A safety monitoring method and system for construction of a city regional railway

Abstract

The present application relates to the technical field of electric digital data processing, in particular to a safety monitoring method and system for city railway construction, wherein the safety monitoring method for city railway construction comprises: acquiring a three-dimensional geometric model, a thermal infrared image and stress data of a city railway construction tunnel; acquiring a construction influence degree of a construction operation on a corresponding position of a stress sensor; acquiring a construction influence degree of the construction operation on all positions in the three-dimensional geometric model; determining a maximum value point of the influence degree; determining a suspected crack point; and determining a crack region. The present application comprehensively uses the data collected by the three-dimensional geometric model, the thermal infrared image and the stress sensor to realize comprehensive perception of the tunnel construction condition, and through interpolation reconstruction of the stress field and construction of a stress distribution cloud image, the potential stress concentration position can be identified, then combined with the thermal infrared image for spatial matching and correlation verification, early accurate positioning and risk assessment of the crack are realized, and the monitoring timeliness is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric digital data processing, and particularly relates to a safety monitoring method and system for construction of a city railway. BACKGROUND

[0002] Tunnel safety monitoring is crucial in the construction of a city railway and is a core link for ensuring project safety, quality and progress. The existing method is to install stress sensors in different areas inside the tunnel to monitor the stress changes in each area. If the stress change in a certain area exceeds a fixed threshold, it is determined that there may be a crack in the area.

[0003] The existing problems are: if the crack position is far from the sensor, the sensor may not be able to monitor the existence of the crack in time due to the attenuation and delay effect of stress transmission; or when the sensor monitors a large stress change, the crack may have expanded to a large scale, thereby possibly causing tunnel structure instability, collapse and other dangers, resulting in serious consequences. SUMMARY

[0004] The present application provides a safety monitoring method and system for construction of a city railway to solve the existing problems.

[0005] The safety monitoring method for construction of a city railway provided by the present application adopts the following technical scheme:

[0006] One embodiment of the present application provides a safety monitoring method for construction of a city railway, comprising: acquiring a three-dimensional geometric model of a city railway construction tunnel, a thermal infrared image and stress data collected by stress sensors arranged in the city railway construction tunnel; based on the stress data, acquiring a construction influence degree of construction work on a position corresponding to the stress sensor; based on the position of the stress sensor in the three-dimensional geometric model and the construction influence degree of construction work on the corresponding area of the stress sensor, acquiring the construction influence degree of construction work on all positions in the three-dimensional geometric model; determining a maximum influence degree point in the three-dimensional geometric model; based on the maximum influence degree point, determining a suspected crack point; based on the thermal infrared image and the suspected crack point, determining a crack area.

[0007] Further, the stress data is used to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor, including: obtaining a stress reference value and a stress increase trend index of the position corresponding to the stress sensor for each stress sensor; the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset period; calculating the ratio of the stress reference value to the maximum stress reference value of all stress reference values corresponding to the stress sensor; calculating the product of the ratio and the increasing trend index to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor.

[0008] Further, the stress increase trend index is obtained, including: for the stress data collected by each stress sensor, taking the time number as the horizontal coordinate and the stress data corresponding to each time number as the vertical coordinate to obtain a plurality of coordinate points; performing principal component analysis on all coordinate points to obtain a plurality of projection values and their corresponding projection vectors; the projection value is used to represent the projection length of the coordinate point in the direction of the corresponding projection vector; calculating the ratio of the vertical coordinate to the horizontal coordinate of the projection vector corresponding to the maximum projection value; calculating the normalized value of the inverse tangent angle of the ratio to obtain the stress increase trend index at the corresponding position.

[0009] Further, the stress data is used to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor, including: obtaining a stress reference value and a stress increase trend index of the position corresponding to the stress sensor for each stress sensor; the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset period; calculating the ratio of the stress reference value to the maximum stress reference value of all stress reference values corresponding to the stress sensor; calculating the product of the ratio and the increasing trend index to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor.

[0010] Further, the stress data is used to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor, including: obtaining a stress reference value and a stress increase trend index of the position corresponding to the stress sensor for each stress sensor; the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset period; calculating the ratio of the stress reference value to the maximum stress reference value of all stress reference values corresponding to the stress sensor; calculating the product of the ratio and the increasing trend index to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor.

[0011] Further, the determining the suspected crack point based on the influence degree maximum point comprises: taking the construction influence degree of all positions in the three-dimensional geometric model as a label, obtaining multiple communities and a unified label of all nodes in each community by using a community detection algorithm based on a preset label propagation rule; the preset label propagation rule is that each node selects a label having the largest difference with a reference node label value from adjacent nodes to transform the label when transforming the label through the adjacent nodes; obtaining a source node of each community; the source node is used to represent a construction position, and the source node is a node in the community whose initial label is the unified label; obtaining an influence degree prediction value at the influence degree maximum point based on the source node; obtaining a difference absolute value between the construction influence degree and the influence degree prediction value of the influence degree maximum point; calculating a ratio of the difference absolute value to the corresponding influence degree prediction value; obtaining a suspected crack point based on the ratio; the suspected crack point is the influence degree maximum point whose ratio is greater than a preset ratio.

[0012] Further, the obtaining the influence degree prediction value at the influence degree maximum point based on the source node comprises: for each community, establishing a regression model based on the construction influence degree of all non-influence degree maximum points in the community and distances between the non-influence degree maximum points and the source node; taking the influence degree maximum point in the community and the distance between the influence degree maximum point and the source node as inputs of the regression model to obtain the influence degree prediction value of the influence degree maximum point.

[0013] Further, the determining the crack region based on the thermal infrared image and the suspected crack point comprises: performing feature point matching on the three-dimensional geometric model and the thermal infrared image to obtain a thermal point corresponding to each suspected crack point; the thermal point is a corresponding position point of the suspected crack point on the thermal infrared image; obtaining a high-temperature pixel point in the thermal infrared image; the high-temperature pixel point is a pixel point having a temperature higher than an average temperature of the whole image in the thermal infrared image; obtaining multiple high-temperature regions by region growing based on the high-temperature pixel point; obtaining a suspected crack region where the suspected crack point is located; the suspected crack region is obtained by region expansion on a target point; the target point is a point in a community where the suspected crack region is located and greater than a community influence degree average value; for each suspected crack point, calculating an intersection-over-union of the suspected crack region and the high-temperature region; determining a crack region based on the intersection-over-union; the crack region is a suspected crack region whose intersection-over-union is greater than a preset intersection-over-union threshold.

[0014] Further, after the crack region is determined based on the thermal infrared image and the suspected crack point, the method comprises: acquiring a crack parameter of the crack region; acquiring a risk level of the crack region based on the crack parameter; acquiring a processing scheme for the crack region based on the risk level; and periodically acquiring a structural health condition of the crack region after the crack region is processed.

[0015] The safety monitoring system for the construction of the city railway adopts the technical scheme as follows:

[0016] An embodiment of the present application provides a safety monitoring system for the construction of a city railway, comprising: an upper computer and a distributed stress sensor network, a thermal infrared sensor and a three-dimensional laser scanner arranged in a city railway construction tunnel, wherein the distributed stress sensor network, the thermal infrared sensor and the three-dimensional laser scanner are in communication connection with the upper computer.

[0017] The distributed stress sensor network is used for collecting stress data of the city railway construction tunnel.

[0018] The thermal infrared sensor is used for collecting thermal infrared images of the city railway construction tunnel.

[0019] The three-dimensional laser scanner is used for collecting full-section point cloud data of the city railway construction tunnel.

[0020] The upper computer is used for acquiring a three-dimensional geometric model of the city railway construction tunnel based on the full-section point cloud data, and is used for acquiring the three-dimensional geometric model, the thermal infrared images and stress data collected by each stress sensor in the distributed stress sensor network; based on the stress data, acquiring a construction influence degree of construction work on the corresponding position of the stress sensor; based on the position of the stress sensor in the three-dimensional geometric model and the construction influence degree of construction work on the corresponding region of the stress sensor, acquiring the construction influence degree of construction work on all positions in the three-dimensional geometric model; determining a maximum influence degree point in the three-dimensional geometric model; based on the maximum influence degree point, determining a suspected crack point; and based on the thermal infrared image and the suspected crack point, determining a crack region.

[0021] The technical scheme of the present application has the following beneficial effects:

[0022] In the embodiment of the present application, the stress change data of different regions in the construction process is collected in real time through the distributed stress sensor network arranged in the key area of the tunnel; then the stress field between different sensors is interpolated with high precision, the stress distribution of the whole region of the tunnel is obtained, and the potential stress concentration position is identified. Finally, combined with the tunnel surface temperature field distribution data obtained by the infrared thermal imager, the stress concentration area and the temperature abnormal area in the thermal infrared image are spatially matched and correlated through the multi-physical field coupling analysis method. Thus, the present application realizes the comprehensive perception of the tunnel construction condition by comprehensively using the data collected by the three-dimensional geometric model, the thermal infrared image and the stress sensor, and through the interpolation reconstruction of the stress field and the construction of the stress distribution cloud picture, the potential stress concentration position can be identified, and then the spatial matching and correlation verification are carried out in combination with the thermal infrared image, the early accurate positioning and risk assessment of the crack are realized, and the timeliness of the monitoring is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0024] Figure 1 The flowchart of the safety monitoring method for the construction of the city railway provided by the embodiment of the present application is shown in the figure.

[0025] Figure 2 The schematic diagram of the initial node state before the SLPA label propagation provided by the embodiment of the present application is shown in the figure.

[0026] Figure 3 The schematic diagram of the final node state after the SLPA label propagation provided by the embodiment of the present application is shown in the figure.

[0027] Figure 4 The structural schematic diagram of the safety monitoring system for the construction of the city railway provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0028] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific implementation, structure, features and effects of the safety monitoring method and system for the construction of the city railway according to the present application are described in detail as follows by combining with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0030] The following will specifically describe the specific scheme of the safety monitoring method and system for city railway construction provided by the present application in combination with the drawings.

[0031] Please refer to Figure 1 which shows a safety monitoring method for city railway construction provided by an embodiment of the present application, which method comprises:

[0032] Step S110: obtaining a three-dimensional geometric model of the city railway construction tunnel, a thermal infrared image, and stress data collected by stress sensors arranged in the city railway construction tunnel.

[0033] It should be noted that the above-mentioned stress data can be obtained by stress sensors arranged in the city railway construction tunnel. The stress sensor is a sensor that can sense and measure the stress of an object. The embodiment of the present application can use a fiber optic stress sensor or a vibrating wire stress sensor to obtain the stress data. The fiber optic stress sensor is based on the principle of fiber optics. When the measured object is deformed by stress, the optical properties of the optical fiber such as the optical path length will change. By measuring the change of these optical properties, the size and direction of the stress can be calculated. It has the advantages of anti-electromagnetic interference, high precision, corrosion resistance, etc. The vibrating wire stress sensor mainly consists of a vibrating wire and an elastic element. When the elastic element is deformed by stress, the vibration frequency of the vibrating wire will change. By measuring the vibration frequency of the vibrating wire, the stress of the measured object can be calculated. Its characteristics are high measurement accuracy, stable performance, and strong adaptability. The stress sensor can be arranged at every 5-20m along the tunnel vault, side wall and other key parts, and the high-risk area can be encrypted to 3-5m. Stress data is collected every minute during construction.

[0034] It should be noted that the above-mentioned thermal infrared image can be obtained by a thermal infrared sensor arranged in the city railway construction tunnel. The thermal infrared sensor is a device that can detect the infrared radiation emitted by an object and convert it into temperature data. A fixed thermal infrared sensor can be installed every 50-100m in the city railway construction tunnel, and a mobile inspection device can be used to scan the whole tunnel at least once a day, focusing on monitoring the areas corresponding to stress anomalies. The mobile inspection device can move on the track or be carried by the staff for inspection.

[0035] It should be noted that the three-dimensional geometric model described above can be obtained based on tunnel full-section high-precision point cloud data, and the tunnel full-section high-precision point cloud data can be obtained by a three-dimensional laser scanner, which is a measuring device capable of quickly and accurately obtaining three-dimensional coordinate data of an object surface. The three-dimensional laser scanner can quickly obtain the high-precision point cloud data of the full section of the tunnel, and the data precision can reach millimeter level. Through these point cloud data, the three-dimensional geometric shape of the tunnel can be accurately reconstructed. It can be understood that the scheme of constructing a three-dimensional geometric model based on point cloud data is a known technology, and the specific implementation mode can be referred to related technologies, and the embodiment of the present application will not be described again.

[0036] In the process of tunnel construction, when cracks appear in surrounding rock or supporting structure, due to the interruption of material continuity, the stress of the surrounding area will be concentrated and transferred to the crack, resulting in a significant stress concentration zone at the crack tip. The stress concentration zone has two characteristics: one is that the stress amplitude is significantly higher than the surrounding area, and the other is that the stress change rate is rapidly accelerated. This concentration effect will continue to strengthen with the construction disturbance. Based on the two characteristics, the embodiment of the present application determines the crack area by the following steps S120 to S160.

[0037] Step S120: Based on the stress data, the construction influence degree of the construction operation on the position corresponding to the stress sensor is obtained.

[0038] Preferably, in an embodiment of the present application, the above step S120 can include: for each stress sensor, obtaining a stress reference value at the position corresponding to the stress sensor and a stress increasing trend index; wherein the stress increasing trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor in a preset period; calculating the ratio of the stress reference value to the maximum reference value among all stress reference values corresponding to the stress sensor; calculating the product of the ratio and the increasing trend index to obtain the construction influence degree of the construction operation on the position corresponding to the stress sensor. For example, the implementation mode is as follows:

[0039] For each stress sensor, the stress reference value at the position corresponding to the stress sensor and the stress increasing trend index are calculated; wherein the average value of the stress data within ten minutes can be taken as the stress reference value at the position;

[0040] The ratio of the stress reference value corresponding to each stress sensor to the maximum value among all stress reference values is calculated, and the product of the ratio and the trend value is taken as the construction influence degree. The construction influence degree represents the influence of the construction operation on the position corresponding to the stress sensor. The greater the construction influence degree, the greater the influence of the construction operation on the position.

[0041] Preferably, in one embodiment of the present application, the stress increasing trend index can be obtained by: for the stress data collected by each stress sensor, taking the time number as the horizontal coordinate and the stress data corresponding to each time number as the vertical coordinate to obtain a plurality of coordinate points; performing principal component analysis on all the coordinate points to obtain a plurality of projection values and corresponding projection vectors; wherein the projection value is used to represent the projection length of the coordinate point in the direction of the corresponding projection vector; calculating the ratio of the vertical coordinate to the horizontal coordinate of the projection vector corresponding to the maximum projection value; calculating the normalized value of the inverse tangent angle of the ratio to obtain the stress increasing trend index at the corresponding position. For example:

[0042] For the current monitoring time point, the stress data within ten minutes at the position corresponding to each stress sensor is obtained, and then each time is numbered according to 1s-600s, and the stress data corresponding to each number can be obtained;

[0043] For each stress sensor, the time number is taken as the horizontal coordinate and the stress value corresponding to each time number is taken as the vertical coordinate to obtain a plurality of coordinate points, and all the coordinate points are input into the principal component analysis (PCA) algorithm to obtain a plurality of two-dimensional projection vectors and the projection value corresponding to each projection vector. The projection value is a scalar value representing the projection length of all coordinate points in the direction corresponding to the projection vector.

[0044] The ratio of the vertical coordinate to the horizontal coordinate of the projection vector corresponding to the maximum projection value is taken as the inverse tangent angle value of the ratio to 90° as the stress increasing trend index,

[0045] The greater the trend value is, the faster the stress at the position corresponding to the stress sensor increases, the greater the stress change is, and the more likely it is that the position is a stress concentration area caused by a crack.

[0046] It should be noted that the principal component analysis algorithm, i.e., the PCA algorithm, can project high-dimensional data into a low-dimensional space while retaining as much information as possible of the original data. In the above scheme, each coordinate point includes two dimensions of horizontal and vertical, and the PCA can project these high-dimensional data into a new low-dimensional space to extract the main change trend, so that the overall upward trend of the data can be more clearly understood. In addition, the PCA algorithm can find the direction with the maximum variance in the data, i.e., the principal component, which represents the main change trend of the data. It can be understood that the PCA algorithm is a relatively mature known technology, and its specific implementation mode can be referred to related technologies, and the embodiments of the present application will not be described again.

[0047] Step S130: Obtain the construction influence degree of the construction operation on all positions in the three-dimensional geometric model based on the positions of the stress sensors in the three-dimensional geometric model and the construction influence degree of the construction operation on the corresponding regions of the stress sensors.

[0048] Preferably, in an embodiment of the present application, the above step S130 can include: obtaining the coordinate position of each stress sensor in the three-dimensional geometric model based on the position of the stress sensor; and obtaining the construction influence degree of all positions in the three-dimensional geometric model based on the coordinate position of each stress sensor in the three-dimensional geometric model by using the Kriging interpolation method. This implementation, for example, takes into account that the stress sensors are discretely distributed and cannot comprehensively represent the influence degree of the construction operation on each position in the tunnel. Therefore, the influence degree of each position can be obtained by using the interpolation method, and then further analysis and calculation are performed.

[0049] The coordinate position of each stress sensor in the three-dimensional geometric model can be obtained by using the position sensor on the stress sensor, and the construction influence degree of the corresponding position of the different stress sensors in the three-dimensional geometric model can be obtained.

[0050] By using the Kriging interpolation method, the construction influence degree of each position in the three-dimensional geometric model can be obtained based on the construction influence degree of the different stress sensors in the three-dimensional geometric model, that is, the influence degree of each position on the continuous surface subjected to the construction operation is obtained.

[0051] It should be noted that the above Kriging interpolation is a spatial interpolation method based on geographic statistics, which is used to estimate the value of a regionalized variable at an unsampled position. It predicts the value of an unknown position by using the values of known sample points and the spatial correlation structure, and provides an accuracy estimate of the prediction. It can be understood that the interpolation scheme using the Kriging interpolation method is a relatively mature known technology, and the specific implementation manner thereof can be understood by referring to related technologies, and thus will not be described herein.

[0052] Step S140: Determine the maximum influence degree point in the three-dimensional geometric model.

[0053] Preferably, in one embodiment of the present application, the step S140 can include: obtaining a spatial gradient of each position in the three-dimensional geometric model based on the construction influence degree of all positions in the three-dimensional geometric model; calculating a Hessian matrix eigenvalue of each position based on the spatial gradient; and obtaining an influence degree maximum value point based on the Hessian matrix eigenvalue, wherein the influence degree maximum value point is a point in the three-dimensional geometric model with a negative Hessian matrix eigenvalue. For example, based on the construction influence degree of all positions in the three-dimensional geometric model, the spatial gradient of each position is calculated, and then the Hessian matrix eigenvalue of each position is calculated, and the position with a negative eigenvalue is recorded as the influence degree maximum value point.

[0054] It should be noted that the spatial gradient is a vector representing the rate and direction of change of a function at a given point along each coordinate axis. In the above scheme, the function is the construction influence degree of each position in the three-dimensional geometric model. By calculating the spatial gradient of each position, the change trend and direction of the construction influence degree in the three-dimensional space can be determined. The magnitude of the spatial gradient reflects the degree of change in the construction influence degree, and the direction of the spatial gradient points to the direction in which the construction influence degree increases most rapidly. This helps to identify areas where the construction influence degree changes significantly, which may be potential stress concentration areas. The Hessian matrix is a square matrix composed of the second-order partial derivatives of a function, which describes the curvature information of the function at a point, i.e., the concave-convexity of the function in each direction. The eigenvalues of the Hessian matrix provide information about the curvature, and the sign and magnitude of the eigenvalues can help determine whether the position is a local maximum or minimum point. In the above scheme, by calculating the Hessian matrix of the construction influence degree of each position, the change characteristics of the influence degree can be further analyzed. The position with a negative eigenvalue is recorded as the influence degree maximum value point because at these points the function exhibits a local maximum value, i.e., around these points the construction influence degree converges from all directions to the point, forming a peak. The spatial gradient and the Hessian matrix are successive levels of analysis. First, the spatial gradient determines the direction and area where the influence degree changes significantly, and then the Hessian matrix further analyzes the curvature characteristics of these areas to more accurately identify the influence degree maximum value point. The spatial gradient provides a direction, and the Hessian matrix provides a verification, and finally the position of the influence degree maximum value point is determined. It can be understood that the calculation methods of the spatial gradient and the Hessian matrix eigenvalue are relatively mature and well-known techniques, and their specific implementation methods can be found in related technologies, which will not be described in detail in the present application.

[0055] Step S150: determining a suspected crack point based on the influence degree maximum value point.

[0056] It is required to be explained: some influence degree maximum points are due to the structure of the tunnel itself, not the influence of the crack, for example: the junction of soft and hard rock, the area where the rock layer inclination changes suddenly, and the uneven pressure transmission of surrounding rock, which leads to local stress concentration, so these points need to be removed from the influence maximum points. In the process of tunnel construction, stress influence usually decays radially with the construction point as the center, but the inherent structural characteristics of the tunnel (such as lithological interface, tectonic weak zone) and the construction-induced cracks will form abnormal stress concentration areas. Compared with the inherent stress concentration of stable structure, the stress distribution in the new crack area shows more significant nonlinear characteristics: when the stress-distance decay model established based on the normal section is used for prediction, the measured stress value in the crack area will deviate significantly from the predicted value, while the deviation in the structural stress concentration area is relatively small. Based on this, the following scheme is provided:

[0057] Preferably, in an embodiment of the present application, the above step S150 can include: tagging the construction influence degree of all positions in the three-dimensional geometric model, obtaining multiple communities and the unified label of all nodes in each community based on a preset label propagation rule using a community detection algorithm; wherein the preset label propagation rule is: when each node is converted through adjacent nodes, the label with the largest difference from the reference node label value is selected from the adjacent nodes for conversion; obtaining the source node of each community; wherein the source node is used to represent the construction position, and the source node is the node in the community whose initial label is the unified label; obtaining the influence degree prediction value at the influence degree maximum point based on the source node; obtaining the absolute value of the difference between the construction influence degree and the influence degree prediction value of the influence degree maximum point; calculating the ratio of the absolute value of the difference and the corresponding influence degree prediction value; obtaining the suspected crack point based on the ratio; wherein the suspected crack point is the influence degree maximum point with a ratio greater than a preset ratio. For example, in the construction process of a city railway, multiple working faces are often constructed at the same time, and this embodiment obtains multiple community structures by the idea of SLPA (Speaker-Listener Label Propagation Algorithm) label propagation, each community structure representing multiple nodes with progressive relationship in influence propagation, and the tunnel area corresponding to being greatly affected by the same working face;

[0058] The SLPA label propagation method is shown in Figure 2 and Figure 3 , the Figure 2 is the initial state of multiple nodes, each node corresponds to a label, and the label of each node is converted into the label of the adjacent node through the label with the highest frequency or the most similar label in the adjacent node. After multiple rounds of conversion, multiple communities in Figure 3 are obtained, and the labels of nodes in the same community are the same. Figure 2 andFigure 3 The numbers below each node are the labels corresponding to the nodes, as shown in the initial state of the plurality of nodes are different. Figure 2 Figure 3 As shown in the final node state after SLPA label propagation, nodes A, B and D have the same label after transformation and belong to the same community; nodes C and F have the same label after transformation and belong to the same community; nodes G, H and I have the same label after transformation and belong to the same community, and node E is a community.

[0059] In this embodiment, the construction influence degree of each position on the three-dimensional geometric model can be taken as a label.

[0060] In addition, in order to facilitate finding the construction source of each working face, the label transformation rule of each node can be set as: when each node transforms the label through an adjacent node, the label of the node with the largest difference in label value from the reference node is selected from the adjacent nodes for transformation, and then the influence source can be gradually found through continuous transformation of the label.

[0061] Through SLPA label propagation, a plurality of communities and a unified label of nodes in each community can be obtained. The node with the initial node label as the unified label is taken as a source node in each community. The source node indicates a construction site. The construction at the site causes an influence on the regions corresponding to the nodes in the community, and then stress changes occur.

[0062] Based on the source node, an influence degree prediction value at the influence degree maximum point is obtained.

[0063] The absolute value of the difference between the influence degree prediction value and the actual influence degree of each influence degree maximum point in each community is calculated, and the ratio of the absolute value to the corresponding influence degree prediction value is calculated. The influence degree maximum point with a ratio greater than 0.6 is retained and recorded as a suspected crack point.

[0064] It should be noted that the above SLPA algorithm is a community monitoring algorithm based on label propagation. Each node corresponds to a label at the beginning. In multiple rounds of propagation, the node updates its own label according to the label of the adjacent node. Through the SLPA label propagation algorithm, a plurality of communities and a unified label of nodes in each community can be obtained, so as to identify the construction source node. The source node indicates a construction site. The construction at the site causes an influence on the regions corresponding to the nodes in the community, and then stress changes occur. It can be understood that the SLPA algorithm is a relatively mature known technology, and its specific implementation mode can be referred to related technologies. The embodiments of the present application will not be described again.

[0065] ​Preferably, in one embodiment of the present application, the source node-based obtaining of the influence degree prediction value at the influence degree maximum point comprises: for each community, establishing a regression model based on the construction influence degrees of all non-influence degree maximum points in the community and the distances between the non-influence degree maximum points and the source node; and inputting the influence degree maximum points in the community and the distances between the influence degree maximum points and the source node into the regression model to obtain the influence degree prediction value of the influence degree maximum point.

[0066] It should be noted that the regression model is a mathematical modeling method for analyzing the relationship between the influence degree maximum points in the community and the source node and screening out suspected crack points. The input variables of the regression model are the influence degree maximum points in the community and the distances between the influence degree maximum points and the source node, and the output variable is the influence degree prediction value of the influence degree maximum point. The regression model can be established by: for each community, establishing a regression model based on the construction influence degrees of all non-influence degree maximum points in the community and the distances between the non-influence degree maximum points and the source node. By fitting the data of the construction influence degrees of the known non-influence degree maximum points in the community and the distances between the non-influence degree maximum points and the source node, a mathematical relationship model between the distance and the construction influence degree is established, so that the influence degree can be predicted according to the distance. It can be understood that the regression model is a relatively mature known technology, and its establishment method and application method can be referred to related technologies, and the embodiments of the present application will not be described in detail.

[0067] Step S160: determining a crack region based on the thermal infrared image and the suspected crack points.

[0068] It should be noted that for each suspected crack point, if the point indeed belongs to a crack point, the area around the point will have a microstructure change and a friction heat generation effect due to rock mass rupture, thereby showing abnormal thermal radiation characteristics. Based on this, the following scheme is provided:

[0069] Preferably, in one embodiment of the present application, the step S160 can include: performing feature point matching on the three-dimensional geometric model and the thermal infrared image to obtain a thermal point corresponding to each suspected crack point; wherein the thermal point is a corresponding position point of the suspected crack point on the thermal infrared image; obtaining a high-temperature pixel point in the thermal infrared image; wherein the high-temperature pixel point is a pixel point in the thermal infrared image with a temperature higher than the average temperature of the whole image; obtaining a plurality of high-temperature regions by region growing based on the high-temperature pixel point; obtaining a suspected crack region where the suspected crack point is located; wherein the suspected crack region is obtained by region expansion on a target point; the target point is a point in the community where the suspected crack region is located and greater than the average community influence degree; for each suspected crack point, calculating the intersection over union of the suspected crack region and the high-temperature region; determining a crack region based on the intersection over union; wherein the crack region is a suspected crack region with an intersection over union greater than a preset intersection over union threshold. For example:

[0070] By the method of feature point matching, the matching relationship between the three-dimensional geometric model and the points on the thermal infrared image is obtained, and then the corresponding position point of each suspected crack point on the thermal infrared image, denoted as a thermal point, can be obtained.

[0071] The average temperature of all pixel points calculated from the thermal infrared image is obtained, and the pixel points greater than the average temperature are recorded as high-temperature pixel points. A plurality of high-temperature regions are obtained by the method of region growing for the high-temperature pixel points.

[0072] For each suspected crack point, when the attribute of each point in the geometric three-dimensional geometric model is the influence degree, the average construction influence degree of all points in the community where the suspected crack point is located is calculated. The region where the suspected crack point is located is obtained by region expansion for the points greater than the average, denoted as a suspected crack region.

[0073] For each suspected crack point, the intersection over union of the suspected crack region and the high-temperature region corresponding to the thermal point is calculated, and the suspected crack region with an intersection over union greater than 0.7 is recorded as a crack region.

[0074] It should be noted that the intersection over union (IoU) is a commonly used index for measuring the degree of overlap between two regions (such as suspected crack regions and high-temperature regions). In the above scheme, the intersection over union is used to determine whether the suspected crack region matches the high-temperature region in the thermal infrared image, thereby determining the crack region. It can be understood that the intersection over union is a relatively mature known technology, and its specific implementation mode can be referred to related technologies, and the embodiments of the present application will not be described again.

[0075] Preferably, in one embodiment of the present application, after the above step S160, the safety monitoring method for the construction of the city railway further comprises: acquiring a crack parameter of the crack area; acquiring a risk level of the crack area based on the crack parameter; acquiring a treatment scheme for the crack area based on the risk level; and periodically acquiring a structural health condition of the crack area after the crack area is treated. For example, in this embodiment:

[0076] For the crack area, first, the crack parameter (such as width, direction, water seepage, etc.) is acquired, the risk level of the crack is evaluated, and temporary closure or support measures are taken;

[0077] According to the crack risk level, a permanent treatment scheme is selected, for example: surface closure is performed on small cracks, grouting repair is performed on medium cracks, structural reinforcement is performed on serious cracks, and blocking and draining are combined for water seepage cracks, and the grouting pressure and material performance are strictly controlled during construction to ensure dense filling;

[0078] After the crack area is treated, the crack stability is periodically re-measured, and is included in the long-term monitoring system, and the structural health condition is tracked in combination with BIM technology, and if the crack is related to the deformation of the stratum, the surrounding rock can also be simultaneously reinforced.

[0079] It should be noted that the scheme for evaluating the crack risk based on the crack parameter is a relatively mature known technology, and the specific implementation manner can be referred to related technologies, and the embodiment of the present application will not be described herein.

[0080] Please refer to Figure 4 which shows a safety monitoring system 200 for the construction of the city railway provided by one embodiment of the present application, and the system comprises: an upper computer 210 and a distributed stress sensor network 220, a thermal infrared sensor 230 and a three-dimensional laser scanner 240 arranged in a city railway construction tunnel, and the distributed stress sensor network 220, the thermal infrared sensor 230 and the three-dimensional laser scanner 240 are respectively in communication connection with the upper computer, wherein:

[0081] The distributed stress sensor network 220 is used to collect stress data of the city railway construction tunnel;

[0082] The thermal infrared sensor 230 is used to collect thermal infrared images of the city railway construction tunnel;

[0083] The three-dimensional laser scanner 240 is used to collect full-section point cloud data of the city railway construction tunnel;

[0084] The host computer 210 is configured to acquire a three-dimensional geometric model of a tunnel under construction of the urban railway based on full-section point cloud data, and is configured to acquire the three-dimensional geometric model, a thermal infrared image, and stress data collected by each stress sensor in the distributed stress sensor network. Based on the stress data, the construction influence degree of the construction operation on the corresponding position of the stress sensor is acquired. Based on the position of the stress sensor in the three-dimensional geometric model and the construction influence degree of the construction operation on the corresponding region of the stress sensor, the construction influence degree of the construction operation on all positions in the three-dimensional geometric model is acquired. The maximum value point of the influence degree is determined in the three-dimensional geometric model. Based on the maximum value point of the influence degree, the suspected crack point is determined. Based on the thermal infrared image and the suspected crack point, the crack region is determined.

[0085] It should be noted that the stress sensors in the distributed stress sensor network 220 described above can be arranged at every 5-20m along the tunnel vault, side wall and other key positions, and the high-risk area is encrypted to 3-5m. Optical fiber or vibrating wire sensors are used, and data is collected once per minute during the construction period. The thermal infrared sensor 230 can be a fixed thermal imager installed every 50-100m, cooperating with a mobile inspection device, scanning the whole tunnel at least once a day, and focusing on monitoring the region corresponding to the stress anomaly. The three-dimensional laser scanner can be used to quickly acquire full-section high-precision point cloud data (millimeter-level precision) of the tunnel, so as to construct the three-dimensional geometric model of the tunnel based on the full-section high-precision point cloud data of the tunnel.

[0086] It should be noted that the host computer 210 described above can be used to execute the safety monitoring method for the construction of the urban railway. For other functions of the host computer 210, please refer to the description of the safety monitoring method for the construction of the urban railway. Here, it will not be repeated.

[0087] So far, the application is completed.

[0088] In summary, in the embodiment of the present application, a three-dimensional geometric model of a city railway construction tunnel, a thermal infrared image and stress data collected by stress sensors arranged in the city railway construction tunnel are acquired; based on the stress data, the construction influence degree of the construction operation on the corresponding position of the stress sensor is acquired; based on the position of the stress sensor in the three-dimensional geometric model and the construction influence degree of the construction operation on the corresponding region of the stress sensor, the construction influence degree of the construction operation on all positions in the three-dimensional geometric model is acquired; the influence degree maximum point is determined in the three-dimensional geometric model; based on the influence degree maximum point, the suspected crack point is determined; based on the thermal infrared image and the suspected crack point, the crack region is determined. Thus, through the distributed stress sensor network arranged in the key region of the tunnel, the stress change data of different regions in the construction process is collected in real time; then, the stress field between different sensors is reconstructed by high-precision interpolation, the stress distribution of the whole region of the tunnel is acquired, and the potential stress concentration position is identified. Finally, combined with the tunnel surface temperature field distribution data acquired by the infrared thermal imager, through the multi-physical field coupling analysis method, the stress concentration region and the temperature abnormal region in the thermal infrared image are spatially matched and verified. Thus, the present application comprehensively uses the data collected by the three-dimensional geometric model, the thermal infrared image and the stress sensor, realizes the comprehensive perception of the tunnel construction condition, identifies the potential stress concentration position through the interpolation and reconstruction of the stress field and the construction of the stress distribution cloud diagram, and then combines the thermal infrared image to realize the spatial matching and the correlation verification, realizes the early accurate positioning and risk assessment of the crack, and greatly improves the monitoring timeliness.

[0089] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A safety monitoring method for construction of a metropolitan railway, characterized by, The method comprises: acquiring a three-dimensional geometric model of a city railway construction tunnel, a thermal infrared image, and stress data collected by stress sensors arranged in the city railway construction tunnel; based on the stress data, acquiring the construction influence degree of construction work on the positions corresponding to the stress sensors; based on the positions of the stress sensors in the three-dimensional geometric model and the construction influence degree of construction work on the regions corresponding to the stress sensors, acquiring the construction influence degree of construction work on all positions in the three-dimensional geometric model; determining an influence degree maximum point in the three-dimensional geometric model; based on the influence degree maximum point, determining a suspected crack point; based on the thermal infrared image and the suspected crack point, determining a crack region; the method comprises:

2. The safety monitoring method for the construction of a metropolitan railway according to claim 1, characterized in that, for each stress sensor, acquiring a stress reference value at the position corresponding to the stress sensor and a stress increase trend index; wherein the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset time period; calculating the ratio of the stress reference value to the maximum stress reference value among all stress reference values corresponding to all stress sensors; calculating the product of the ratio of the stress reference value to the maximum stress reference value and the increasing trend index to obtain the construction influence degree of construction work on the position corresponding to the stress sensor. The method comprises: for each stress sensor, acquiring a stress reference value at the position corresponding to the stress sensor and a stress increase trend index; wherein the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset time period; calculating the ratio of the stress reference value to the maximum stress reference value among all stress reference values corresponding to all stress sensors; calculating the product of the ratio of the stress reference value to the maximum stress reference value and the increasing trend index to obtain the construction influence degree of construction work on the position corresponding to the stress sensor. The method comprises: for each stress sensor, acquiring a stress reference value at the position corresponding to the stress sensor and a stress increase trend index; wherein the stress increase trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset time period; calculating the ratio of the stress reference value to the maximum stress reference value among all stress reference values corresponding to all stress sensors; calculating the product of the ratio of the stress reference value to the maximum stress reference value and the increasing trend index to obtain the construction influence degree of construction work on the position corresponding to the stress sensor.

3. The safety monitoring method for the construction of a metropolitan railway according to claim 1, characterized in that, The method comprises: based on the positions of the stress sensors, acquiring the coordinate positions of each stress sensor in the three-dimensional geometric model; based on the coordinate positions of each stress sensor in the three-dimensional geometric model, acquiring the construction influence degree of all positions in the three-dimensional geometric model by means of Kriging interpolation.

4. The safety monitoring method for the construction of a metropolitan railway according to claim 1, characterized in that, The method comprises: based on the construction influence degree of all positions in the three-dimensional geometric model, acquiring the spatial gradient of each position in the three-dimensional geometric model; Based on the spatial gradient, a Hessian matrix eigenvalue of each position is calculated; Based on the Hessian matrix eigenvalue, an influence degree maximum value point is obtained; wherein, the influence degree maximum value point is a point in the three-dimensional geometric model where the Hessian matrix eigenvalue is negative.

5. The safety monitoring method for the construction of a metropolitan railway according to claim 1, characterized in that, Based on the influence degree maximum value point, a suspected crack point is determined, including: With the construction influence degree of all positions in the three-dimensional geometric model as a label, a plurality of communities and a unified label of all nodes in each community are obtained by using a community detection algorithm based on a preset label propagation rule; wherein, the preset label propagation rule is that each node selects the label with the largest difference from the label value of the reference node from the adjacent nodes when converting the label through the adjacent nodes; A source node of each community is obtained; wherein, the source node is used to represent the construction position, and the source node is a node in the community whose initial label is the unified label; Based on the source node, an influence degree prediction value at the influence degree maximum value point is obtained; An absolute value of the difference between the construction influence degree of the influence degree maximum value point and the influence degree prediction value is obtained; The ratio of the absolute value of the difference to the corresponding influence degree prediction value is calculated; Based on the ratio of the absolute value of the difference to the corresponding influence degree prediction value, a suspected crack point is obtained; wherein, the suspected crack point is the influence degree maximum value point whose ratio of the absolute value of the difference to the corresponding influence degree prediction value is greater than a preset ratio value.

6. The safety monitoring method for the construction of a metropolitan railway according to claim 5, characterized in that, Based on the source node, an influence degree prediction value at the influence degree maximum value point is obtained, including: For each community, a regression model is established based on the construction influence degree of all non-influence degree maximum value points in the community and the distance between the non-influence degree maximum value points and the source node; The influence degree prediction value of the influence degree maximum value point is obtained by taking the influence degree maximum value point in the community and the distance between the influence degree maximum value point and the source node as inputs of the regression model.

7. The safety monitoring method for the construction of a metropolitan railway according to claim 5, characterized in that, Based on the thermal infrared image and the suspected crack point, a crack region is determined, including: Feature point matching is performed on the three-dimensional geometric model and the thermal infrared image to obtain a corresponding thermal point of each suspected crack point; wherein, the thermal point is a corresponding position point of the suspected crack point on the thermal infrared image; High-temperature pixel points in the thermal infrared image are obtained; wherein, the high-temperature pixel points are pixel points in the thermal infrared image whose temperature is higher than the average temperature of the whole image; Based on the high-temperature pixel points, a plurality of high-temperature regions are obtained by region growing; A suspected crack region where the suspected crack point is located is obtained; wherein, the suspected crack region is obtained by region expansion on a target point; the target point is a point in the community where the suspected crack region is located and whose community influence degree is greater than the average community influence degree; For each suspected crack point, the intersection-over-union ratio of the suspected crack region and the high-temperature region is calculated; Based on the intersection-over-union ratio, a crack region is determined; wherein, the crack region is a suspected crack region whose intersection-over-union ratio is greater than a preset intersection-over-union ratio threshold. 8.The method for safety monitoring of city regional railway construction according to any one of claims 1-7, characterized in that, After determining the crack region based on the thermal infrared image and the suspected crack point, the method comprises: obtaining a crack parameter of the crack region; obtaining a risk level of the crack region based on the crack parameter; obtaining a treatment scheme for the crack region based on the risk level; obtaining a structural health condition of the crack region periodically after treating the crack region.

9. A safety monitoring system for construction of a commuter rail, characterized by, The system comprises a host computer and a distributed stress sensor network, a thermal infrared sensor and a three-dimensional laser scanner arranged in a city railway construction tunnel, and the distributed stress sensor network, the thermal infrared sensor and the three-dimensional laser scanner are in communication connection with the host computer, wherein: The distributed stress sensor network is used to collect stress data of the city railway construction tunnel. The thermal infrared sensor is used to collect thermal infrared images of the city railway construction tunnel. The three-dimensional laser scanner is used to collect full-section point cloud data of the city railway construction tunnel. The host computer is used to obtain a three-dimensional geometric model of the city railway construction tunnel based on the full-section point cloud data, and is used to obtain the three-dimensional geometric model, the thermal infrared image and the stress data collected by each stress sensor in the distributed stress sensor network; based on the stress data, the construction influence degree of construction operation on the corresponding position of the stress sensor is obtained; based on the position of the stress sensor in the three-dimensional geometric model and the construction influence degree of construction operation on the corresponding region of the stress sensor, the construction influence degree of construction operation on all positions in the three-dimensional geometric model is obtained; the maximum influence degree point in the three-dimensional geometric model is determined; based on the maximum influence degree point, a suspected crack point is determined; based on the thermal infrared image and the suspected crack point, a crack region is determined. Based on the stress data, the construction influence degree of construction operation on the corresponding position of the stress sensor is obtained, which comprises: for each stress sensor, obtaining a stress reference value and a stress increasing trend index at the corresponding position of the stress sensor; wherein the stress increasing trend index is used to represent the increasing trend of the stress data at the corresponding position; the stress reference value is the average value of the stress data collected by the stress sensor within a preset period; calculating the ratio of the stress reference value to the maximum stress reference value among all stress reference values corresponding to the stress sensors; calculating the product of the ratio of the stress reference value to the maximum stress reference value and the increasing trend index, to obtain the construction influence degree of construction operation on the corresponding position of the stress sensor.

Images