Water-rich karst tunnel disaster early warning method and system

Abstract

The invention discloses a water-rich karst tunnel disaster early warning method and system, and belongs to the technical field of tunnel disaster early warning, and the method comprises the steps: obtaining a tunnel face image of a tunnel, carrying out the preprocessing of the tunnel face image, inputting the preprocessed tunnel face image into a pre-trained crack recognition model, and obtaining tunnel face crack distribution data; the method comprises the following steps: acquiring an eddy current field signal formed by pulse electromagnetic waves of a tunnel face, performing inversion analysis on the eddy current field signal to obtain underground medium electrical distribution, and obtaining two-dimensional images of different underground layers according to the underground medium electrical distribution corresponding to the pulse electromagnetic waves at different angles; establishing three-dimensional data of a water body in front of the tunnel face by adopting a three-dimensional inversion technology and a data fusion method and combining the two-dimensional images of different layers; obtaining vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; determining the structural stability of the tunnel according to the tunnel face crack distribution data, the three-dimensional data of the water body, the vertical deformation data and the horizontal deformation data of the tunnel and the tunnel structural stress data; according to the method, monitoring and early warning of the whole construction stage are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel disaster early warning, and in particular to a method and system for early warning of water-rich karst tunnel disasters. Background Art

[0002] The common caves and karst holes in karst areas bring great uncertainty to tunnel excavation, and karst geology is also full of cracks and faults. These geological disasters may suddenly appear during the construction process, causing tunnel collapse. Secondly, there is the problem of groundwater in karst tunnels. Karst geology usually has abundant groundwater resources, which may cause water gushing and seepage during construction, posing a huge challenge to the stability of tunnel construction.

[0003] Existing tunnel disaster warning methods all predict and output warning information based on a single tunnel structural parameter. However, there are many possible causes of tunnel disasters, such as cracks, tunnel deformation, water gushing and seepage. Therefore, a single-parameter warning system cannot accurately predict tunnel disasters. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a water-rich karst tunnel disaster early warning method and system, which can realize continuous and accurate monitoring of the safety status of the tunnel structure and timely early warning.

[0005] The present invention is achieved through the following technical solutions: A water-rich karst tunnel disaster early warning method comprises the following steps: Acquire the tunnel face image of the tunnel, preprocess the tunnel face image, input the preprocessed tunnel face image into the pre-trained crack recognition model, and obtain the tunnel face crack distribution data; The eddy current field signal formed by the pulse electromagnetic wave at the tunnel face is obtained, and the eddy current field signal is inverted and analyzed to obtain the electrical distribution of the underground medium. According to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles, two-dimensional images of different underground layers are obtained. The three-dimensional data of the water body in front of the tunnel face is established by using three-dimensional inversion technology and data fusion methods and combining two-dimensional images of different layers. Obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; The structural stability of the tunnel is determined based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

[0006] Preferably, the preprocessing of the tunnel face image includes grayscale processing, filtering processing and binarization processing.

[0007] Preferably, the crack recognition model is a pre-trained YOLOv5 model; The YOLOv5 model includes an input layer, a backbone layer, a neck layer and a head layer; The input layer is used to read image data, the backbone layer is used to extract feature information of the image, the neck layer is used to fuse the feature information, and the head layer is used to identify cracks according to the fused feature information.

[0008] Preferably, a water body detector is used to transmit pulse electromagnetic waves at different angles to the tunnel face, and the signal of the secondary eddy current field is received by a receiver.

[0009] Preferably, the method for acquiring the vertical deformation data includes: Three settlement measuring points are installed on the top of the tunnel, and the settlement deformation data of the tunnel is determined based on the images of the settlement measuring points and combined with the template feature matching method.

[0010] Preferably, the method for determining the settlement deformation data is as follows: The image data of the settlement measuring points are collected, and preprocessing operations such as denoising and enhancement are performed on the collected image data. The template feature matching algorithm is used to slide the template in the image to be matched, and the similarity measurement between the template and the image sub-region is calculated. The position with the highest similarity is selected as the matching result, and the matched pixel coordinates are converted into three-dimensional coordinates in the camera coordinate system. The coordinates in the camera coordinate system are converted into actual coordinates in the world coordinate system by combining the camera extrinsic parameters. The actual coordinate changes of the same measuring point in the images at different time points are compared to obtain the settlement deformation data.

[0011] Preferably, the method for determining the horizontal deformation data is as follows: According to the construction method, a horizontal convergence survey line is set in the tunnel, and the horizontal deformation data of the tunnel is obtained according to the horizontal convergence survey line; The method for determining the tunnel structure stress data is as follows: A structural stress monitor is set inside the initial support of the tunnel, and the tunnel structure stress data is obtained based on the structural stress monitor.

[0012] Preferably, the method for determining the structural stability of the tunnel is as follows: Determine the length and width of each crack based on the crack distribution data on the tunnel face, and determine the stability of the tunnel structure based on the comparison between the total area of ​​all cracks and the threshold value; The volume of the water body ahead is determined based on the three-dimensional data of the water body, and the stability of the tunnel structure is determined based on a comparison result between the volume of the water body and a first water body volume threshold; Determining the stability of the tunnel structure based on a comparison result of the vertical deformation data and a corresponding first vertical threshold value; The stability of the tunnel structure is determined based on a comparison result between the horizontal deformation data and a first horizontal threshold.

[0013] Preferably, the method for determining the structural stability of the tunnel is as follows: When cracks exist in the area where the water body is located, the volume of the water body in the area is calculated, and the current water body volume is compared with the second water body volume threshold to determine the stability of the tunnel structure; If the vertical deformation data reaches the second vertical threshold and the horizontal deformation data reaches the set second horizontal threshold, the tunnel structure is unstable.

[0014] A water-rich karst tunnel disaster early warning system comprises the following steps: A crack acquisition module is used to obtain the tunnel face image of the tunnel, pre-process the tunnel face image, input the pre-processed tunnel face image into the pre-trained crack recognition model, and obtain the tunnel face crack distribution data; The water body acquisition module is used to obtain the eddy current field signal formed by the pulse electromagnetic wave of the tunnel face, perform inversion analysis on the eddy current field signal, obtain the electrical distribution of the underground medium, and obtain two-dimensional images of different underground layers according to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles. The three-dimensional data of the water body in front of the tunnel face is established by using three-dimensional inversion technology and data fusion methods and combining two-dimensional images of different layers; Structural data acquisition module, used to obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; The stability module is used to determine the structural stability of the tunnel based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

[0015] Compared with the prior art, the present invention has the following beneficial technical effects: The present application proposes a method for early warning of water-rich karst tunnel disasters. The method combines a variety of monitoring methods, including face image analysis, pulse electromagnetic wave detection, and tunnel deformation and stress monitoring, and can obtain key information in the tunnel construction process in an all-round way. This comprehensive monitoring method enables the early warning system to more accurately evaluate the structural stability of the tunnel and timely discover potential geological disaster risks. By acquiring the face image, eddy current field signal, and deformation and stress data of the tunnel in real time, the method can achieve real-time monitoring and early warning. Once an abnormal situation is found, such as crack expansion, water body abnormality, or tunnel deformation beyond the safe range, the system can immediately issue an early warning to provide construction personnel with valuable escape and response time. Using pre-trained crack recognition models and three-dimensional inversion technology, the method can improve the accuracy of crack recognition and the positioning accuracy of water body distribution. At the same time, combined with a variety of data sources for comprehensive analysis, the risk of false alarms and missed alarms can be further reduced, and the reliability of the early warning system can be improved. Through data fusion technology, the method can effectively integrate data from different monitoring methods to form a comprehensive tunnel stability assessment report. At the same time, with the help of intelligent algorithms and models, the data can be deeply mined and analyzed to discover potential disaster risk points and provide a scientific basis for construction decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 This is a schematic diagram of the structure of the water-rich karst tunnel disaster early warning system of the present invention; Figure 2 It is a flow chart of the water-rich karst tunnel disaster early warning method of the present invention; Figure 3 is a structural diagram of the early warning device of the present invention; Figure 4 This is a schematic diagram of the structural flow of the crack identification instrument of the present invention; Figure 5 This is a schematic diagram of the structure flow of the water body detector of the present invention; Figure 6 This is a schematic diagram of the survey line arrangement of the water body detector of the present invention; Figure 7 It is a schematic diagram of the arrangement of the structural stress monitoring instrument and the deformation monitoring instrument of the present invention; Figure 8 It is a schematic diagram of the deformation monitoring process of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0020] See also Figure 1-8 , a water-rich karst tunnel disaster early warning method, comprising the following steps: Step 1: Acquire a tunnel face image of the tunnel, preprocess the tunnel face image, input the preprocessed tunnel face image into a pre-trained crack recognition model, and obtain tunnel face crack distribution data.

[0021] The crack distribution map of the tunnel face is mainly used to judge the stability of the surrounding rock, determine whether the surrounding rock has a tendency to deteriorate, and provide early warning for the deterioration of the surrounding rock.

[0022] S1.1. Use an industrial camera to obtain images of the tunnel.

[0023] The image is collected by a camera set up in front of the face of the tunnel. The camera is set up about ten meters away from the face of the tunnel and roughly perpendicular to the face of the tunnel to avoid squint. Then the flashlight is turned on and a water spray gun is used to reduce the dust concentration to obtain a better crack image.

[0024] S1.2. Preprocess the tunnel image to improve the image quality and the accuracy of crack identification.

[0025] The preprocessing of tunnel images includes grayscale processing, filtering processing and binarization processing.

[0026] 1. Grayscale: Due to the influence of the concrete lining on the tunnel surface and the uneven lighting, the collected images are mostly gray, and the distinction between cracks and background is small. In order to make the image present more levels of color depth, the image needs to be processed in depth. The principle is to sum the three components of RGB and then take the average value, so that the contrast of RGB can be concentrated on the grayscale value. The operation command in OpenCV is [cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)] 2. This application adopts median filtering, and its basic principle is to divide the image into several windows of equal size in the gray value space, and calculate the ratio of the gray value of each pixel point to the gray value of the pixel point in the neighborhood window in each window, and use this ratio as a gray feature vector of the pixel point, and then perform median filtering on this feature vector, and perform weighted average of the ratio of the gray value of all pixels in the neighborhood window to obtain a smoothed image. In the filtering process, noise is often added to the area where noise needs to be filtered out to achieve a better filtering effect. In this way, some pixels with unclear gray features and more noise will appear in the area where noise does not need to be filtered out, thereby affecting the image quality. Therefore, the median filtering algorithm not only has the ability to filter out noise, but also has good edge retention capabilities. The operation command in OpebCV is [cv2.medianBlur(img, 5)] In this command, the convolution kernel size used for filtering is 5*5. The larger the convolution kernel, the better the noise reduction effect, but the computer resources consumed will also increase.

[0027] 3. Binarization processing, the principle is to set a threshold, compare the gray value of the pixel in the image with the threshold, set the pixel greater than the threshold to white, and set the pixel less than the threshold to black. This process further increases the contrast between the crack area in the image and other areas, making the crack easier to identify. The operation steps in OpenCV are: use a for loop to traverse each pixel in the image, if it is greater than the threshold, directly set its gray value to 255, otherwise set the gray value to 0.

[0028] S1.3. Use the YOLOv5 model as the image recognition model.

[0029] The trained YOLOv5 model is used to identify the preprocessed image to obtain the crack image of the tunnel.

[0030] First, the preprocessed image is cropped and scaled to 640*640 to facilitate importing the trained YOLOv5 model.

[0031] The YOLOv5 model includes input layer, backbone layer, neck layer and head layer; the input layer is used to read image data, the backbone layer is used to extract feature information of the image, the neck layer is used to fuse feature information, and the head layer is used to identify cracks based on the fused feature information.

[0032] The training process of the YOLOv5 model is as follows: The captured images are randomly divided into two groups: training set and validation set according to a certain ratio. Labelme is used to make the data set labels. The specific method is to select the cracks on the image in labelme and enter the name of the label. After all labels are made, they are kept in .josn format. Secondly, since YOLOv5 does not support the .josn format, the file format needs to be converted from .josn to .txt. For the adjustment of model training parameters, open the train.py file and modify the data in this file to adjust the training parameters. The adjustable parameters include but are not limited to: initial weight, training model location, data set parameter file location, hyperparameters, number of training rounds, number of batch processing files, GPU acceleration settings, and multi-threading settings. When the model is trained, the weight file is obtained. Finally, open the detect.py file to test the effect of the model, fill in the location of the weight file, the location of the validation set, etc. If the test is not satisfactory, you can open the train.py file again to reset different parameters to adjust the model and repeat the test step. This process is repeated until the test step shows satisfactory results.

[0033] Step 2: Obtain the eddy current field signal formed by the pulse electromagnetic wave at the tunnel face, perform inversion analysis on the eddy current field signal, obtain the electrical distribution of the underground medium, and obtain two-dimensional images of different underground levels according to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles. Use three-dimensional inversion technology and data fusion methods and combine the two-dimensional images of different levels to establish three-dimensional data of the water body in front of the tunnel face.

[0034] S2.1. Use a water detector to send pulse electromagnetic waves to the tunnel face.

[0035] The measuring line is Figure 6 This arrangement method is not unique and can be adjusted according to the construction situation. This example only lists one of them.

[0036] S2.2, a pulse electromagnetic field is emitted by a transmitter, and a controller in the transmitter controls the waveform of the emission, such as a rectangular, triangular, and half-sine shape. In addition to the waveform, the emission power, emission current, and emission current cut-off time of the transmitter need to be determined according to the specific environment.

[0037] S2.3. During the first pulse electromagnetic field interval, the receiver receives the signal of the secondary eddy current field, and performs inversion analysis on the secondary eddy current electric field in these signals to obtain the electrical distribution of the underground medium. By obtaining the measurement data on the survey lines at multiple different angles, two-dimensional images at different levels can be obtained. Combining these two-dimensional images at different levels, using three-dimensional inversion technology and data fusion methods, the three-dimensional morphology and distribution characteristics of the water body ahead can be accurately reconstructed, thereby providing more detailed and comprehensive groundwater characteristics information, which helps to improve the detection accuracy and interpretation ability of complex underground structures.

[0038] S2.4, in the process of fusing two-dimensional images at different levels into three-dimensional images, Voxler software is used to process and visualize the data. Before drawing the three-dimensional slice map, Voxler software first needs to pre-process the transient electromagnetic method (TEM) inversion data of the survey area.

[0039] The specific operation steps are as follows: organize the data of the measuring points on each survey line, including the actual coordinates or relative coordinates, into a data format containing X, Y, Z and ρ (apparent resistivity), where X and Y are plane coordinates, Z is the elevation of the measuring point, and ρ is the apparent resistivity value corresponding to the X, Y, and Z coordinate points. Next, merge the sorted data of all survey lines into a complete data file, so as to construct the data body required for the Voxler software to draw the three-dimensional slice map. After starting the Voxler software, the user needs to import the sorted three-dimensional data body, and then gradually generate and display the three-dimensional slice map of the target layer by executing the relevant module operations in the software, such as interpolation algorithm settings, data gridding processing, and slice map drawing commands.

[0040] This three-dimensional slice map can intuitively display the structural characteristics and electrical distribution of each underground layer, provide more comprehensive and accurate three-dimensional morphological information of groundwater bodies or other geological bodies, and facilitate more in-depth geological analysis and decision support.

[0041] Step 3: Obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel.

[0042] S3.1. Vertical deformation data: A deformation detector is installed on the top of the tunnel to monitor the settlement deformation of the tunnel.

[0043] In this embodiment, three settlement measuring points are installed on the top of the tunnel, and the settlement deformation data of the tunnel is determined by combining the template feature matching method. The specific method is as follows: 1. Obtain the internal and external parameters of the camera, including focal length, principal point position, lens distortion coefficient, camera rotation and translation matrix, etc.

[0044] Intrinsic parameters describe the internal characteristics of the camera (such as focal length and optical center), while extrinsic parameters describe the position and orientation of the camera in the world coordinate system. Through camera calibration, the relationship between the image coordinate system and the world coordinate system can be established.

[0045] 2. Evenly arrange three settlement measuring points on the top of the tunnel to ensure that the measuring points can fully reflect the settlement conditions of the tunnel.

[0046] 3. Use a deformation detector to regularly collect image data of settlement measuring points, and perform pre-processing operations such as denoising and enhancement on the collected image data to improve image quality.

[0047] 4. Use template feature matching algorithms, such as grayscale, edge or feature point matching methods. Slide the template in the image to be matched, calculate the similarity measure between the template and the image sub-region (such as cross-correlation, mean square error, normalized cross-correlation coefficient, etc.), and select the position with the highest similarity as the matching result.

[0048] 5. Use the camera intrinsic parameter matrix to convert the matched pixel coordinates (u, v) into three-dimensional coordinates (x, y, z) in the camera coordinate system. Combined with the camera extrinsic parameters (rotation matrix R and translation vector t), convert the coordinates in the camera coordinate system into actual coordinates in the world coordinate system.

[0049] Using the camera intrinsic parameters, the pixel coordinates (u, v) are converted to coordinates (x, y, z) in the camera coordinate system. This is calculated by inverse perspective projection, and the formula is:

[0050] Among them, K is the camera intrinsic parameter matrix, and d is the depth information of the object from the camera.

[0051] Camera coordinate system to world coordinate system: Convert the camera coordinates to the world coordinates, which requires the external parameters of the camera. The formula is:

[0052] Among them, R is the rotation matrix and t is the translation vector.

[0053] Compare the actual coordinate changes of the same measuring point in the images at different time points, calculate the settlement amount (ΔX, ΔY, ΔZ), perform statistical analysis on the settlement data, draw settlement curves, and evaluate the settlement trend and stability of the tunnel.

[0054] S3.2. The horizontal deformation data of the tunnel shall be determined based on the horizontal convergence survey line.

[0055] Structural stress monitors are installed inside the primary lining of the tunnel.

[0056] The horizontal convergence survey line is arranged according to the construction method. If it is a full-section construction, only one survey line is needed. If it is a step method construction, two survey lines are needed. The schematic diagram shows the arrangement of horizontal convergence survey lines under the step method construction state.

[0057] S3.3. The tunnel structure stress data is measured using a structural force monitoring instrument, and the internal temperature data of the tunnel is obtained at the same time.

[0058] In this step, by installing structural stress monitors and deformation monitors, it is possible to provide accurate and timely basis for evaluating the feasibility of tunnel construction methods, the rationality of design parameters, and understanding the actual surrounding rock level and deformation characteristics of tunnel construction, which is of decisive significance to the construction time of the secondary lining of the tunnel; therefore, it is an important means to ensure the success of tunnel construction. The main tasks of tunnel monitoring and measurement should be to improve safety, correct design, guide construction, accumulate construction experience, and provide analysis data to the construction party, supervision party, design party and owner in a timely manner through on-site analysis and processing of measured data.

[0059] Step 4: Determine the structural stability of the tunnel based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

[0060] 1. Determine the length and width of each crack based on the crack distribution data of the tunnel face. When the total area of ​​the cracks reaches a certain threshold, the tunnel structure is unstable and an early warning message is issued.

[0061] 2. Determine the volume of the water body ahead based on the three-dimensional data of the water body. When the volume of the water body reaches the threshold, the tunnel structure is unstable and an early warning message is issued.

[0062] 3. Combine the crack data with the water body property data for analysis. If there are cracks in the area where the water body is located, calculate the water volume in the area. When the water volume reaches the set threshold, issue an early warning message.

[0063] 4. When the vertical deformation data reaches the set first vertical threshold, an early warning is issued to the monitor.

[0064] When the horizontal deformation data reaches the set first horizontal threshold, an early warning is issued to the monitor.

[0065] When the vertical deformation data reaches the second vertical threshold and the horizontal deformation data reaches the set second horizontal threshold, an early warning is issued to the monitor.

[0066] The first vertical threshold is greater than the second vertical threshold, and the first horizontal threshold is greater than the second horizontal threshold.

[0067] After an early warning is issued, the staff can start the emergency treatment device, which consists of a drainage pump and a mud interception net, mainly for water inrush and mud in the tunnel. If a water inrush occurs, the pump can quickly pump water out of the tunnel. If a mud inrush occurs, the mud interception net can block stones and mud outside the net to prevent mud from entering the pump and affecting its working efficiency. Secondly, when other emergencies occur in the tunnel, the personnel in the tunnel can also contact the staff through the active alarm device for processing.

[0068] This method is applicable to different types of water-rich karst tunnels. No matter how complex the geological conditions of the tunnel are, the monitoring parameters and models can be adjusted to suit the actual situation. This strong adaptability enables the method to play a role in a wide range of tunnel construction projects. The method not only focuses on early warning and prevention of disasters, but also provides strong support for the management and repair of tunnels through real-time monitoring data. Once signs of disasters are found, measures can be taken quickly to manage them and prevent the disaster from further expanding and spreading.

[0069] In summary, the water-rich karst tunnel disaster early warning method has the advantages of comprehensiveness and integration, real-time monitoring and early warning, high precision and reliability, strong adaptability, data fusion and intelligence, and equal emphasis on prevention and control, which provides a strong guarantee for the safety and stability of tunnel construction.

[0070] Based on the above-mentioned water-rich karst tunnel disaster warning method, the present application also proposes a water-rich karst tunnel disaster warning system, including a multi-sensor array, a data collection center, a 3GPP R16 wireless communication module and a central processor, aiming to achieve continuous and accurate monitoring and timely warning of the safety status of the tunnel structure.

[0071] The multi-sensor array is configured to be deployed inside the tunnel and is responsible for collecting multi-dimensional physical quantity data including crack characteristics, water properties, stress distribution, structural deformation and ambient temperature. Sensor types include crack identification instruments, water detectors, stress and deformation monitors, and temperature monitors to ensure the comprehensiveness and accuracy of data collection.

[0072] The data collection center is configured to integrate a high-performance analog-to-digital converter (such as ADC0832) to convert the analog signal output by the sensor into a digital signal for subsequent data processing and analysis.

[0073] The wireless transmission module is configured to: utilize advanced 3GPP R16 wireless communication technology to achieve real-time and reliable data transmission between the data acquisition module and the central processing unit, thereby ensuring the continuity and integrity of the data.

[0074] The central processor is configured to: as the core of the system, receive, integrate and analyze data from various sensors, use algorithm models to evaluate the safety status of the tunnel, and output risk types and levels. When potential danger is detected and the risk level exceeds the preset threshold, the alarm mechanism is automatically triggered and remote control of emergency response equipment is allowed.

[0075] The crack identification instrument uses advanced image processing technology (such as grayscale, filtering, and binarization) to remove image noise and improve image quality. Combined with deep learning algorithms (such as YOLOv5), it realizes automatic identification and quantitative analysis of cracks and evaluates the impact of cracks on tunnel structures.

[0076] The water body detector is based on the principle of electromagnetic induction. It sends pulsed electromagnetic fields and receives secondary eddy current field signals, and combines inversion algorithms to construct a three-dimensional image of the water body, providing an intuitive basis for water hazard risk assessment.

[0077] The system workflow includes four main links: data collection, preprocessing, analysis and evaluation, and early warning response. The central processor integrates multi-source information such as crack data, water properties, stress distribution, structural deformation and temperature, and uses complex risk assessment algorithms for comprehensive analysis to determine the safety risk level of the tunnel. When the risk level exceeds the preset threshold, the system automatically triggers an alarm, and the staff can start the emergency treatment device through remote control.

[0078] Correspondingly, the present application also proposes a water-rich karst tunnel disaster early warning system, comprising the following steps: A crack acquisition module is used to obtain the tunnel face image of the tunnel, pre-process the tunnel face image, input the pre-processed tunnel face image into the pre-trained crack recognition model, and obtain the tunnel face crack distribution data; The water body acquisition module is used to obtain the eddy current field signal formed by the pulse electromagnetic wave of the tunnel face, perform inversion analysis on the eddy current field signal, obtain the electrical distribution of the underground medium, and obtain two-dimensional images of different underground layers according to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles. The three-dimensional data of the water body in front of the tunnel face is established by using three-dimensional inversion technology and data fusion methods and combining two-dimensional images of different layers; Structural data acquisition module, used to obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; The stability module is used to determine the structural stability of the tunnel based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

[0079] It should be noted that in the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each module is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules can be combined or integrated into another device, or some features can be ignored or not executed. The module described as a separate component may or may not be physically separated. The component displayed as a module may be a physical unit or multiple physical units, that is, it may be located in one place, or it may be distributed in multiple different places. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0080] In addition, each module in each embodiment of the present invention may be integrated into a processing unit, each module may exist physically separately, or two or more modules may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0081] An electronic device provided in an embodiment of the present application includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of a water-rich karst tunnel disaster warning method described in any of the above embodiments are implemented.

[0082] Another electronic device provided in the embodiment of the present application may also include: an input port connected to the processor, used to transmit multimodal data collected by an external acquisition device to the processor; and a display unit connected to the processor, used to display the processing results of the processor to the outside world; a communication module connected to the processor, used to realize the communication between the electronic device and the outside world. The display unit can be a display panel, a laser scanning display, etc.; the communication mode adopted by the communication module includes but is not limited to mobile high-definition link technology (HML), universal serial bus (USB), high-definition multimedia interface (HDMI), wireless connection (including wireless fidelity technology (WiFi), Bluetooth communication technology, low-power Bluetooth communication technology, and communication technology based on IEEE802.11s).

[0083] An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of a water-rich karst tunnel disaster warning method described in any of the above embodiments are implemented.

[0084] For the description of the relevant parts of the water-rich karst tunnel disaster warning system, electronic device and computer-readable storage medium provided in the embodiments of the present application, please refer to the detailed description of the corresponding parts in the water-rich karst tunnel disaster warning method provided in the embodiments of the present application, which will not be repeated here. In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive elaboration.

[0085] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A water-rich karst tunnel disaster early warning method, characterized in that: The following steps are involved: Acquire the tunnel face image of the tunnel, preprocess the tunnel face image, input the preprocessed tunnel face image into the pre-trained crack recognition model, and obtain the tunnel face crack distribution data; The eddy current field signal formed by the pulse electromagnetic wave at the tunnel face is obtained, and the eddy current field signal is inverted and analyzed to obtain the electrical distribution of the underground medium. According to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles, two-dimensional images of different underground layers are obtained. The three-dimensional data of the water body in front of the tunnel face is established by using three-dimensional inversion technology and data fusion methods and combining two-dimensional images of different layers. Obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; The structural stability of the tunnel is determined based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

2. A water-rich karst tunnel disaster early warning method according to claim 1, characterized in that: The preprocessing of the tunnel face image includes grayscale processing, filtering processing and binarization processing.

3. A water-rich karst tunnel disaster early warning method according to claim 1, characterized in that: The crack recognition model is a pre-trained YOLOv5 model; The YOLOv5 model includes an input layer, a backbone layer, a neck layer and a head layer; The input layer is used to read image data, the backbone layer is used to extract feature information of the image, the neck layer is used to fuse the feature information, and the head layer is used to identify cracks according to the fused feature information.

4. A water-rich karst tunnel disaster early warning method according to claim 1, characterized in that: A water body detector is used to transmit pulse electromagnetic waves at different angles to the tunnel face, and the signal of the secondary eddy current field is received by the receiver.

5. The method for early warning of water-rich karst tunnel disaster according to claim 1, characterized in that: The method for obtaining the vertical deformation data comprises: Three settlement measuring points are installed on the top of the tunnel, and the settlement deformation data of the tunnel is determined based on the images of the settlement measuring points and combined with the template feature matching method.

6. A water-rich karst tunnel disaster early warning method according to claim 5, characterized in that: The method for determining the settlement deformation data is as follows: The image data of the settlement measuring points are collected, and preprocessing operations such as denoising and enhancement are performed on the collected image data. The template feature matching algorithm is used to slide the template in the image to be matched, and the similarity measurement between the template and the image sub-region is calculated. The position with the highest similarity is selected as the matching result, and the matched pixel coordinates are converted into three-dimensional coordinates in the camera coordinate system. The coordinates in the camera coordinate system are converted into actual coordinates in the world coordinate system by combining the camera extrinsic parameters. The actual coordinate changes of the same measuring point in the images at different time points are compared to obtain the settlement deformation data.

7. The method for early warning of water-rich karst tunnel disaster according to claim 1, characterized in that: The method for determining the horizontal deformation data is as follows: According to the construction method, a horizontal convergence survey line is set in the tunnel, and the horizontal deformation data of the tunnel is obtained according to the horizontal convergence survey line; The method for determining the tunnel structure stress data is as follows: A structural stress monitor is set inside the initial support of the tunnel, and the tunnel structure stress data is obtained based on the structural stress monitor.

8. The method for early warning of water-rich karst tunnel disaster according to claim 1, characterized in that: The structural stability of the tunnel is determined as follows: Determine the length and width of each crack based on the crack distribution data on the tunnel face, and determine the stability of the tunnel structure based on the comparison between the total area of ​​all cracks and the threshold value; The volume of the water body ahead is determined based on the three-dimensional data of the water body, and the stability of the tunnel structure is determined based on a comparison result between the volume of the water body and a first water body volume threshold; Determining the stability of the tunnel structure based on a comparison result of the vertical deformation data and a corresponding first vertical threshold value; The stability of the tunnel structure is determined based on a comparison result between the horizontal deformation data and a first horizontal threshold.

9. A water-rich karst tunnel disaster early warning method according to claim 8, characterized in that: The structural stability of the tunnel is determined as follows: When cracks exist in the area where the water body is located, the volume of the water body in the area is calculated, and the current water body volume is compared with the second water body volume threshold to determine the stability of the tunnel structure; If the vertical deformation data reaches the second vertical threshold and the horizontal deformation data reaches the set second horizontal threshold, the tunnel structure is unstable.

10. A water-rich karst tunnel disaster early warning system, characterized in that: The following steps are involved: A crack acquisition module is used to obtain the tunnel face image of the tunnel, pre-process the tunnel face image, input the pre-processed tunnel face image into the pre-trained crack recognition model, and obtain the tunnel face crack distribution data; The water body acquisition module is used to obtain the eddy current field signal formed by the pulse electromagnetic wave of the tunnel face, perform inversion analysis on the eddy current field signal, obtain the electrical distribution of the underground medium, and obtain two-dimensional images of different underground layers according to the electrical distribution of the underground medium corresponding to the pulse electromagnetic waves at different angles. The three-dimensional data of the water body in front of the tunnel face is established by using three-dimensional inversion technology and data fusion methods and combining two-dimensional images of different layers; Structural data acquisition module, used to obtain the vertical deformation data, horizontal deformation data and tunnel structure stress data of the tunnel; The stability module is used to determine the structural stability of the tunnel based on the crack distribution data of the tunnel face, the three-dimensional data of the water body, the vertical deformation data of the tunnel, the horizontal deformation data and the tunnel structure stress data.

Images