Safety protection monitoring method and system for pipe gallery construction

Abstract

The invention relates to the technical field of safety monitoring, in particular to a pipe gallery construction safety protection monitoring method and system.The method comprises the following steps that a stress sensor is installed at a key bearing structure of a pipe gallery, and the stress sensor monitors and records stress changes of the key bearing structure in the construction process; the stress sensors acquire structural stress data in real time, transmit the acquired data to the central processing unit in real time, and integrally process the data from the stress sensors; the central processing unit performs deep analysis on the stress data by using a multi-dimensional stress correlation analysis method, and identifies a potential risk area; when the abnormal stress level or trend is detected, early warning is automatically triggered; and assisting verification based on an image recognition technology. According to the invention, by combining the stress sensor data and the image recognition technology, the structural stress change and potential risk in the pipe gallery construction process can be monitored and analyzed more accurately.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology, and in particular to a method and system for safety protection and monitoring of utility tunnel construction. Background Technology

[0002] In the field of utility tunnel construction, safety protection and risk management are key factors in ensuring project quality and the safety of workers. Traditional safety monitoring methods for utility tunnel construction rely on physical inspections and basic monitoring tools. These methods often cannot provide real-time data analysis and early warning, and have limitations in dealing with complex structural stresses and potential risks. Therefore, developing efficient and accurate monitoring technologies is crucial for improving the safety and efficiency of utility tunnel construction.

[0003] In recent years, the development of sensor technology, wireless communication technology, and image processing technology has provided new possibilities for enhancing the monitoring and safety management of utility tunnel construction. In particular, technological advancements in stress monitoring, data processing, and image analysis have laid the foundation for achieving a more comprehensive and efficient monitoring system.

[0004] However, existing technologies still have some shortcomings, such as insufficient comprehensive monitoring of structural stress, a lack of effective data fusion and analysis methods, and limitations in applying image recognition technology to practical engineering monitoring. These problems limit the ability of existing systems to identify and respond to potential risks in a timely manner.

[0005] In view of this, the present invention proposes a method for monitoring the safety of pipe gallery construction by comprehensively applying multidimensional stress correlation analysis, image recognition technology, and advanced data processing. This method aims to provide a comprehensive, real-time, and efficient pipe gallery construction monitoring system by combining high-precision stress sensors and image monitoring equipment with powerful data analysis and image processing algorithms. Summary of the Invention

[0006] To achieve the above objectives, the present invention provides a method and system for safety protection and monitoring during the construction of utility tunnels.

[0007] A method for safety monitoring during the construction of utility tunnels includes the following steps:

[0008] S1: Stress sensor layout: High-sensitivity stress sensors are installed at the key load-bearing structures of the pipe gallery. The key load-bearing structures include the top, bottom and support points of the pipe gallery. The stress sensors monitor and record the stress changes of the key load-bearing structures during the construction process.

[0009] S2: Real-time data acquisition. Stress sensors acquire structural stress data in real time, including stress magnitude and stress point location. The stress point location is achieved by pre-marking the location of each stress sensor. The acquired data is transmitted to the central processing unit in real time for integrated processing of data from each stress sensor.

[0010] S3: Stress data analysis. The central processing unit uses multidimensional stress correlation analysis to perform in-depth analysis of stress data and identify potential risk areas.

[0011] S4: Early warning mechanism. When an abnormal stress level or trend is detected (exceeding the preset threshold or changing rapidly), an early warning is automatically triggered to promptly notify the project management personnel and construction team.

[0012] S5: Image recognition technology is used to assist in verification. Cameras installed in the construction area of ​​the utility tunnel are used for visual monitoring to assist in verifying stress monitoring data.

[0013] Furthermore, S1 specifically includes:

[0014] Determine installation points: Analyze the structural design drawings of the utility tunnel to determine the location of load-bearing structures, such as beams, columns, support points, and joints;

[0015] S11, Sensor type selection: Select a resistance strain gauge or fiber Bragg grating sensor suitable for monitoring stress in large structures.

[0016] S12, use welding, bonding or mechanical fixing methods to fix the stress sensor, and calibrate each stress sensor after installation.

[0017] Furthermore, S2 specifically includes:

[0018] S21, Pre-configuration of stress sensors: Before installing stress sensors, each stress sensor is assigned a unique identifier and associated with its actual installation location. The location information is pre-entered into the central processing unit to identify the specific location of each stress sensor during data acquisition.

[0019] S22, Stress point location tracking: Through the central processing unit, the specific location of the stress point is tracked and recorded in real time using the pre-marked location information of each stress sensor;

[0020] S23, Stress Change Trend Analysis: Monitor and record the stress level change trend over time;

[0021] S24 transmits the collected stress magnitude and stress point location data to the central processing unit in real time via wireless or wired connection.

[0022] S25, Data Synchronization and Calibration: Regularly calibrate and synchronize the data to ensure the accuracy and consistency of stress sensor data.

[0023] Furthermore, the multidimensional stress correlation analysis method in S3 specifically includes:

[0024] S31, Data preprocessing: Collect stress magnitude and location from each stress sensor, acquire timestamp data, and perform data standardization processing;

[0025] S32, construct a stress correlation matrix, analyze the stress data of each stress sensor and the stress correlation between nearby stress sensors, and quantify it through a correlation coefficient algorithm;

[0026] S33, Spatial Relationship Mapping: Maps the physical location of the stress sensor to the created three-dimensional spatial model to ensure that the stress data is consistent with its actual location in the pipe gallery structure;

[0027] S34. By applying the stress distribution network analysis method, stress sensors are regarded as nodes in the stress distribution network, and the stress correlation between nodes is regarded as edges. Through graph theory analysis, key nodes and paths in the stress distribution network are identified to point out the weak points of the structure.

[0028] Furthermore, the correlation coefficient algorithm in S32 uses the Spearman rank correlation coefficient. For any two sensors i and j, the Spearman rank correlation coefficient is expressed as:

[0029] Where, ρ ij d represents the Spearman rank correlation coefficient between the i-th and j-th stress sensors. ij (k) represents the grade difference, which reflects the relative grade difference of the stress values ​​recorded by the two sensors among all measurements at time point k. n represents the number of time points, which refers to the total number of time points considered throughout the monitoring process.

[0030] The stress correlation matrix in S32 is a square matrix whose size is the square of the number of stress sensors. Each element of the matrix represents the Spearman rank correlation coefficient between a pair of stress sensors.

[0031] Examine the stress correlation matrix to identify sensor pairs with correlation coefficients higher than a predetermined threshold. High correlation indicates that the monitored stresses are influenced by each other. Map the sensor pairs onto the physical structure of the utility tunnel to identify their specific locations and analyze the structural importance of these locations, including whether they are close to load-bearing points or joint structures. Perform structural integrity analysis on the areas where sensor pairs showing high correlation are located.

[0032] Furthermore, S34 specifically includes:

[0033] Node definition: In the stress distribution network, each stress sensor is regarded as a node. Each node represents the physical location of the stress sensor and the stress data monitored at that physical location.

[0034] Edge establishment: The stress correlation matrix constructed based on the Spearman rank correlation coefficient is used to define the edges between nodes. When the correlation coefficient between two stress sensors exceeds a predetermined threshold, an edge is established between the two nodes. The weight of the edge is determined by the absolute value of the correlation coefficient.

[0035] Graph theory analysis: A stress distribution network graph is constructed using the definitions of all nodes and edges. The stress distribution network graph reflects the stress correlation between various monitoring points in the entire pipe gallery construction area. The degree centrality analysis in graph theory is used to identify important nodes in the stress distribution network, and the shortest path algorithm is used to identify important paths in the stress distribution network.

[0036] Structural weak point indication: By analyzing important nodes and important paths, potential weak points in the utility tunnel structure can be identified. A node with high degree centrality (highly connected nodes) indicates an important node that bears stress in multiple directions, and an important path indicates a potential structural stress concentration line.

[0037] Furthermore, the degree centrality analysis is defined as how many other nodes a node is directly connected to. In a stress distribution network, a high degree centrality of a node (i.e., a stress sensor) indicates that the stress sensor has a significant stress correlation with multiple other stress sensors. For a node v in the stress distribution network, its degree centrality C... D (v) is defined as:

[0038] The shortest path algorithm is used to find the shortest path between two nodes. In the stress distribution network, it is used to identify the main stress transmission paths. Dijkstra's algorithm is used as the shortest path algorithm. For two nodes u and v in the network, the calculation steps are as follows:

[0039] Except for the starting point u, set the initial distance of all nodes to infinity, and set the distance of the starting point u to 0;

[0040] Find the node with the smallest distance and update the distances of its neighboring nodes;

[0041] Repeat the above process of retrieving and updating until all nodes have been processed.

[0042] The shortest distance to the final node v is the shortest path length from u to v.

[0043] Furthermore, S5 specifically includes:

[0044] A verification section is selected in the construction area of ​​the utility tunnel, the verification section including the key load-bearing structure, and cameras are installed in the verification section;

[0045] Centralized image data collection is carried out for the verification section. The camera continuously monitors and collects real-time image data to ensure that the timestamp of the image data is synchronized with the timestamp of the stress monitoring data in the verification section.

[0046] An image recognition algorithm based on convolutional neural networks (CNN) was used to analyze the images of the verification section and identify structural changes and crack development issues.

[0047] The image recognition results are compared and analyzed with the stress monitoring data in the verification section, and the image data is used to assist in verifying the stress monitoring data.

[0048] Furthermore, the CNN analysis verification segment image specifically includes:

[0049] Preprocessing: The validation segment image is preprocessed to the input format required by the CNN, including resizing and normalization;

[0050] Feature extraction: Image features are extracted using convolutional and pooling layers of a CNN. Different filters capture different features in the image, including the edges of cracks or shape changes of structures.

[0051] Crack and change detection: The extracted features are analyzed using a trained CNN model to identify feature patterns related to crack development or structural changes.

[0052] Output Explanation: The fully connected layers of a CNN provide classification results regarding the presence of cracks or structural changes in an image;

[0053] Post-processing: Based on the CNN output, potential crack regions or structural changes are marked in the original image for analysis.

[0054] A safety monitoring system for utility tunnel construction, used to implement the aforementioned safety monitoring method for utility tunnel construction, includes the following modules:

[0055] Stress monitoring module: Includes multiple stress sensors, installed at key load-bearing structures of the pipe gallery, for real-time monitoring and recording of structural stress data;

[0056] Data transmission and processing module: The collected stress data is transmitted to the central processing unit in real time using wireless communication technology. The central processing unit integrates the stress data and performs in-depth analysis using multidimensional stress correlation analysis to identify potential risk areas.

[0057] Image monitoring and analysis module: Cameras are installed in the verification section of the utility tunnel to collect real-time image data. Convolutional neural network (CNN) image recognition technology is applied to analyze the collected images, identify structural changes and crack development, and fuse them with stress monitoring data for verification.

[0058] Early warning and response module: When abnormal stress levels are detected or image analysis results indicate potential risks, an early warning is automatically triggered to notify project managers and construction teams.

[0059] The beneficial effects of this invention are:

[0060] This invention, by combining stress sensor data and image recognition technology, can more accurately monitor and analyze structural stress changes and potential risks during the construction of utility tunnels. The combined use of stress distribution network analysis and image recognition technology can identify structural weak points, such as cracks and deformations, at a more detailed level, thereby providing early warning of potential structural problems, reducing the need for monitoring a wide area, and saving resources and costs.

[0061] This invention, through multidimensional stress correlation analysis and graph theory algorithms, can accurately identify key stress points and main stress transmission paths in the pipe gallery structure. This in-depth analysis enhances the ability to predict potential risk areas, enabling engineering managers to make more targeted structural assessments and maintenance decisions, thereby improving the safety management level of pipe gallery construction projects.

[0062] This invention achieves effective monitoring and auxiliary verification of key areas by centrally deploying high-resolution cameras and applying image recognition technology in the verification section of the utility tunnel construction, without the need for extensive deployment throughout the entire construction area. This not only reduces the overall cost of the monitoring system but also simplifies the data processing and analysis process, and improves the overall operation and maintenance efficiency of the monitoring system. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is a schematic diagram of the protection and monitoring method according to an embodiment of the present invention;

[0065] Figure 2 This is a schematic diagram of the functional modules of the protection and monitoring system according to an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0067] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0068] like Figure 1 As shown, a method for safety monitoring during the construction of utility tunnels includes the following steps:

[0069] S1: Stress sensor layout: High-sensitivity stress sensors are installed at the critical load-bearing structures of the utility tunnel, including the top, bottom and support points of the tunnel. The stress sensors monitor and record stress changes in the critical load-bearing structures during construction.

[0070] S2: Real-time data acquisition. Stress sensors acquire structural stress data in real time, including stress magnitude and stress point location. The stress point location is achieved by pre-marking the location of each stress sensor. The acquired data is transmitted to the central processing unit in real time for integrated processing of data from each stress sensor.

[0071] S3: Stress data analysis. The central processing unit uses multidimensional stress correlation analysis to perform in-depth analysis of stress data and identify potential risk areas.

[0072] S4: Early warning mechanism. When an abnormal stress level or trend is detected (such as exceeding a preset threshold or changing rapidly), an early warning is automatically triggered to promptly notify the project management personnel and construction team.

[0073] S5: Image recognition technology is used to assist in verification. Cameras installed in the construction area of ​​the utility tunnel are used for visual monitoring to assist in verifying stress monitoring data.

[0074] S1 specifically includes:

[0075] Determine installation points: Analyze the structural design drawings of the utility tunnel to determine the location of load-bearing structures, such as beams, columns, support points, and joints;

[0076] S11, Sensor type selection: Select a resistance strain gauge or fiber Bragg grating sensor suitable for monitoring stress in large structures.

[0077] S12, use welding, bonding or mechanical fixing methods to fix the stress sensor, and calibrate each stress sensor after installation.

[0078] S2 specifically includes:

[0079] S21, Pre-configuration of stress sensors: Before installing stress sensors, each stress sensor is assigned a unique identifier and associated with its actual installation location. The location information is pre-entered into the central processing unit to identify the specific location of each stress sensor during data acquisition.

[0080] S22, Stress point location tracking: Through the central processing unit, the specific location of the stress point is tracked and recorded in real time using the pre-marked location information of each stress sensor;

[0081] S23, Stress Change Trend Analysis: Monitor and record the stress level change trend over time;

[0082] S24 transmits the collected stress magnitude and stress point location data to the central processing unit in real time via wireless or wired connection.

[0083] S25, Data Synchronization and Calibration: Regularly calibrate and synchronize the data to ensure the accuracy and consistency of stress sensor data.

[0084] The multidimensional stress correlation analysis method in S3 specifically includes:

[0085] S31, Data preprocessing: Collect stress magnitude and location from each stress sensor, acquire timestamp data, and perform data standardization processing;

[0086] S32, construct a stress correlation matrix, analyze the stress data of each stress sensor and the stress correlation between nearby stress sensors, and quantify it through a correlation coefficient algorithm;

[0087] S33, Spatial Relationship Mapping: Maps the physical location of the stress sensor to the created three-dimensional spatial model to ensure that the stress data is consistent with its actual location in the pipe gallery structure;

[0088] S34. By applying the stress distribution network analysis method, stress sensors are regarded as nodes in the stress distribution network, and the stress correlation between nodes is regarded as edges. Through graph theory analysis, key nodes and paths in the stress distribution network are identified to point out the weak points of the structure.

[0089] The correlation coefficient algorithm in S32 uses the Spearman rank correlation coefficient, and the original formula is: Where ρ is the Spearman rank correlation coefficient, and d i This represents the difference in ranks between the two variables, where n is the number of data points, and 6 is a fixed coefficient used to correct for the sum of squared rank differences. It appears in the formula to ensure proper scaling of the correlation coefficient, allowing the calculated correlation coefficient to vary within the range of -1 to +1. This range represents a perfect negative correlation (-1) to a perfect positive correlation (+1).

[0090] In the application of this invention, for any two sensors i and j, the Spearman rank correlation coefficient is expressed as:

[0091] Where, ρ ij d represents the Spearman rank correlation coefficient between the i-th and j-th stress sensors. ij (k) represents the rank difference, reflecting the relative rank difference between the stress values ​​recorded by the two sensors at time point k among all measurements. For example, if at a certain moment, the stress recorded by sensor i ranks 3rd among all sensors, while the stress recorded by sensor j ranks 10th, then d ij The value is 7, where n is the number of time points, referring to the total number of time points considered throughout the entire monitoring process. 2 -1) is a scaling factor based on the number of time n, ensuring that the entire expression varies within the correct range;

[0092] The stress correlation matrix in S32 is a square matrix whose size is the square of the number of stress sensors. Each element of the matrix represents the Spearman rank correlation coefficient between a pair of stress sensors. For example, if there are three sensors A, B, and C, the stress correlation matrix would look like this:

[0093] A B C A 1 ρ_AB ρ_AC B ρ_BA 1 ρ_BC C ρ_CA ρ_CB 1

[0094] Based on the stress correlation matrix, the analysis of the interactions between sensors includes the following steps:

[0095] Examine the stress correlation matrix to identify sensor pairs with correlation coefficients higher than a predetermined threshold. High correlation indicates that the monitored stresses are influenced by each other. Map the sensor pairs onto the physical structure of the utility tunnel to identify their specific locations and analyze the structural importance of these locations, including whether they are close to load-bearing points or joint structures. Perform structural integrity analysis on the areas where sensor pairs showing high correlation are located.

[0096] S34 specifically includes:

[0097] Node definition: In the stress distribution network, each stress sensor is regarded as a node. Each node represents the physical location of the stress sensor and the stress data monitored at that physical location.

[0098] Edge establishment: The stress correlation matrix constructed based on the Spearman rank correlation coefficient is used to define the edges between nodes. When the correlation coefficient between two stress sensors exceeds a predetermined threshold, an edge is established between the two nodes. The weight of the edge is determined by the absolute value of the correlation coefficient.

[0099] Graph theory analysis: A stress distribution network graph is constructed using the definitions of all nodes and edges. This graph reflects the stress correlation between monitoring points throughout the entire utility tunnel construction area. Degree centrality analysis in graph theory is used to identify important nodes in the stress distribution network, and the shortest path algorithm is used to identify important paths within the network. These paths represent the main pathways of stress transmission in the structure or areas with potentially high risks.

[0100] Structural weak point indication: By analyzing important nodes and important paths, potential weak points in the utility tunnel structure can be identified. A node with high degree centrality (highly connected nodes) indicates an important node that bears stress in multiple directions, and an important path indicates a potential structural stress concentration line.

[0101] Degree centrality analysis is defined as the number of other nodes directly connected to a node. In a stress distribution network, a high degree centrality of a node (i.e., a stress sensor) indicates that the stress sensor has a significant stress correlation with multiple other stress sensors. For a node v in the stress distribution network, its degree centrality C... D (v) is defined as:

[0102] In a stress distribution network, nodes with high degree centrality are critical stress points on the structure because they show a high stress correlation with multiple other points and may be key areas of stress concentration or dispersion.

[0103] The shortest path algorithm is used to find the shortest path between two nodes. In stress distribution networks, it is used to identify the main stress transmission paths. Dijkstra's algorithm is used as the shortest path algorithm. For two nodes u and v in the network, the calculation steps are as follows:

[0104] Except for the starting point u, set the initial distance of all nodes to infinity, and set the distance of the starting point u to 0;

[0105] Find the node with the smallest distance and update the distances of its neighboring nodes;

[0106] Repeat the above process of retrieving and updating until all nodes have been processed.

[0107] The shortest distance to the final node v is the shortest path length from u to v.

[0108] In stress distribution networks, shortest path analysis can reveal how stress is transmitted from one point to another in a structure, helping to identify potential stress concentration areas or structural weaknesses.

[0109] S5 specifically includes:

[0110] Select a verification section within the utility tunnel construction area. The verification section includes the critical load-bearing structure. Install cameras within the verification section, placing them near stress sensors.

[0111] Centralized image data collection is carried out for the verification section. The camera continuously monitors and collects real-time image data to ensure that the timestamp of the image data is synchronized with the timestamp of the stress monitoring data in the verification section.

[0112] An image recognition algorithm based on convolutional neural networks (CNN) was used to analyze the images of the verification section and identify structural changes and crack development issues.

[0113] The image recognition results are compared and analyzed with the stress monitoring data in the verification section. The image data is used to assist in verifying the stress monitoring data, especially when the stress data shows anomalies, the image data is used to confirm whether there is a real structural problem.

[0114] The CNN analysis of the verification segment images specifically includes:

[0115] Preprocessing: The validation segment image is preprocessed to the input format required by the CNN, including resizing and normalization;

[0116] Feature extraction: Image features are extracted using convolutional and pooling layers of a CNN. Different filters capture different features in the image, including the edges of cracks or shape changes of structures.

[0117] Crack and change detection: The extracted features are analyzed using a trained CNN model to identify feature patterns related to crack development or structural changes.

[0118] Output Interpretation: The fully connected layers of a CNN provide classification results about whether there are cracks or structural changes in an image. For example, the output may indicate "cracks present" or "no cracks present".

[0119] Post-processing: Based on the CNN output, potential crack regions or structural changes are marked in the original image for analysis.

[0120] The computation steps of CNN are as follows.

[0121] Convolutional layers: Convolutional layers extract features by sliding filters (convolutional kernels) across the input image. Each filter corresponds to a specific feature, including edges, color, or texture. The calculation formula is as follows:

[0122] in, It is the output feature map of the k-th filter at position (i,j), I is the input image, and W is the output feature map of the k-th filter at position (i,j). k It is the weight of the k-th filter, b k It is a bias term.

[0123] Activation layers: Using non-linear activation functions increases the non-linearity of the network and enhances the expressive power of features. in, It is the activated feature map;

[0124] Pooling layers: Reduce the spatial size of feature maps through downsampling, increasing the robustness of the model and reducing computational cost; max pooling selects the maximum value.

[0125] in, This is the feature map after pooling, and M×N is the size of the pooling window.

[0126] Fully connected layer: Converts the feature maps extracted by convolution and pooling into one-dimensional vectors for final classification or regression analysis.

[0127] like Figure 2 As shown, a safety monitoring system for utility tunnel construction is used to implement the aforementioned safety monitoring method for utility tunnel construction, and includes the following modules:

[0128] Stress monitoring module: Includes multiple stress sensors, installed at key load-bearing structures of the pipe gallery, for real-time monitoring and recording of structural stress data;

[0129] Data transmission and processing module: The collected stress data is transmitted to the central processing unit in real time using wireless communication technology. The central processing unit integrates the stress data and performs in-depth analysis using multidimensional stress correlation analysis to identify potential risk areas.

[0130] Image monitoring and analysis module: Cameras are installed in the verification section of the utility tunnel to collect real-time image data. Convolutional neural network (CNN) image recognition technology is applied to analyze the collected images, identify structural changes and crack development, and fuse them with stress monitoring data for verification.

[0131] Early warning and response module: When abnormal stress levels are detected or image analysis results indicate potential risks, an early warning is automatically triggered to notify project managers and construction teams.

[0132] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.

[0133] This invention is intended to cover all such substitutions, modifications, and variations falling within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for safety protection and monitoring during the construction of utility tunnels, characterized in that, Includes the following steps: S1: Stress sensor layout: Stress sensors are installed at the key load-bearing structures of the utility tunnel. The key load-bearing structures include the top, bottom and support points of the utility tunnel. The stress sensors monitor and record the stress changes of the key load-bearing structures during the construction process. S2: Real-time data acquisition. Stress sensors acquire structural stress data in real time, including stress magnitude and stress point location. The stress point location is achieved by pre-marking the location of each stress sensor. The acquired data is transmitted to the central processing unit in real time for integrated processing of data from each stress sensor. S3: Stress data analysis. The central processing unit uses multidimensional stress correlation analysis to perform in-depth analysis of stress data and identify potential risk areas. S4: Early warning mechanism. When an abnormal stress level or trend is detected, an early warning will be automatically triggered to promptly notify the project management personnel and construction team. S5: Image recognition technology is used to assist in verification. Cameras installed in the construction area of ​​the utility tunnel are used for visual monitoring to assist in verifying stress monitoring data.

2. The method for safety protection and monitoring during the construction of utility tunnels according to claim 1, characterized in that, S1 specifically includes: Determine the installation points: Analyze the structural design drawings of the utility tunnel to determine the location of the load-bearing structures; S11, Sensor type selection: Select a resistance strain gauge or fiber Bragg grating sensor suitable for monitoring stress in large structures. S12, use welding, bonding or mechanical fixing methods to fix the stress sensor, and calibrate each stress sensor after installation.

3. The method for safety protection and monitoring during the construction of utility tunnels according to claim 2, characterized in that, S2 specifically includes: S21, Pre-configuration of stress sensors: Before installing stress sensors, each stress sensor is assigned a unique identifier and associated with its actual installation location. The location information is pre-entered into the central processing unit to identify the specific location of each stress sensor during data acquisition. S22, Stress point location tracking: Through the central processing unit, the specific location of the stress point is tracked and recorded in real time using the pre-marked location information of each stress sensor; S23, Stress Change Trend Analysis: Monitor and record the stress level change trend over time; S24 transmits the collected stress magnitude and stress point location data to the central processing unit in real time via wireless or wired connection. S25, Data Synchronization and Calibration: Regularly calibrate and synchronize the data to ensure the accuracy and consistency of stress sensor data.

4. The method for safety protection and monitoring of utility tunnel construction according to claim 3, characterized in that, The multidimensional stress correlation analysis method in S3 specifically includes: S31, Data preprocessing: Collect stress magnitude and location from each stress sensor, acquire timestamp data, and perform data standardization processing; S32, construct a stress correlation matrix, analyze the stress data of each stress sensor and the stress correlation between nearby stress sensors, and quantify it through a correlation coefficient algorithm; S33, Spatial Relationship Mapping: Maps the physical location of the stress sensor to the created three-dimensional spatial model to ensure that the stress data is consistent with its actual location in the pipe gallery structure; S34. By applying the stress distribution network analysis method, stress sensors are regarded as nodes in the stress distribution network, and the stress correlation between nodes is regarded as edges. Through graph theory analysis, key nodes and paths in the stress distribution network are identified to point out the weak points of the structure.

5. The method for safety protection and monitoring during the construction of utility tunnels according to claim 4, characterized in that, The correlation coefficient algorithm in S32 uses the Spearman rank correlation coefficient. For any two sensors i and j, the Spearman rank correlation coefficient is expressed as: Where, ρ ij d represents the Spearman rank correlation coefficient between the i-th and j-th stress sensors. ij (k) represents the grade difference, which reflects the relative grade difference of the stress values ​​recorded by the two sensors among all measurements at time point k. n represents the number of time points, which refers to the total number of time points considered throughout the monitoring process. The stress correlation matrix in S32 is a square matrix whose size is the square of the number of stress sensors. Each element of the matrix represents the Spearman rank correlation coefficient between a pair of stress sensors. Examine the stress correlation matrix to identify sensor pairs with correlation coefficients higher than a predetermined threshold. High correlation indicates that the monitored stresses are influenced by each other. Map the sensor pairs onto the physical structure of the utility tunnel to identify their specific locations and analyze the structural importance of these locations, including whether they are close to load-bearing points or joint structures. Perform structural integrity analysis on the areas where sensor pairs showing high correlation are located.

6. The method for safety protection and monitoring during the construction of a utility tunnel according to claim 5, characterized in that, S34 specifically includes: Node definition: In the stress distribution network, each stress sensor is regarded as a node. Each node represents the physical location of the stress sensor and the stress data monitored at that physical location. Edge establishment: The stress correlation matrix constructed based on the Spearman rank correlation coefficient is used to define the edges between nodes. When the correlation coefficient between two stress sensors exceeds a predetermined threshold, an edge is established between the two nodes. The weight of the edge is determined by the absolute value of the correlation coefficient. Graph theory analysis: A stress distribution network graph is constructed using the definitions of all nodes and edges. The stress distribution network graph reflects the stress correlation between various monitoring points in the entire pipe gallery construction area. The degree centrality analysis in graph theory is used to identify important nodes in the stress distribution network, and the shortest path algorithm is used to identify important paths in the stress distribution network. Structural weak point indication: By analyzing important nodes and important paths, potential weak points in the pipe gallery structure are identified. High node centrality indicates that it is an important node that bears stress in multiple directions, and an important path indicates a potential structural stress concentration line.

7. The method for safety protection and monitoring of utility tunnel construction according to claim 6, characterized in that, The degree centrality analysis is defined as the number of other nodes directly connected to a node. In a stress distribution network, a high degree centrality of a node indicates that the stress sensor has a significant stress correlation with multiple other stress sensors. For a node v in the stress distribution network, its degree centrality C... D (v) is defined as: The shortest path algorithm is used to find the shortest path between two nodes. In the stress distribution network, it is used to identify the main stress transmission paths. Dijkstra's algorithm is used as the shortest path algorithm. For two nodes u and v in the network, the calculation steps are as follows: Except for the starting point u, set the initial distance of all nodes to infinity, and set the distance of the starting point u to 0; Find the node with the smallest distance and update the distances of its neighboring nodes; Repeat the above process of retrieving and updating until all nodes have been processed. The shortest distance to the final node v is the shortest path length from u to v.

8. A method for safety protection and monitoring during the construction of utility tunnels according to claim 7, characterized in that, S5 specifically includes: A verification section is selected in the construction area of ​​the utility tunnel, the verification section including the key load-bearing structure, and cameras are installed in the verification section; Centralized image data collection is carried out for the verification section. The camera continuously monitors and collects real-time image data to ensure that the timestamp of the image data is synchronized with the timestamp of the stress monitoring data in the verification section. An image recognition algorithm based on convolutional neural networks (CNN) was used to analyze the images of the verification section and identify structural changes and crack development issues. The image recognition results are compared and analyzed with the stress monitoring data in the verification section, and the image data is used to assist in verifying the stress monitoring data.

9. A method for safety protection and monitoring during the construction of utility tunnels according to claim 8, characterized in that, The CNN analysis verification segment image specifically includes: Preprocessing: The validation segment image is preprocessed to the input format required by the CNN, including resizing and normalization; Feature extraction: Image features are extracted using convolutional and pooling layers of a CNN. Different filters capture different features in the image, including the edges of cracks or shape changes of structures. Crack and change detection: The extracted features are analyzed using a trained CNN model to identify feature patterns related to crack development or structural changes. Output Explanation: The fully connected layers of a CNN provide classification results regarding the presence of cracks or structural changes in an image; Post-processing: Based on the CNN output, potential crack regions or structural changes are marked in the original image for analysis.

10. A safety monitoring system for utility tunnel construction, used to implement the safety monitoring method for utility tunnel construction as described in any one of claims 1-9, characterized in that, Includes the following modules: Stress monitoring module: Includes multiple stress sensors, installed at key load-bearing structures of the pipe gallery, for real-time monitoring and recording of structural stress data; Data transmission and processing module: The collected stress data is transmitted to the central processing unit in real time using wireless communication technology. The central processing unit integrates the stress data and performs in-depth analysis using multidimensional stress correlation analysis to identify potential risk areas. Image monitoring and analysis module: Cameras are installed in the verification section of the utility tunnel to collect real-time image data. Convolutional neural network (CNN) image recognition technology is applied to analyze the collected images, identify structural changes and crack development, and fuse them with stress monitoring data for verification. Early warning and response module: When abnormal stress levels are detected or image analysis results indicate potential risks, an early warning is automatically triggered to notify project managers and construction teams.

Images