A highway tunnel construction safety intelligent monitoring system

Abstract

This invention relates to the field of safety monitoring technology, and more particularly to an intelligent monitoring system for highway tunnel construction safety. The system includes a data acquisition module, a contour analysis module, a feature fusion module, and a data compensation module. By acquiring images of the tunnel surface, dust concentration within the tunnel, and water vapor humidity, the system divides the tunnel into several sub-image regions. The system determines the unevenness and texture complexity of each sub-image region, calculates region-sensitive feature values, and uses the tunnel dust concentration and water vapor humidity, combined with the region-sensitive feature values, to determine whether each sub-image region meets preset conditions. The system also calculates data missing characterization values ​​to determine the modeling anomaly tendency of each sub-image region. Based on the modeling anomaly tendency and the data missing characterization values, the system determines the data acquisition method for each sub-image region. This invention improves the accuracy and efficiency of model building through comprehensive analysis, data acquisition, and compensation of the tunnel environment, thereby ensuring construction safety.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology, and in particular to an intelligent monitoring system for safety during highway tunnel construction. Background Technology

[0002] Highway tunnel construction faces complex environments and challenges related to geological risks, machinery operation, personnel safety, and traffic control. In particular, operations during traffic hours significantly increase safety risks, and traditional manual monitoring and decentralized management are insufficient to meet the efficiency and safety requirements of modern tunnel construction. With the development of technologies such as the Internet of Things (IoT), artificial intelligence (AI), 5G communication, and big data analytics, installing various sensors and monitoring equipment within tunnels allows for real-time collection of geological data, machinery operating status, personnel locations, and traffic flow. Utilizing big data analytics and AI algorithms for in-depth data mining and analysis enables real-time monitoring, risk assessment, and predictive warnings of construction progress. Intelligent monitoring systems are gradually becoming an important tool for tunnel construction safety management, playing a crucial role in advancing tunnel construction technology and improving construction management levels.

[0003] Chinese Patent Publication No. CN106593534A discloses an intelligent tunnel construction safety monitoring system, including a deformation monitoring sensor group, a video data acquisition module, an environmental monitoring sensor group, a monitoring point internal stress acquisition module, a data processing module, a central processing unit, a predictive analysis module, an expert evaluation module, a dangerous action judgment module, a mathematical model building module, virtual actuators, virtual sensors, and a human-machine operation module. This invention can comprehensively monitor tunnel construction in real time, thereby obtaining multiple evaluation results of the tunnel construction situation. The detection results are highly accurate, and through the construction of a tunnel construction mathematical model, subsequent tunnel construction situations can be predicted and simulated, enabling timely detection of problems arising during tunnel construction and achieving reasonable selection of tunnel construction plans. It also has automatic assessment functions for dangerous actions and hazardous environments, further ensuring the safety of construction personnel.

[0004] Chinese Patent Publication No. CN113720499A discloses an intelligent tunnel construction safety monitoring system, including a mobile platform, an arched component and a terminal installed on the mobile platform. The arched component abuts against the tunnel top and sidewalls, and its surface is matrix-distributed with several stress detection elements. The stress detection elements are made of elastic material and are arched, allowing them to resonate and sense pressure against the tunnel top and sidewalls. The stress detection elements are connected to the terminal for uploading detection data. The terminal has built-in resonance models, stress models, and tunnel collapse models for the tunnel top and sidewalls under different geological conditions during tunnel excavation and in a static state. When the detected data is greater than normal and approaches the tunnel collapse model, the terminal issues an alarm. By improving the detection structure, the system can continuously detect stress changes within the tunnel during tunnel excavation, thereby achieving dynamic monitoring within the tunnel and improving data reliability.

[0005] However, the following problems still exist in the existing technology.

[0006] In practice, when collecting modeling data for tunnel construction, laser scanning is often used directly, ignoring the fact that uneven tunnel surfaces can affect the accuracy of the collected data. In particular, the presence of moisture or dust in the tunnel air can distort the raw data, leading to errors in the modeling results, biases in the tunnel analysis, and increased construction risks. Summary of the Invention

[0007] To address this issue, the present invention provides an intelligent monitoring system for highway tunnel construction safety. This system solves the problem that, in practice, when collecting modeling data for tunnel construction, most data is collected directly through laser scanning, ignoring the fact that uneven tunnel surfaces can affect the accuracy of the collected data. In particular, the presence of moisture or dust in the tunnel air can distort the collected raw data, leading to errors in the modeling results, deviations in the tunnel analysis, and increased construction risks.

[0008] To achieve the above objectives, the present invention provides an intelligent monitoring system for safety during highway tunnel construction, comprising:

[0009] The data acquisition module includes an image acquisition unit for acquiring images of the tunnel surface, a dust acquisition unit for acquiring dust concentration inside the tunnel, and a humidity acquisition unit for acquiring water vapor humidity inside the tunnel.

[0010] The contour analysis module, which is connected to the data acquisition module, is used to divide the tunnel surface image into several sub-image regions based on impurities, determine the unevenness of the sub-image regions based on the depth region, determine the texture complexity of the sub-image regions based on the crack distribution state, and calculate the region sensitive feature value of each sub-image region.

[0011] The feature fusion module, which is connected to the data acquisition module and the contour analysis module, is used to determine whether the sub-image region meets the preset conditions based on the dust concentration and the water vapor humidity, combined with the region sensitive feature value, calculate the data missing characterization value, and determine the modeling anomaly tendency of each sub-image region.

[0012] The data compensation module, which is connected to the feature fusion module, determines the data acquisition method of the sub-image region based on the modeling anomaly tendency of the sub-image region and the data missing characterization value, and performs modeling.

[0013] Furthermore, the contour analysis module divides the image into several sub-image regions, including:

[0014] Used to identify several impurities in the tunnel surface image, and to determine the center point of each impurity;

[0015] If the distance between the center point of the impurity and the center point of the adjacent impurity is less than a preset distance, then the impurity and the adjacent impurity are divided into the same region.

[0016] This is used to traverse all impurities within the sub-image region, forming several regions;

[0017] Used to determine that the aforementioned regions are several sub-image regions.

[0018] Furthermore, the contour analysis module determines the convexity / concaveness of the sub-image region, including,

[0019] Used to determine the depth region within the sub-image region;

[0020] The ratio of the total area of ​​the depth region to the area of ​​the sub-image region is used to determine the undulation.

[0021] The depth region is the region with a chromaticity greater than a preset chromaticity.

[0022] Furthermore, the contour analysis module determines the texture complexity of the sub-image region, including,

[0023] Used to determine several consecutive texture segments within the sub-image region;

[0024] This is used to vectorize each of the continuous texture segments and determine the variance;

[0025] The inverse of the variance is used to determine the texture complexity.

[0026] Further, the contour analysis module calculates the region-sensitive feature values ​​of each of the sub-image regions, including,

[0027] The first sensitivity factor is used to determine the ratio of the unevenness to the reference unevenness.

[0028] The second sensitivity factor is used to determine the ratio of the texture complexity to the baseline texture complexity;

[0029] This is used to determine the weighted sum of the first and second sensitive factors as the region sensitive feature value.

[0030] Furthermore, the feature fusion module determines whether the sub-image region meets preset conditions, wherein,

[0031] If the sub-image region meets the preset conditions, data acquisition is performed directly;

[0032] If the sub-image region does not meet the preset conditions, the missing data representation value is calculated to determine the modeling anomaly tendency of each sub-image region;

[0033] The preset conditions are that the regional sensitive feature value is less than the benchmark regional sensitive feature value, the dust concentration is less than the benchmark dust concentration, and the water vapor humidity is less than the benchmark water vapor humidity.

[0034] Further, the feature fusion module calculates the missing data representation value, including,

[0035] The first missing factor is used to determine the ratio of the dust concentration to the reference dust concentration;

[0036] The second missing factor is used to determine the ratio of the water vapor humidity to the reference water vapor humidity;

[0037] The ratio of the region-sensitive feature value to the benchmark region-sensitive feature value is used to determine the third missing factor;

[0038] The weighted sum of the first missing factor, the second missing factor, and the third missing factor is used to determine the data missing characterization value.

[0039] Furthermore, the feature fusion module determines the modeling anomaly tendency of each of the sub-image regions, wherein,

[0040] If the data missing characterization value is greater than the preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a strong anomaly tendency.

[0041] If the data missing characterization value is less than or equal to a preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a weak anomaly tendency.

[0042] Furthermore, the data compensation module determines the data acquisition method for the sub-image region, wherein,

[0043] If the modeling anomaly tendency is a weak anomaly tendency, then data compensation is performed on the sub-image region;

[0044] If the modeling anomaly tendency is a strong anomaly tendency, then multiple data collections are performed on the sub-image region to conduct temporary modeling and determine valid data.

[0045] Furthermore, the data compensation module determines valid data, including:

[0046] This is used to construct several temporary models, stack the temporary models, and determine the number of times the same part appears;

[0047] This is used to confirm that data whose occurrence count is greater than a predetermined number is valid data.

[0048] Compared with existing technologies, this invention includes a data acquisition module, a contour analysis module, a feature fusion module, and a data compensation module. By acquiring images of the tunnel surface, dust concentration within the tunnel, and water vapor humidity within the tunnel, it divides the tunnel into several sub-image regions. The unevenness and texture complexity of each sub-image region are determined, and region-sensitive feature values ​​are calculated. The dust concentration and water vapor humidity within the tunnel, combined with the region-sensitive feature values, are used to determine whether each sub-image region meets preset conditions. Data missing characterization values ​​are calculated to determine the modeling anomaly tendency of each sub-image region. Based on the modeling anomaly tendency of the sub-image regions, combined with the data missing characterization values, the data acquisition method for each sub-image region is determined, and modeling is performed. This invention improves the accuracy and efficiency of model building by comprehensively analyzing the tunnel environment, performing data acquisition and compensation.

[0049] In particular, the tunnel surface image is decomposed into several sub-image regions. These sub-image regions may or may not contain impurities. It is understandable that the region sensitivity feature value corresponding to the sub-image region without impurities approaches 0 infinitely. In this case, even if the data is slightly missing on a smooth surface, it is easy to recover it through interpolation. However, for regions with uneven surfaces or mottled textures, even a small data loss during actual data acquisition may lead to the loss of key features or misjudgment, thereby reducing the reliability and integrity of the data and affecting subsequent modeling and analysis. Based on this, this invention calculates the region sensitivity feature value. The larger the region sensitivity feature value, the more significant the combined interference from complex surface morphology and texture factors on the tunnel-related data collected in that region. By calculating the region sensitivity feature value, a quantitative data basis is provided for the subsequent modeling anomaly tendency analysis for each sub-image region, improving the accuracy of subsequent modeling and analysis, and ensuring construction safety.

[0050] In particular, analyzing the dust concentration and water vapor humidity within the tunnel provides a data foundation for subsequent calculations of missing data representation values. In practice, during tunnel data collection, differences in dust particle size, shape, and refractive index cause measurement point offsets, introducing errors. For example, high dust concentration within the tunnel leads to laser light scattering, with dust particles altering the laser's propagation direction, resulting in data acquisition location deviations and data errors. Simultaneously, the non-uniform distribution of water vapor molecules causes anisotropic scattering of the laser during propagation, and the strong absorption characteristics of water molecules at specific wavelengths lead to signal attenuation. For instance, high water vapor humidity within the tunnel causes laser scattering and attenuation. Furthermore, in high humidity environments, water vapor condenses into small droplets, further exacerbating the scattering effect and increasing data errors. Therefore, this invention considers dust concentration and water vapor humidity to provide a theoretical and mathematical basis for calculating missing data representation values, determining subsequent data acquisition methods, improving the accuracy of subsequent modeling and analysis, and ensuring construction safety.

[0051] In particular, by comprehensively analyzing the regional sensitive feature values ​​along with the dust concentration and water vapor humidity within the tunnel, the missing data characterization value is calculated to determine the modeling anomaly tendency of each sub-image region. In practice, when data is collected using lasers, individual dust concentration, water vapor humidity, or complex contours can all affect the accuracy of data collection. In this case, if the tunnel surface contour is complex and the environmental water vapor humidity is high, the error will increase exponentially. For example, high humidity and uneven tunnel surfaces can amplify the tunnel modeling error to 4-6 times that of a normal environment. It is understandable that if the same data collection method is used for both single-factor and multi-factor influences, data loss in some areas and discrepancies between the collected data and the actual tunnel data will occur. Based on this, this invention analyzes the modeling anomaly tendency from both the tunnel itself and the environment, and specifically determines the data collection methods for high and low anomaly tendencies to ensure the comprehensiveness and accuracy of data collection. This provides precise data basis for subsequent mathematical modeling, improves the accuracy of modeling and analysis, and ensures construction safety. Attached Figure Description

[0052] Figure 1 A schematic diagram of the intelligent monitoring system for safety monitoring during highway tunnel construction, as described in an embodiment of the invention.

[0053] Figure 2 This is a schematic diagram of the sub-image region boundary division according to an embodiment of the invention;

[0054] Figure 3 A logic block diagram for determining whether the sub-image region meets preset conditions in an embodiment of the invention;

[0055] Figure 4 A logic block diagram for determining the modeling anomaly tendency of each of the sub-image regions in an embodiment of the invention;

[0056] Figure 5 This is a logic block diagram illustrating the data acquisition method for determining the sub-image region in an embodiment of the invention.

[0057] Among them, 2.1, 2.2, 2.3, and 2.4 are all sub-image regions. Detailed Implementation

[0058] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0059] It should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0060] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0061] Please see Figure 1 , Figure 1 This is a schematic diagram of the intelligent monitoring system for highway tunnel construction safety according to an embodiment of the invention. The intelligent monitoring system for highway tunnel construction safety according to the invention includes:

[0062] The data acquisition module includes an image acquisition unit for acquiring images of the tunnel surface, a dust acquisition unit for acquiring dust concentration inside the tunnel, and a humidity acquisition unit for acquiring water vapor humidity inside the tunnel.

[0063] The contour analysis module, which is connected to the data acquisition module, is used to divide the tunnel surface image into several sub-image regions based on impurities, determine the unevenness of the sub-image regions based on the depth region, determine the texture complexity of the sub-image regions based on the crack distribution state, and calculate the region sensitive feature value of each sub-image region.

[0064] The feature fusion module, which is connected to the data acquisition module and the contour analysis module, is used to determine whether the sub-image region meets the preset conditions based on the dust concentration and the water vapor humidity, combined with the region sensitive feature value, calculate the data missing characterization value, and determine the modeling anomaly tendency of each sub-image region.

[0065] The data compensation module, which is connected to the feature fusion module, determines the data acquisition method of the sub-image region based on the modeling anomaly tendency of the sub-image region and the data missing characterization value, and performs modeling.

[0066] Specifically, there are no restrictions on the specific devices for the image acquisition unit, dust acquisition unit, and humidity acquisition unit. In practice, the image acquisition unit can be a high-definition camera, the dust acquisition unit can be a particle concentration sensor set at a predetermined distance, and the humidity acquisition unit can be a humidity sensor set at a predetermined distance, wherein the predetermined distance is 0.2 times the total length of the tunnel.

[0067] Specifically, the contour analysis module divides the image into several sub-image regions, including:

[0068] Used to identify several impurities in the tunnel surface image, and to determine the center point of each impurity;

[0069] If the distance between the center point of the impurity and the center point of the adjacent impurity is less than a preset distance, then the impurity and the adjacent impurity are divided into the same region.

[0070] This is used to traverse all impurities within the sub-image region, forming several regions;

[0071] Used to determine that the aforementioned regions are several sub-image regions.

[0072] Specifically, there is no limitation on the method of determining impurities. For example, in practice, if the color of a closed shape is different from that of its adjacent shapes, then the closed shape is determined to be an impurity. This will not be elaborated further.

[0073] Specifically, there is no limitation on the method of determining the center point. In practice, the closed figure can be decomposed into several triangles, and the center point can be obtained by weighting the triangles by their areas. This will not be elaborated further.

[0074] Please see Figure 2 , Figure 2 This is a schematic diagram illustrating the sub-image region boundary division according to an embodiment of the invention. Specifically, the sub-image region has region boundaries, and the method for determining these boundaries is not limited. In practice, the method for determining the region boundaries is as follows:

[0075] Impurities that are grouped into the same region are identified as impurity groups;

[0076] Based on OpenCV, the smallest outer region of the impurity group is determined as the region boundary.

[0077] Specifically, the preset distance is 0.1 times the longest side of the tunnel surface image. Of course, those skilled in the art can also determine the preset distance according to the actual situation, as long as it is reasonable, which will not be elaborated here.

[0078] Specifically, the contour analysis module determines the convexity / concaveness of the sub-image region, including,

[0079] Used to determine the depth region within the sub-image region;

[0080] The ratio of the total area of ​​the depth region to the area of ​​the sub-image region is used to determine the undulation.

[0081] The depth region is the region with a chromaticity greater than a preset chromaticity.

[0082] Specifically, in implementation, the preset chromaticity is determined to be the chromaticity corresponding to the tunnel surface image without impurities.

[0083] Specifically, the contour analysis module determines the texture complexity of the sub-image region, including,

[0084] Used to determine several consecutive texture segments within the sub-image region;

[0085] This is used to vectorize each of the continuous texture segments and determine the variance;

[0086] The inverse of the variance is used to determine the texture complexity.

[0087] Specifically, the contour analysis module calculates the region-sensitive feature values ​​for each of the sub-image regions, including:

[0088] The first sensitivity factor is used to determine the ratio of the unevenness to the reference unevenness.

[0089] The second sensitivity factor is used to determine the ratio of the texture complexity to the baseline texture complexity;

[0090] This is used to determine the weighted sum of the first and second sensitive factors as the region sensitive feature value.

[0091] Specifically, the baseline undulation is calculated in advance, and the historical undulation of several tunnel surface images is recorded in advance. The average of the historical undulation is determined as the baseline undulation.

[0092] Specifically, the baseline texture complexity is calculated in advance. The historical texture complexity of several tunnel surface images is recorded in advance, and the average of the historical texture complexity is determined as the baseline texture complexity.

[0093] Specifically, the sum of the weight coefficients of the first sensitive factor and the second sensitive factor is 1. When weighting the first sensitive factor and the second sensitive factor, since the unevenness has a greater impact on the accuracy of data acquisition, the weight coefficient of the first sensitive factor is greater than the weight coefficient of the second sensitive factor. In practice, the weight coefficient of the first sensitive factor is determined to be 0.6 and the weight coefficient of the second sensitive factor is determined to be 0.4.

[0094] Specifically, the tunnel surface image is decomposed into several sub-image regions. These sub-image regions may or may not contain impurities. It is understood that the region sensitivity feature value corresponding to a sub-image region without impurities approaches 0. In this case, even if the data is slightly missing on a smooth surface, it can be easily recovered through interpolation. However, for regions with uneven surfaces or mottled textures, even a small data loss during actual data acquisition can lead to the loss or misjudgment of key features, thereby reducing the reliability and integrity of the data and affecting subsequent modeling and analysis. Based on this, this invention calculates the region sensitivity feature value. The larger the region sensitivity feature value, the more significant the combined interference from complex surface morphology and texture factors on the tunnel-related data collected in that region. By calculating the region sensitivity feature value, a quantitative data foundation is provided for subsequent modeling anomaly tendency analysis of each sub-image region, improving the accuracy of subsequent modeling and analysis, and ensuring construction safety.

[0095] Please see Figure 3 , Figure 3 This is a logic block diagram illustrating how to determine whether a sub-image region meets preset conditions, according to an embodiment of the invention. Specifically, the feature fusion module determines whether the sub-image region meets the preset conditions, wherein...

[0096] If the sub-image region meets the preset conditions, data acquisition is performed directly;

[0097] If the sub-image region does not meet the preset conditions, the missing data representation value is calculated to determine the modeling anomaly tendency of each sub-image region;

[0098] The preset conditions are that the regional sensitive feature value is less than the benchmark regional sensitive feature value, the dust concentration is less than the benchmark dust concentration, and the water vapor humidity is less than the benchmark water vapor humidity.

[0099] Specifically, the sensitive characteristic value of the benchmark area is calculated in advance, the historical sensitive characteristic values ​​of the area during the construction process are recorded in advance, and the average value of the historical sensitive characteristic values ​​is determined as the sensitive characteristic value of the benchmark area.

[0100] Specifically, the baseline dust concentration is calculated in advance, and the historical dust concentration during the construction process is recorded in advance. The average of the historical dust concentration is determined as the baseline dust concentration.

[0101] Specifically, the baseline water vapor humidity is calculated in advance, and the historical water vapor humidity during the construction process is recorded in advance. The average value of the historical water vapor humidity is determined as the baseline water vapor humidity.

[0102] Specifically, this invention analyzes the dust concentration and water vapor humidity within the tunnel environment to provide a data foundation for subsequent calculations of missing data representation values. In practice, during tunnel data collection, differences in dust particle size, shape, and refractive index cause measurement point offsets, introducing errors. For example, high dust concentration in the tunnel can lead to laser light scattering, with dust particles altering the laser's propagation direction, resulting in data acquisition location deviations and data errors. Simultaneously, the non-uniform distribution of water vapor molecules causes anisotropic scattering of the laser during propagation, and the strong absorption characteristics of water molecules at specific wavelengths lead to signal attenuation. For instance, high humidity in the tunnel can cause laser scattering and attenuation. Furthermore, in high humidity environments, water vapor condenses into small droplets, further exacerbating the scattering effect and increasing data errors. Therefore, this invention considers dust concentration and water vapor humidity to provide a theoretical and mathematical basis for calculating missing data representation values, determining subsequent data collection methods, improving the accuracy of subsequent modeling and analysis, and ensuring construction safety.

[0103] Specifically, the feature fusion module calculates missing data representation values, including:

[0104] The first missing factor is used to determine the ratio of the dust concentration to the reference dust concentration;

[0105] The second missing factor is used to determine the ratio of the water vapor humidity to the reference water vapor humidity;

[0106] The ratio of the region-sensitive feature value to the benchmark region-sensitive feature value is used to determine the third missing factor;

[0107] The weighted sum of the first missing factor, the second missing factor, and the third missing factor is used to determine the data missing characterization value.

[0108] Specifically, the sum of the weight coefficients of the first missing factor, the second missing factor, and the third missing factor is 1. When adjusting the weight coefficients, since the regional sensitive feature value has a more serious impact on the data than dust concentration and water vapor humidity, the weight coefficient of the first missing factor is determined to be 0.3, the weight coefficient of the second missing factor is 0.3, and the weight coefficient of the third missing factor is 0.4.

[0109] Please refer to example 4. Figure 4This is a logic block diagram illustrating the modeling anomaly tendency of each of the sub-image regions in an embodiment of the invention. Specifically, the feature fusion module determines the modeling anomaly tendency of each of the sub-image regions, wherein...

[0110] If the data missing characterization value is greater than the preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a strong anomaly tendency.

[0111] If the data missing characterization value is less than or equal to a preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a weak anomaly tendency.

[0112] Specifically, the purpose of setting a data missing characterization threshold is to classify the modeling anomaly tendency through a numerical value. The data missing characterization threshold is calculated in advance, and several historical data missing characterization values ​​corresponding to the construction process are recorded in advance. The average value of the historical data missing characterization values ​​is determined as the data missing baseline value under normal circumstances. The data missing characterization threshold is set as a predetermined multiple of the data missing baseline value. Usually, in order to characterize the case of strong data missing, the predetermined multiple is selected in the range [1.2, 1.4]. In practice, 1.3 is preferred.

[0113] Specifically, by comprehensively analyzing regional sensitive feature values ​​along with dust concentration and moisture humidity within the tunnel, the missing data characterization value is calculated to determine the modeling anomaly tendency of each sub-image region. In practice, when data is collected using lasers, individual dust concentration, moisture humidity, or complex contours can all affect the accuracy of data collection. For example, if the tunnel surface contour is complex and the ambient moisture humidity is high, the error will increase exponentially. High humidity combined with an uneven tunnel surface can amplify the tunnel modeling error to 4-6 times that of a normal environment. It is understandable that using the same data collection method for both single-factor and multi-factor influences can lead to missing data in some areas, resulting in data that does not match the actual tunnel data. Therefore, this invention analyzes modeling anomaly tendencies from both the tunnel itself and the environment, and specifically determines data collection methods for high and low anomaly tendencies to ensure comprehensive and accurate data collection. This provides precise data for subsequent mathematical modeling, improves modeling accuracy and analytical precision, and ensures construction safety.

[0114] Please see Figure 5 , Figure 5 This is a logic block diagram illustrating the determination of the data acquisition method for the sub-image region according to an embodiment of the invention. Specifically, the data compensation module determines the data acquisition method for the sub-image region, wherein...

[0115] If the modeling anomaly tendency is a weak anomaly tendency, then data compensation is performed on the sub-image region;

[0116] If the modeling anomaly tendency is a strong anomaly tendency, then multiple data collections are performed on the sub-image region to conduct temporary modeling and determine valid data.

[0117] Understandably, if the modeling anomaly tendency of the sub-image region is weak, data compensation can be directly performed based on the historical data correspondence table. The historical data correspondence table is pre-built, and for abnormal data corresponding to the data missing characterization value threshold less than or equal to the preset data missing characterization value threshold, normal data corresponding to the abnormal data is pre-stored. The abnormal data is directly replaced to complete the compensation.

[0118] Specifically, the number of data collections is positively correlated with the data missing characterization value.

[0119] Specifically, the data compensation module determines valid data, including:

[0120] This is used to construct several temporary models, stack the temporary models, and determine the number of times the same part appears;

[0121] This is used to confirm that data whose occurrence count is greater than a predetermined number is valid data.

[0122] Specifically, in implementation, "identical parts" means that the models can overlap when stacked.

[0123] Specifically, the number of scheduled collections is 0.7 times the number of collections.

[0124] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0125] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A smart monitoring system for safety during highway tunnel construction, characterized in that, include: The data acquisition module includes an image acquisition unit for acquiring images of the tunnel surface, a dust acquisition unit for acquiring dust concentration inside the tunnel, and a humidity acquisition unit for acquiring water vapor humidity inside the tunnel. The contour analysis module, which is connected to the data acquisition module, is used to divide the tunnel surface image into several sub-image regions based on impurities, determine the unevenness of the sub-image regions based on the depth region, determine the texture complexity of the sub-image regions based on the crack distribution state, and calculate the region sensitive feature value of each sub-image region. The feature fusion module, which is connected to the data acquisition module and the contour analysis module, is used to determine whether the sub-image region meets the preset conditions based on the dust concentration and the water vapor humidity, combined with the region sensitive feature value, calculate the data missing characterization value, and determine the modeling anomaly tendency of each sub-image region. A data compensation module, which is connected to the feature fusion module, determines the data acquisition method of the sub-image region based on the modeling anomaly tendency of the sub-image region and combines the data missing characterization value to perform modeling; The feature fusion module determines whether the sub-image region meets preset conditions, wherein... If the sub-image region meets the preset conditions, data acquisition is performed directly; If the sub-image region does not meet the preset conditions, the missing data representation value is calculated to determine the modeling anomaly tendency of each sub-image region; The preset conditions are that the regional sensitive feature value is less than the reference regional sensitive feature value, the dust concentration is less than the reference dust concentration, and the water vapor humidity is less than the reference water vapor humidity. The feature fusion module calculates the missing data representation value, including: The first missing factor is used to determine the ratio of the dust concentration to the reference dust concentration; The second missing factor is used to determine the ratio of the water vapor humidity to the reference water vapor humidity; The ratio of the region-sensitive feature value to the benchmark region-sensitive feature value is used to determine the third missing factor; The weighted sum of the first missing factor, the second missing factor, and the third missing factor is used to determine the data missing characterization value. The feature fusion module determines the modeling anomaly tendency of each of the sub-image regions, wherein... If the data missing characterization value is greater than the preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a strong anomaly tendency. If the data missing characterization value is less than or equal to the preset data missing characterization value threshold, then the modeling anomaly tendency is determined to be a weak anomaly tendency. The data compensation module determines the data acquisition method for the sub-image region, wherein... If the modeling anomaly tendency is a weak anomaly tendency, then data compensation is performed on the sub-image region; If the modeling anomaly tendency is a strong anomaly tendency, then multiple data collections are performed on the sub-image region to conduct temporary modeling and determine valid data; The data compensation module determines valid data, including: This is used to construct several temporary models, stack the temporary models, and determine the number of times the same part appears; This is used to confirm that data whose occurrence count is greater than a predetermined number is valid data.

2. The intelligent monitoring system for highway tunnel construction safety according to claim 1, characterized in that, The contour analysis module divides the image into several sub-image regions, including: Used to identify several impurities in the tunnel surface image, and to determine the center point of each impurity; If the distance between the center point of the impurity and the center point of the adjacent impurity is less than a preset distance, then the impurity and the adjacent impurity are divided into the same region. This is used to traverse all impurities within the sub-image region, forming several regions; Used to determine that the aforementioned regions are several sub-image regions.

3. The intelligent monitoring system for highway tunnel construction safety according to claim 1, characterized in that, The contour analysis module determines the convexity / concaveness of the sub-image region, including: Used to determine the depth region within the sub-image region; The ratio of the total area of ​​the depth region to the area of ​​the sub-image region is used to determine the undulation. The depth region is the region with a chromaticity greater than a preset chromaticity.

4. The intelligent monitoring system for highway tunnel construction safety according to claim 1, characterized in that, The contour analysis module determines the texture complexity of the sub-image region, including: Used to determine several consecutive texture segments within the sub-image region; This is used to vectorize each of the continuous texture segments and determine the variance; The inverse of the variance is used to determine the texture complexity.

5. The intelligent monitoring system for highway tunnel construction safety according to claim 1, characterized in that, The contour analysis module calculates the region-sensitive feature values ​​for each of the sub-image regions. include, The first sensitivity factor is used to determine the ratio of the unevenness to the reference unevenness. The second sensitivity factor is used to determine the ratio of the texture complexity to the baseline texture complexity; This is used to determine the weighted sum of the first and second sensitive factors as the region sensitive feature value.

Images