre-processing on defects, water seepage areas, and risk areas, which can improve the level of refinement of construction and provide the effectiveness of waterproof engineering.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of construction simulation, and in particular relates to an integral waterproof construction method for a tunnel. Background Art

[0002] Tunnels are underground projects that are affected by a variety of external factors, such as groundwater and surface water. If there are no effective waterproofing measures inside the tunnel, groundwater may seep into the tunnel, causing damage to the tunnel structure, or even seepage and leakage, which will endanger the safety of the tunnel. The overall waterproofing construction of the tunnel effectively prevents groundwater leakage and erosion of the tunnel structure, prolongs the service life of the tunnel, and reduces maintenance and reinforcement costs.

[0003] At present, tunnel waterproofing construction is mostly carried out by laying waterproof materials, which often fails to fully consider the impact of site conditions, working surface flatness, on-site water seepage and other factors on waterproofing performance, which can easily lead to insufficient targeted construction and water seepage after completion.

[0004] With the development of detection technology and digital technology, construction simulation related technologies such as 3D modeling and risk analysis can be used for targeted detection before construction to guide engineering operations. Related technologies are widely used in civil engineering fields such as slope deformation monitoring, rock and gravel identification, and tunnel disease identification and measurement. However, they have not been used in the overall waterproofing construction of tunnels. Summary of the invention

[0005] The purpose of the present invention is to propose a tunnel integral waterproofing construction method to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present invention, a method for integral waterproofing construction of a tunnel is provided, and the method comprises the following steps: The method comprises the following steps: S100, using a three-dimensional laser scanning device to obtain a three-dimensional model of the tunnel, comparing it with the design data, and obtaining the data of the pretreatment area of ​​the working face; S200, using an image acquisition device to obtain a surface image of the tunnel, and identifying the water seepage area through image processing to obtain the data of the key waterproofing area; S300, using on-site surveys to obtain hydrological data and tunnel geological data around the tunnel, the three-dimensional model of the tunnel obtained in step S100 and the data of the pretreatment area of ​​the working face, and the data of the key waterproofing area obtained in step 200, construct a finite element analysis model, and obtain risk area data; S400, performing working face pretreatment construction operations according to the data of the pretreatment area of ​​the working face obtained in step S100; S500, performing key waterproofing area pretreatment construction operations according to the key waterproofing area data and the risk area data obtained in steps S200 and S300; S600, performing conventional waterproofing operations.

[0007] Furthermore, the working surface pretreatment area data in steps S100 and S200 include but are not limited to the defect type, defect location, and defect area of ​​the protrusions, recesses, and cracks; the waterproof key area data include the location and area of ​​the seepage area.

[0008] Furthermore, the image acquisition device in step S200 includes a high-definition camera and a fill light; the image processing and identification in step S200 are as follows: the acquired image is imported into the image processing software, the image is preprocessed by denoising and contrast enhancement, the water seepage area is identified by color space conversion and threshold setting, and the area of ​​the water seepage area is obtained by edge detection and morphological operations.

[0009] Furthermore, the image processing and identification steps in step S200 are as follows: S2001, in the acquisition of tunnel surface images, for the same position, the image acquisition device obtains several images at the same position and under the same fill light conditions after a certain time interval, and the image data includes image data, position data, and acquisition number data; S2002, the acquired images are uniformly selected for a certain acquisition data, imported into the image processing software, and denoising and contrast enhancement preprocessing operations are performed on the images, the first water seepage area is identified by color space conversion and threshold setting, and the edge of the first water seepage area is obtained by edge detection and morphological operations; S2003, for the image of the first water seepage area, several images acquired multiple times at the same position are obtained, and a unified preprocessing is performed on the multiple images, including denoising and contrast enhancement; S2004, the Lucas-Kanade method of the optical flow method is used to evaluate the motion vectors of the feature points in the image sequence, if the feature points with a consistent motion pattern in the moving feature points exceed a preset ratio, it is determined that the area has a water seepage problem and is marked as a waterproof key area; the feature points include but are not limited to points on the edge of the first water seepage area.

[0010] Furthermore, the image processing and recognition steps in step S200 are as follows: S2001, in acquiring the tunnel surface image, the image acquisition device gradually acquires the image according to the moving distance a, the image frame size in the tunnel direction is 3a, and the image data includes image data and position data; the fill light is a rectangular fill light, and the light-emitting surface size is less than or equal to a; S2002, the acquired image is separated into 3 images with a distance a in the tunnel direction, thereby obtaining 3 images at the same position, which are a front oblique view image, a front view image, and a rear oblique view image; S2003, importing the front view image of the same position into the image processing software, performing denoising and contrast enhancement preprocessing operations on the image, and identifying the first water seepage area through color space conversion and threshold setting; S2004, obtaining the front oblique view image, the front view image, and the rear oblique view image for the image of the first water seepage area, analyzing the shadow and highlight area data thereof, and comparing the differences in the shadow and highlight area data of the front oblique view image, the front view image, and the rear oblique view. If the difference exceeds the preset threshold, it is determined that the area has a water seepage problem and is marked as a key waterproofing area.

[0011] Furthermore, the judgment of the difference in shadow and highlight area data in step S2004 includes the following steps: using threshold segmentation and edge detection to identify shadow and highlight areas for the front oblique view image, the front view image, and the rear oblique view image; comparing the shadow and highlight areas of the same water seepage area in the three images, if the changes in their size and position exceed a preset threshold, it is judged that there is a water seepage problem in the area.

[0012] Furthermore, the finite element analysis in step S300 includes the following steps: S3001. Importing a three-dimensional model, importing the three-dimensional tunnel model obtained in step S100 into the PLAXIS software; S3002. Defining the material properties of different geological layers in the model based on the tunnel geological data obtained from the on-site survey; S3003. Applying a water pressure load, determining the water pressure using the hydrological data around the tunnel, and applying a water pressure load to the corresponding position in the model; S3004. Meshing, meshing the three-dimensional tunnel model, and refining the mesh for the working face pretreatment area and the water seepage area; S3005. Water seepage simulation, obtaining the deformation of the model mesh through finite element analysis by applying a water pressure load, and marking the mesh as a risk mesh when the deformation exceeds a threshold value, treating multiple adjacent risk meshes and risk meshes with a number of interval meshes less than a preset number and the interval meshes as a continuous risk area, and obtaining the risk level based on the deformation data of each mesh in the risk area.

[0013] Furthermore, the finite element analysis in step S300 also includes the following steps: S3006. Model verification and parameter adjustment, comparing the risk area obtained by the seepage simulation with the key waterproofing area. If the deviation is within the set range, the model is considered accurate; if not, adjust the relevant parameters in the seepage simulation, and then repeat steps S3005 and S3006; S3007. After the model is accurate, the hydrological data is amplified according to the site needs to obtain the risk area data under extreme preset conditions.

[0014] The present invention provides a method for overall waterproofing construction of a tunnel. Relying on three-dimensional laser scanning and image acquisition equipment, the defects on the tunnel waterproofing construction working surface can be quickly determined and the water seepage area can be identified. Then, the risk area data can be obtained through finite element analysis. Then, targeted pretreatment is performed on the defects, water seepage areas and risk areas, which can improve the level of refinement of the construction and provide the effectiveness of the waterproofing project.

[0015] In identifying water seepage areas, the present application can improve the accuracy of identification through optical flow method or multi-angle image method in image analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other features of the present invention will become more obvious by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present invention represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings:

[0017] Figure 1 Shown is a flow chart of the tunnel integral waterproofing construction method of the present invention;

[0018] Figure 2 The figure shows a schematic diagram of an image acquisition scheme in water seepage area identification according to an embodiment of the present invention;

[0019] Figure 3 Shown is a schematic diagram of an image acquisition scheme for water seepage area identification according to another embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0021] like Figure 1 The figure shows a flow chart of a tunnel integral waterproofing construction method. Figure 1A tunnel integral waterproofing construction method according to an embodiment of the present invention is described below. The method comprises the following steps:

[0022] S100, using a 3D laser scanning device to scan along the tunnel longitudinally to obtain a 3D model of the tunnel, comparing the 3D model collected on site with the design data in detail, analyzing the difference between the actual surface shape of the tunnel and the design, and obtaining data of the pre-processed area of ​​the working face, including but not limited to the defect type, defect location, and defect area of ​​the protruding part, the concave part, and the crack part;

[0023] S200, using an image acquisition device to scan along the longitudinal direction of the tunnel to obtain a tunnel surface image, and identifying the water seepage area through image processing to obtain waterproof key area data, including the location of the water seepage area and the area of ​​the water seepage area;

[0024] S300, using on-site survey to obtain hydrological data around the tunnel, tunnel geology, and structure (crack) related data; constructing a finite element analysis model based on the tunnel three-dimensional model and working face pre-processing area data obtained in step S100 and the waterproof key area data obtained in step 200 to obtain risk area data;

[0025] S400, performing work surface pretreatment construction work according to the work surface pretreatment area data obtained in step S100;

[0026] S500, performing pre-treatment construction work on the key waterproofing areas according to the key waterproofing area data and risk area data obtained in steps S200 and S300;

[0027] S600, for routine waterproofing work.

[0028] The three-dimensional laser scanning device in step S100 is a conventional three-dimensional laser scanning device in the field. To improve work efficiency, multiple devices can be used simultaneously. The three-dimensional laser scanning device collects tunnel surface morphology data to obtain a tunnel three-dimensional model L1; the design data is the tunnel design three-dimensional model L2 in the original design. By comparing the tunnel three-dimensional model L1 and the tunnel design three-dimensional model L2, the difference between the two models is obtained, and defects occurring in tunnel construction are identified. The defect types include but are not limited to protrusions, depressions, and cracks, and the defect locations and defect areas are determined.

[0029] The conventional knowledge in the field is used to obtain the tunnel three-dimensional model L1 by using a three-dimensional laser scanning device. The commercially available three-dimensional laser scanning devices generally have built-in related software. The comparison between the tunnel three-dimensional model L1 and the tunnel design three-dimensional model L2 can be carried out by conventional methods in the field.

[0030] In one embodiment of the present application, defects are identified by directly comparing point clouds and calculating the distance and distribution differences between point clouds. First, the acquired tunnel 3D model L1 is filtered to remove small fluctuations, such as fluctuations within 2 cm. Then, the tunnel 3D model L1 and the tunnel design 3D model L2 are directly compared in point clouds. When the distance difference between the two points exceeds a preset value, it is a defective area.

[0031] The image acquisition device in step S200 can be a common high-definition camera in the field. In order to avoid changes in lighting, a fill light is also provided. The scanning is moved longitudinally along the tunnel. To improve work efficiency, multiple devices can be used simultaneously.

[0032] The following scheme can be adopted for the process of processing and identifying the tunnel surface images: import the collected images into image processing software, perform preprocessing operations such as denoising and contrast enhancement on the images to highlight the characteristics of the water seepage area. Taking advantage of the fact that the water seepage area is generally darker than the surrounding area, the water seepage area is identified through color space conversion (such as from RGB to grayscale) and threshold setting, and the area of ​​the water seepage area is obtained through edge detection and morphological operations (such as opening and closing operations to remove noise and fill small holes).

[0033] As attached Figure 2 As shown, in order to further improve the recognition quality, the present application also provides a recognition method based on the optical flow method. For the same position, several images are obtained after a certain interval. For the water seepage area, the water surface will flow on the rock surface, and its brightness data in the image will also change. By analyzing the optical flow method of multiple images, the optical flow field change data in the image can be obtained, and the optical flow field change data can be used to assist in determining whether there is a water seepage problem in the area. The specific steps of this method are as follows:

[0034] 1. In the acquisition of tunnel surface images, for the same position, the image acquisition device obtains several images at the same position and under the same fill light conditions after a certain interval. The image data includes image data, position data, and acquisition times data;

[0035] 2. The collected images are uniformly selected from a certain collection data, imported into the image processing software, and pre-processed by denoising, contrast enhancement and other operations. The first water seepage area is identified by color space conversion (such as from RGB to grayscale) and threshold setting, and the edge of the first water seepage area is obtained by edge detection and morphological operations;

[0036] 3. For the image of the first water seepage area, a number of images collected multiple times at the same position are obtained, and a unified preprocessing is applied to the multiple images, including denoising, contrast enhancement, etc.;

[0037] 4. Use the optical flow method to determine whether the first water seepage area has water seepage problems. Specifically, use the Lucas-Kanade method to evaluate the motion vectors of feature points in the image sequence. If the number of feature points with consistent motion patterns among the moving feature points exceeds a preset ratio, it is determined that the area has water seepage problems and is marked as a key waterproof area. Feature points include but are not limited to points on the edge of the first water seepage area.

[0038] For the determination of consistent motion patterns, the motion vectors of all feature points are statistically analyzed to calculate their average and covariance. If the motion vectors of most feature points are concentrated in a certain direction, this can be regarded as a consistent motion pattern.

[0039] As attached Figure 3 As shown, in order to further improve the recognition quality, this application also provides another recognition method, the image acquisition device moves along the tunnel direction to shoot, the size of the picture taken at a certain position in the tunnel direction is 3a, after shooting is completed, the image acquisition device moves forward a distance a to shoot again, and the tunnel surface of size a is used as a judgment area. For the same area, 3 images of different angles will be formed, located in the front, center, and back of the shooting picture, with different imaging angles, and the shadows and highlights caused by different angles of illumination are analyzed to improve the recognition effect. The specific steps of this method are as follows:

[0040] 1. In acquiring images on the tunnel surface, the image acquisition device gradually acquires images according to the moving distance a. The image frame size in the tunnel direction is 3a. The image data includes image data and position data. The fill light is a rectangular fill light, and the size of the luminous surface is less than or equal to a.

[0041] 2. The collected image is divided into three images with a distance of a in the tunnel direction, thereby obtaining three images at the same position, namely, a front oblique view image, a front view image, and a rear oblique view image;

[0042] 3. Import the front view image of the same position into the image processing software, perform pre-processing operations such as denoising and contrast enhancement on the image, and identify the first water seepage area through color space conversion (such as from RGB to grayscale) and threshold setting;

[0043] 4. For the image of the first water seepage area, obtain the front oblique image, front view image, and rear oblique image to analyze the shadow and highlight area data. By comparing the differences in the shadow and highlight area data of the front oblique image, front view image, and rear oblique image, if the difference in the shadow and highlight area data exceeds the preset threshold, it is determined that the area has a water seepage problem and is marked as a key waterproof area. Since the water surface has a clear direction for the reflection of the fill light source, and the rock surface generally exhibits diffuse reflection, its reflection direction is significantly different from that of the water surface.

[0044] The process of judging the difference of shadow and highlight area data is as follows: For the front oblique image, front view image, and rear oblique image, threshold segmentation and edge detection are used to identify the shadow and highlight areas. The shadow area usually appears as a dark area in the image, while the highlight area appears as an area with higher brightness. Compare the shadow and highlight areas of the same water seepage area in the three images. If the changes in their size and position exceed the preset threshold, it is judged that there is a water seepage problem in the area.

[0045] The specific steps may include:

[0046] 1. Determine the threshold and set the brightness threshold for the shadow and highlight areas. The shadow threshold is usually lower than the average brightness value of the image, and the highlight threshold is higher than the average brightness value. Use the determined threshold to binarize each image to separate the shadow and highlight areas.

[0047] 2. Edge detection: Apply edge detection algorithms (such as Canny, Sobel, or Laplacian algorithms) to the binarized image to determine the boundaries of shadow and highlight areas.

[0048] 3. Count the number of pixels in each shadow and highlight area to quantify its size.

[0049] 4. Size and position comparison: The size of the shadow area and highlight area is obtained by comparing the difference in the total number of pixels in the corresponding area. For position comparison, the overlap area ratio is calculated by the area overlap method. The overlap area ratio is obtained by using software such as MATLAB to obtain the total number of pixels in the intersection of the shadow area and highlight area of ​​the two images, and then the total number of pixels in the shadow area and highlight area of ​​the two images is obtained. For the front oblique image, the front view image, and the rear oblique image, three sets of pixel number difference and overlap area ratio data are obtained by comparing them pairwise. The average value of the above three sets of data is taken to represent the difference.

[0050] 5. If the difference in the above steps is greater than the preset value, it is considered that there is a water seepage problem in the area. Specifically, the average value of the pixel number difference and the overlap area ratio after comparing the front oblique image, the front view image, and the rear oblique image is obtained, and the preset values ​​of the pixel number difference and the overlap area ratio are set. If one of the values ​​exceeds the preset value, it is determined that there is a water seepage problem.

[0051] The above image processing can be performed using MATLAB, OpenCV, and Python. The specific implementation processes of threshold segmentation, edge detection, and other solutions mentioned in the above scheme are common knowledge in the field.

[0052] The specific process of processing the finite element analysis model in step S300 is as follows: Import the model obtained in step S100 into the finite element analysis software. In one embodiment of the present application, PLAXIS finite element software is used, and Visual-modflow, ABAQUS or other finite element software may also be used. Use field surveys to obtain tunnel geological data, including but not limited to material properties of different geological layers, such as soil layers, rock layers, etc. to define material properties. Use field surveys to obtain hydrological data around the tunnel to determine boundary conditions and water pressure loads. Perform grid division, refine the grid for the seepage area, perform seepage simulation, and determine the risk area data. Specifically include the following steps: 1. Import the three-dimensional model, import the three-dimensional tunnel model obtained in step S100 into the PLAXIS software.

[0053] 2. Based on the tunnel geological data obtained from the on-site survey, define the material properties of different geological layers in the model. In one embodiment of the present application, the geological data includes soil and rock layer types, and the elastic modulus, Poisson's ratio, internal friction, cohesion, permeability, porosity, etc. of the soil and rock layers are obtained based on on-site sampling.

[0054] 3. Applying water pressure loads, using the hydrological data around the tunnel, apply water pressure loads to the corresponding positions in the model. In one embodiment of the present application, the geological data includes groundwater elevation data and underground river location data. The boundary conditions of the model are set according to the hydrological data. Specifically, the water flow direction is determined according to the on-site hydrological data, and the boundary on one side (or multiple sides) is set as a free seepage boundary, indicating that the groundwater around the tunnel can flow freely. The boundary on the other side (or multiple sides) can be set to a fixed water level boundary / head boundary condition, indicating that the groundwater level on this side is fixed at a certain elevation value, and then the groundwater flow is numerically simulated (such as using hydrogeological software such as MODFLOW to simulate the groundwater level and water pressure). The water pressure distribution can also be obtained by on-site measurement.

[0055] 4. Meshing: Mesh the three-dimensional model of the tunnel and refine the mesh in the working surface pretreatment area (the working surface pretreatment area includes but is not limited to areas with defects such as protrusions, depressions, and cracks) and key waterproofing areas to improve the accuracy of the analysis.

[0056] 5. Water seepage simulation: by applying water pressure load, the deformation of the model grid is obtained through finite element analysis. When the deformation exceeds the threshold, the grid is marked as a risk grid. Multiple adjacent risk grids and risk grids with less than a preset number of interval grids and interval grids are regarded as a continuous risk area, and the risk level is obtained according to the deformation data of each grid in the risk area. The risk level includes the total deformation of each grid, the average deformation of the grid, and the maximum deformation of the grid. The risk level data is obtained by weighted average of the above data. The risk level data is a relative value, which is used to characterize the risk of water seepage and to match the subsequent pretreatment operations.

[0057] The mesh deformation distribution map obtained by finite element analysis can also be used to identify risk meshes through color space conversion and threshold setting, and the risk area can be obtained through edge detection, morphological operations (such as opening and closing operations to remove noise and fill small holes).

[0058] Furthermore, since the above-mentioned finite element analysis requires a large amount of field data, the relevant data is often incomplete and difficult to collect in practice. Therefore, the present application also provides a solution for using the identified waterproof key areas to perform model verification and parameter adjustment on the above-mentioned finite element analysis results.

[0059] On the basis of the above scheme, steps 5 and 6 are also included, namely, model verification and parameter adjustment. The risk area obtained by water seepage simulation is compared with the key waterproofing area. If the deviation is within the set range, the model is considered accurate. If not, the relevant parameters in the water seepage simulation are adjusted, and then steps 5 and 6 are repeated.

[0060] 7. After the model is accurate, the hydrological data is amplified according to the needs of the site to obtain the risk area data under extreme preset conditions.

[0061] In this application, mesh deformation data under water pressure is used to characterize water seepage risk, which can simplify the water seepage analysis process. The water seepage simulation in step 5 can be performed in the following steps:

[0062] ΔV x = k(F x *A x *(1-2v)) / E

[0063] ΔV y = k(F y *A y *(1-2v)) / E

[0064] ΔV z = k(F z *A z *(1-2v)) / E

[0065] K=f(n,i)

[0066]

[0067] Where ΔV is the volume deformation of the grid, Fx, Fy, Fz are the water pressures on the grid in the x, y, z directions, ΔVx, ΔVz, ΔVz are the deformations of the grid in the x, y, z directions, Ax, Ay, Az are the grid areas affected by the water pressure in the x, y, z directions; E is the elastic modulus of the material, ν is the Poisson's ratio of the material. n, i are the porosity and permeability of the material corresponding to the grid, respectively. K is the correction coefficient used to reflect the influence of the material's water permeability on the deformation.

[0068] Correspondingly, in step 5, the model verification and parameter adjustment include the following processes:

[0069] 6.1 According to the waterproof key area data obtained in step S200, the grid corresponding to the waterproof key area;

[0070] 6.2 Compare the risk grid obtained in step 5 with the grid corresponding to the waterproof key area obtained in step 6.1, and evaluate the simulation effect of step 5 through the comparison result; the simulation effect can be evaluated by the accuracy of the risk grid predicted in step 5;

[0071] 6.3 Adjust the parameter settings in the model in step 5 based on the evaluation results.

[0072] Furthermore, the above step 6.3 can be assisted by a neural network, which specifically includes the following steps:

[0073] 6.3.1. A typical section is selected in the tunnel, and the physical parameters corresponding to each grid after the three-dimensional model of the tunnel is divided are obtained as input values. The physical parameters include but are not limited to the elastic modulus (E), Poisson's ratio (ν), porosity (n), and permeability (i) of the material. Whether the grid belongs to the grid corresponding to the waterproof key area is used as the output value, if so, it is recorded as 1, otherwise it is recorded as 0; specifically, the above data set is divided into a training set and a validation set.

[0074] 6.3.2, construct a BP neural network model and perform model training. When the model error value converges, determine the mutual weights between the input values, and adjust the parameter settings in the model in step 5 according to the weights.

[0075] Specifically, a back-propagation neural network model can be selected, which is provided with an input layer, a hidden layer, and an output layer, wherein the number of neurons in the input layer is equal to the number of physical parameters, and the number of neurons in the hidden layer is twice that. A model with an error within a set range is obtained through iterative training and model verification, thereby obtaining the mutual weights between the input values.

[0076] In terms of loss function, this application adopts the following loss function:

[0077] L = Lfn + (1-Lf) Lfp

[0078] Where Lfn is the false negative loss (the model incorrectly predicts the waterproof key area as a non-key area), which can be defined as the ratio of the number of false negative predictions to the total number of predictions. Lfp is the false positive loss (the model incorrectly predicts the non-key area as the key area), which can be defined as the ratio of the number of false positive predictions to the total number of predictions. λ is a weight parameter used to adjust the relative importance between false negatives and false positives. Specifically, the false negative loss and false positive loss are as follows:

[0079]

[0080] Where I is the indicator function, y i is a real label, that is, whether it belongs to the grid corresponding to the waterproof key area, is the prediction value, t is the classification threshold, and in one embodiment of the present application, it is set to 0.5. The λ value can be set according to the risk tendency, such as paying more attention to the occurrence of false negatives, and the λ value can be greater than 0.5.

[0081] Step S400: The work surface pretreatment operation is targeted at the defect type in the work surface pretreatment area data, such as polishing the protrusions, filling the concave parts and cracks with concrete. The specific process is conventional knowledge in the field.

[0082] In step S500, pre-treatment construction work is performed on the waterproof key area data and risk area data obtained in S200 and S300. Different construction plans need to be selected according to the specific conditions of the waterproof key area and risk area during the specific construction work: For water seepage caused by cracks, grouting technology can be used, such as cement slurry, epoxy resin or other chemical slurry, which are injected into the cracks to seal the cracks. For water seepage surfaces without obvious cracks, cleaning operations are carried out before waterproofing construction to remove loose particles, dust and other impurities, provide a clean surface for the subsequent bonding of waterproof materials, and perform air drying before construction; for large-flow water seepage areas, a drainage system is set up to guide water to the tunnel drainage system to reduce the pressure of water on the tunnel surface.

[0083] In addition, different response methods are adopted for different risk levels, such as setting up multi-layer waterproof membranes and strengthening pads in the pretreatment of high-risk locations.

[0084] Conventional waterproofing operations in step S600 are conventional knowledge in the field. The following are the process steps of an embodiment of the present application: construct a waterproof layer on the initial support surface, which is commonly sprayed with a waterproof membrane, laid with a waterproof board, and applied with a waterproof coating. After the waterproof layer is constructed, a secondary lining is carried out, usually by pouring a concrete lining, to form the final inner surface of the tunnel. Seal the construction joints, deformation joints and other parts, and install water stop strips, water stop strips, etc. After that, install drainage pipes, blind ditches, etc. to form a drainage system for the tunnel to guide the seepage water to the outside of the tunnel.

[0085] Although the description of the present invention has been quite detailed and has been described in particular with respect to several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, so as to effectively cover the intended scope of the present invention. In addition, the present invention is described above with the embodiments foreseeable by the inventors, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that are not currently foreseen may still represent equivalent changes of the present invention.

Claims

1. A tunnel integral waterproofing construction method, characterized in that: The method The method comprises the following steps: S100, using a three-dimensional laser scanning device to obtain a three-dimensional model of the tunnel, comparing it with design data, and obtaining data of a pre-processed area of ​​the working face; S200, using an image acquisition device to obtain a tunnel surface image, identifying a water seepage area through image processing to obtain waterproof key area data; S300, using the hydrological data and geological data of the tunnel surrounding obtained by the on-site survey, the tunnel three-dimensional model and the working face pre-processing area data obtained in step S100, and the waterproof key area data obtained in step 200, to construct a finite element analysis model and obtain risk area data; S400, performing work surface pretreatment construction work according to the work surface pretreatment area data obtained in step S100; S500, performing pre-treatment construction work on the key waterproofing areas according to the key waterproofing area data and risk area data obtained in steps S200 and S300; S600, for conventional waterproofing work; The image processing and identification steps in step S200 are as follows: S2001, in the acquisition of tunnel surface images, for the same position, the image acquisition device obtains several images at the same position and under the same fill light conditions after a certain interval, and the image data includes image data, position data, and acquisition number data; S2002, the acquired images are uniformly selected for a certain acquisition data, imported into the image processing software, and the images are preprocessed by denoising and contrast enhancement, the first water seepage area is identified by color space conversion and threshold setting, and the edge of the first water seepage area is obtained by edge detection and morphological operations; S2003, for the image of the first water seepage area, several images acquired multiple times at the same position are obtained, and a unified preprocessing is performed on the multiple images, including denoising and contrast enhancement; S2004, the Lucas-Kanade method of the optical flow method is used to evaluate the motion vectors of the feature points in the image sequence, if the feature points with consistent motion patterns in the moving feature points exceed the preset ratio, it is determined that the area has a water seepage problem and is marked as a waterproof key area; the feature points include but are not limited to points on the edge of the first water seepage area; Alternatively, the image processing and identification steps in step S200 are as follows: S2001, in acquiring the tunnel surface image, the image acquisition device gradually acquires the image according to the moving distance a, the image frame size in the tunnel direction is 3a, and the image data includes image data and position data; the fill light is a rectangular fill light, and the size of the luminous surface is less than or equal to a; S2002, the acquired image is divided into three images with a distance a in the tunnel direction, thereby obtaining three images at the same position, which are a front oblique image, a front view image, and a rear oblique image; S2003, the front view image at the same position is imported into the image processing software, and the image is subjected to denoising and contrast enhancement preprocessing operations, and the first water seepage area is identified by color space conversion and threshold setting; S2004, for the image of the first water seepage area, a front oblique image, a front view image, and a rear oblique image are obtained, and the shadow and highlight area data thereof are analyzed, and by comparing the differences in the shadow and highlight area data of the front oblique image, the front view image, and the rear oblique image, if the differences exceed the preset threshold, it is determined that the area has a water seepage problem and is marked as a key waterproof area.

2. The tunnel integral waterproofing construction method according to claim 1, characterized in that: The working surface pre-treatment area data in steps S100 and S200 include but are not limited to the defect type, defect location, and defect area of ​​the protrusion, depression, and crack; the waterproof key area data include the location and area of ​​the seepage area.

3. The tunnel integral waterproofing construction method according to claim 1, characterized in that: The judgment of the difference in shadow and highlight area data in step S2004 includes the following steps: using threshold segmentation and edge detection to identify shadow and highlight areas for the front oblique view image, the front view image, and the rear oblique view image; comparing the shadow and highlight areas of the same water seepage area in the three images, if the changes in their size and position exceed the preset threshold, it is judged that there is a water seepage problem in the area.

4. The tunnel integral waterproofing construction method according to claim 1, characterized in that: The finite element analysis in step S300 includes the following steps: S3001. Import the three-dimensional model, and import the three-dimensional model of the tunnel obtained in step S100 into the PLAXIS software; S3002. Based on the tunnel geological data obtained from the on-site survey, define the material properties of different geological layers in the model; S3003. Applying water pressure load: using hydrological data around the tunnel to determine water pressure, and applying water pressure load to corresponding positions in the model; S3004. Meshing: Meshing the three-dimensional model of the tunnel, and refining the mesh for the working face pretreatment area and the water seepage area; S3005. Water seepage simulation, through the application of water pressure load, the deformation of the model grid is obtained through finite element analysis. When the deformation exceeds the threshold, the grid is marked as a risk grid. Multiple adjacent risk grids and risk grids with a number of interval grids less than a preset number and the interval grids are regarded as a continuous risk area, and the risk level is obtained based on the deformation data of each grid in the risk area.

5. The tunnel integral waterproofing construction method according to claim 4, characterized in that: The finite element analysis in step S300 further includes the following steps: S3006. Model verification and parameter adjustment. Compare the risk area obtained by water seepage simulation with the key waterproof area. If the deviation is within the set range, the model is considered accurate; if not, adjust the relevant parameters in the water seepage simulation, and then repeat steps S3005 and S3006; S3007. After the model is accurate, the hydrological data is amplified according to site needs to obtain risk area data under extreme preset conditions.

Citation Information

Patent Citations

  • Tunnel disease detection device and method

    CN113252728A