Tunnel bridge monitoring management system and method

Abstract

The invention relates to the field of tunnel and bridge monitoring, and particularly discloses a tunnel and bridge monitoring management system and method.The method comprises the steps that firstly, tunnel and bridge road surface monitoring images collected by cameras and road surface stress data, collected by fiber grating sensors, at multiple time points are obtained, and then the deep learning technology is utilized; and feature extraction and correlation analysis are carried out on the two images, and finally a classification result is obtained through a classifier so as to judge whether tunnel and bridge pavement early warning is sent out, so that abnormal conditions and potential problems of the tunnel and bridge pavement can be found in time, immediate response is realized, early warning and preventive maintenance are realized, and the accident risk and the maintenance cost are further reduced.

Description

Technical Field

[0001] The present application relates to the field of tunnel and bridge monitoring, and more specifically, to a tunnel and bridge monitoring management system and method. Background Art

[0002] Pavement tunnel bridges refer to those traffic structures that can both pass through the ground and cross obstacles. They are usually designed for use in cities or highway systems, and include both underground tunnels and elevated bridges. Pavement tunnels provide vehicles with a channel for rapid movement within the city, effectively reducing traffic congestion and improving traffic efficiency; while bridges connect different regions by crossing rivers, canyons or other terrain obstacles, greatly facilitating traffic flow and urban development. Their design and construction must take into account geological conditions, traffic flow, environmental impact and engineering safety, and are an indispensable and important part of modern urban and highway networks.

[0003] The concrete inside the road surface, tunnel and bridge will crack over time due to various reasons. When some cracks reach the corresponding depth and width, it will bring danger to the current traffic. Generally, road surfaces, tunnels and bridges need to be inspected regularly, but they cannot be monitored normally. Once a problem occurs, it cannot be handled in time, which may cause a major accident. In addition, for the road surface with problems, it is generally necessary to manually find the corresponding location and then manually mark it, which is time-consuming, labor-intensive and inefficient.

[0004] Therefore, a tunnel and bridge monitoring management system and method are desired. Summary of the invention

[0005] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides a tunnel bridge monitoring management system and method, which first obtains the tunnel bridge pavement monitoring image collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally obtains the classification result through the classifier to determine whether to issue a tunnel bridge pavement warning, so as to timely discover the abnormal conditions and potential problems of the tunnel bridge pavement, realize instant response, thereby realizing early warning and preventive maintenance, thereby reducing accident risks and maintenance costs.

[0006] According to one aspect of the present application, a tunnel and bridge monitoring and management system is provided, which includes:

[0007] A tunnel and bridge monitoring data acquisition module is used to acquire tunnel and bridge pavement monitoring images acquired by cameras and pavement stress data at multiple time points acquired by fiber grating sensors;

[0008] A tunnel and bridge monitoring data extraction module, used to extract tunnel and bridge pavement texture enhancement feature vectors and optimized pavement stress space correlation feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor;

[0009] The tunnel and bridge pavement warning judgment module is used to judge whether to issue a tunnel and bridge pavement warning based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector.

[0010] According to another aspect of the present application, a tunnel and bridge monitoring management method is provided, which includes:

[0011] Acquire monitoring images of tunnel and bridge pavement collected by cameras and pavement stress data at multiple time points collected by fiber grating sensors;

[0012] Extracting tunnel and bridge pavement texture enhancement feature vectors and optimized pavement stress space correlation feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor;

[0013] Based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector, it is determined whether to issue a tunnel and bridge pavement warning.

[0014] Compared with the prior art, the present application provides a tunnel and bridge monitoring management system and method, which first obtains the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally obtains the classification result through the classifier to determine whether to issue a tunnel and bridge pavement warning, so as to promptly discover abnormal conditions and potential problems in the tunnel and bridge pavement, and achieve immediate response, thereby achieving early warning and preventive maintenance, thereby reducing accident risks and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 It is a block diagram of a tunnel and bridge monitoring and management system according to an embodiment of the present application.

[0017] Figure 2It is a block diagram of a tunnel and bridge monitoring data extraction module in a tunnel and bridge monitoring and management system according to an embodiment of the present application.

[0018] Figure 3 It is a block diagram of a tunnel and bridge pavement image feature extraction unit in a tunnel and bridge monitoring and management system according to an embodiment of the present application.

[0019] Figure 4 It is a block diagram of a tunnel and bridge pavement early warning judgment module in a tunnel and bridge monitoring and management system according to an embodiment of the present application.

[0020] Figure 5 It is a flow chart of the tunnel and bridge monitoring and management method according to an embodiment of the present application.

[0021] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] Figure 1 FIG. 1 is a block diagram of a tunnel and bridge monitoring and management system according to an embodiment of the present application. Figure 1 As shown, according to the tunnel and bridge monitoring and management system 100 of the embodiment of the present application, it includes: a tunnel and bridge monitoring data acquisition module 110, which is used to obtain the tunnel and bridge pavement monitoring image collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; a tunnel and bridge monitoring data extraction module 120, which is used to extract the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector from the tunnel and bridge pavement monitoring image collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; a tunnel and bridge pavement early warning judgment module 130, which is used to judge whether to issue a tunnel and bridge pavement early warning based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector.

[0024] In the above-mentioned tunnel and bridge monitoring management system 100, the tunnel and bridge monitoring data acquisition module 110 is used to obtain the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor. It should be understood that pavement tunnels and bridges are important types of traffic buildings. They can not only pass through the city underground, but also cross various natural and man-made obstacles, such as rivers and highlands. These structures play a key role in urban and highway systems, including not only underground tunnels but also elevated bridges, providing vehicles with unimpeded passage. Tunnels provide fast-moving channels within the city, effectively alleviating traffic congestion and improving overall traffic efficiency. Bridges connect different areas by crossing geographical obstacles such as rivers and valleys, greatly promoting traffic flow and urban development. The design and construction of pavement tunnels and bridges must consider multiple factors, including geological conditions, traffic flow, environmental impact, and engineering safety. They are not only key components of modern cities and highway networks, but also important supports for urban life and economic activities. However, over time, the concrete inside the pavement tunnel bridge may crack due to a variety of reasons. When these cracks reach a certain depth and width, they may pose a danger to passing vehicles. The current conventional practice is to regularly detect the state of the road surface, but real-time data is usually not monitored. Once a problem occurs, manual intervention is often required to find the location of the cracks and mark them. This process is time-consuming, labor-intensive and inefficient. Therefore, in the technical solution of the present application, by obtaining the monitoring images of the tunnel and bridge pavement collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor, and combining deep learning technology to determine whether to issue a tunnel and bridge pavement warning, abnormal conditions and potential problems of the tunnel and bridge pavement can be discovered in time, and rapid response can be achieved to achieve early warning and preventive maintenance, thereby reducing accident risks and maintenance costs. This system can accurately determine whether a pavement warning is needed and monitor the state changes of the pavement structure through real-time data. This ensures that necessary measures are taken before problems occur to ensure the safe operation and sustainable development of tunnels and bridges.

[0025] Specifically, the camera can provide high-definition monitoring images. Through image feature extraction technology, it can analyze the texture, cracks and deformation of the road surface, providing intuitive data for evaluating the stability and safety of the road surface structure. The fiber grating sensor can collect the stress distribution data of the road surface in real time. Through monitoring at multiple time points, it can detect and analyze the changing trend of the road surface stress, helping to predict and identify potential structural problems or load conditions. The comprehensive analysis of these two types of data can not only effectively improve the real-time monitoring capabilities of tunnels and bridges, but also provide timely decision-making support for engineering managers, ensure structural safety and operational stability, and thus ensure the smooth flow of road traffic and the safety of users.

[0026] In the above-mentioned tunnel and bridge monitoring and management system 100, the tunnel and bridge monitoring data extraction module 120 is used to extract the tunnel and bridge pavement texture enhancement feature vector and optimize the pavement stress space correlation feature vector from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor. It should be understood that extracting feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data collected by the fiber grating sensor is aimed at deeply understanding and evaluating the structural state and health status of the tunnel and bridge. Extracting the pavement texture enhancement feature vector and optimizing the pavement stress space correlation feature vector not only provides an effective means for real-time monitoring of tunnels and bridges, but also provides detailed data support for engineering management personnel, helping them to promptly identify and resolve possible safety hazards, thereby ensuring the long-term stable operation and safety of tunnels and bridges.

[0027] Figure 2 FIG. 4 is a block diagram of a tunnel and bridge monitoring data extraction module in a tunnel and bridge monitoring and management system according to an embodiment of the present application. Figure 2 As shown, in a specific embodiment of the present application, the tunnel bridge monitoring data extraction module 120 includes: a tunnel bridge pavement image feature extraction unit 121, which is used to extract features from the tunnel bridge pavement monitoring image collected by the camera to obtain the tunnel bridge pavement texture enhancement feature vector; a pavement stress data preprocessing unit 122, which is used to preprocess the pavement stress data of multiple time points collected by the fiber grating sensor to obtain a pavement stress correlation matrix; a pavement stress feature encoding unit 123, which is used to feature encode the pavement stress correlation matrix to obtain the optimized pavement stress space correlation feature vector.

[0028] It should be understood that the pavement texture features in the monitoring images, such as cracks, wear, patches, etc., are important indicators of the health of the pavement. The feature extraction process first involves image preprocessing, including grayscale, filtering and enhancement to reduce noise and enhance the contrast of the image. Then, using computer vision technology, specific feature extraction algorithms such as gray-level co-occurrence matrix and local binary pattern are used to extract numerical descriptors representing the pavement texture features from the image. These feature descriptors are integrated into pavement texture enhancement feature vectors, which specifically include information such as texture density, directionality and spatial distribution. These vectors not only reflect the roughness and overall texture of the pavement surface, but also identify minor changes and damage, such as microcracks and local deformation, providing timely feedback and early warning for engineering managers. By comparing historical data or standard models, the system can further analyze and predict the life and health of the pavement, providing data support for maintenance and repair work.

[0029] Furthermore, the pavement stress data collected by the fiber Bragg grating sensor at multiple time points are preprocessed to comprehensively analyze and understand the mechanical response characteristics of the pavement structure, so as to evaluate its safety and stability. First, the pavement stress data is collected in real time by the fiber Bragg grating sensor, which records the stress distribution and changes of the pavement at different time points. The preprocessing process first includes data cleaning and correction to eliminate possible measurement errors and interference to ensure the accuracy and reliability of the data. Secondly, the continuously collected stress data are organized into a pavement stress correlation matrix through time series analysis and statistical methods. This matrix reflects the spatial distribution and change trend of pavement stress, including stress components and magnitudes at different positions and directions. The preprocessing process may also include data interpolation and smoothing to fill in data gaps or reduce noise interference, so that the final matrix has higher data integrity and interpretability.

[0030] Furthermore, the pavement stress correlation matrix is ​​feature encoded to obtain the optimized pavement stress spatial correlation feature vector, aiming to further extract and express the stress distribution characteristics of the pavement structure to support structural health monitoring and safety assessment. Among them, the pavement stress correlation matrix is ​​integrated by multi-time point data collected by fiber grating sensors, reflecting the stress state and change trend of the pavement at different positions and directions. The feature encoding process not only improves the information density and explanatory power of pavement stress data, but also enhances its application value in real-time monitoring and engineering decision-making, ensuring the long-term safe operation and service performance of tunnel and bridge structures.

[0031] Figure 3 FIG. 1 is a block diagram of a tunnel bridge pavement image feature extraction unit in a tunnel bridge monitoring and management system according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the tunnel bridge pavement image feature extraction unit 121 includes: a pavement monitoring image texture feature encoding subunit 1211, which is used to encode the texture features of the tunnel bridge pavement monitoring image collected by the camera to obtain a tunnel bridge pavement texture enhancement feature map; a pavement monitoring image texture feature pooling subunit 1212, which is used to pool the tunnel bridge pavement texture enhancement feature map to obtain the tunnel bridge pavement texture enhancement feature vector.

[0032] It should be understood that the pavement texture features in the monitoring image include cracks, wear, patches, etc. These features directly reflect the use and damage of the pavement. The process of texture feature encoding involves image processing and computer vision technology. First, image preprocessing, such as grayscale, filtering and enhancement, is performed to reduce noise and enhance image contrast. Specific texture feature extraction algorithms, such as grayscale co-occurrence matrix, local binary pattern, etc., are used to extract numerical descriptors representing the pavement texture from the processed image. These descriptors can include texture roughness, directionality, local shape and other features. Through spatial distribution and statistical analysis, a pavement texture enhancement feature map is constructed. This feature map not only presents detailed texture information on the pavement surface, but also quantifies texture changes in different regions and time points, providing accurate visual data support for engineering management personnel.

[0033] Furthermore, the texture enhancement feature map of the tunnel and bridge pavement is pooled to further extract and compress the key features in the image so as to more effectively describe and compare the texture changes and damage of the pavement. First, the texture enhancement feature map is generated by the aforementioned texture feature encoding technology, which reflects the texture features of different areas of the pavement, such as roughness, directionality, and local shape. These feature maps are usually high-dimensional and complex, containing a large amount of pixel-level information. The pooling operation aggregates and simplifies the information in the feature map through certain rules and algorithms to generate a relatively small feature vector that retains key features. Common pooling methods include maximum pooling and average pooling, which can effectively reduce the dimension of the data and extract the main features in the feature map. For example, in maximum pooling, the maximum value in each pooling window is selected as the representative value of the area, thereby retaining the most significant local features. Average pooling calculates the average of the pixel values ​​in the pooling window to generate a smoother and more consistent feature description. Through pooling, the feature vector obtained from the complex texture enhancement feature map not only has a lower dimension, but also can more effectively express the main features of the entire image. These feature vectors can be directly used for classification, identification and evaluation of pavement conditions, such as identifying different types of pavement damage or evaluating the overall quality of the pavement.

[0034] In a specific embodiment of the present application, the road surface monitoring image texture feature encoding subunit 1211 includes: passing the tunnel and bridge road surface monitoring image collected by the camera through a grayscale co-occurrence matrix to obtain a tunnel and bridge road surface texture enhancement feature map; passing the tunnel and bridge road surface texture enhancement feature map through a tunnel and bridge road surface texture enhancement convolutional neural network based on a feature encoder to obtain the tunnel and bridge road surface texture enhancement feature map.

[0035] It should be understood that the purpose of processing the tunnel and bridge pavement monitoring images collected by the camera through the gray level co-occurrence matrix to obtain the pavement texture enhancement feature map is to deeply analyze and quantify the texture characteristics of the pavement surface, which directly reflect the damage, wear and structural health of the pavement. First, the monitoring images are pre-processed, such as graying and denoising, to eliminate unnecessary details and noise to ensure the stability and reliability of image quality. The gray level co-occurrence matrix is ​​a classic texture analysis tool that reveals the spatial relationship between pixels in the image by counting the frequency of occurrence of grayscale pairs between pixels. When processing pavement monitoring images, the gray level co-occurrence matrix calculates the co-occurrence probability of different pixel gray levels in the same image area. These co-occurrence probabilities reflect different texture features in the image, such as particle size, directionality and roughness. By adjusting the parameters of the matrix, such as distance and angle, the pavement texture information at different scales and directions can be captured, thus forming a more comprehensive and detailed feature description.

[0036] Furthermore, the monitoring images of tunnel and bridge pavement are transformed into texture-enhanced feature maps through techniques such as gray-level co-occurrence matrix, which reflect the texture details and changes of different areas of the pavement. These feature maps are usually high-dimensional and contain a lot of information. The feature encoder-based CNN uses deep learning methods to extract feature representations with a higher level of abstraction from complex texture-enhanced feature maps. The encoder part usually consists of multiple layers of convolutional layers and pooling layers, which gradually reduce the spatial size of the feature map and extract the most significant features. During the learning process, these feature encoders can capture important patterns and structural information in the pavement texture, such as cracks, patches, and wear, through a large amount of supervised or unsupervised learning. Through the deconvolution layer or upsampling layer of the convolutional neural network, the encoded high-level feature map is restored to the same size as the original input to form a new pavement texture-enhanced feature map. These feature maps not only retain the spatial and structural information of the original image, but also enhance the expression and resolution of important features in the image, making the analysis of pavement texture more comprehensive and refined. Specifically, each layer of the tunnel and bridge pavement texture enhancement convolutional neural network based on the feature encoder performs convolution processing, mean pooling processing based on the local feature matrix and nonlinear activation processing on the input data in the forward transfer of the layer, so that the last layer of the tunnel and bridge pavement texture enhancement convolutional neural network based on the feature encoder outputs the tunnel and bridge pavement texture enhancement feature map, wherein the input of the tunnel and bridge pavement texture enhancement convolutional neural network based on the feature encoder is the tunnel and bridge pavement texture enhancement feature map.

[0037] In a specific embodiment of the present application, the pavement stress data preprocessing unit 122 includes: arranging the pavement stress data at multiple time points collected by the fiber grating sensor according to the time dimension to obtain a pavement stress input vector; multiplying the pavement stress input vector by its transpose to obtain the pavement stress correlation matrix.

[0038] It should be understood that the stress data collected at multiple time points are sorted and timestamped and arranged according to the time dimension. This arrangement ensures the temporal order of the data and reflects the dynamic changes of pavement stress in different time periods. The data at each time point usually includes data from multiple measurement locations or sensors, forming a multi-dimensional data matrix or data set. By integrating the stress data of the time series, a comprehensive analysis of the stress distribution, change trend, and the impact of periodic loads on the pavement structure in different time periods can be achieved.

[0039] Furthermore, the operation of multiplying the pavement stress input vector by its transpose to obtain the pavement stress correlation matrix is ​​intended to analyze and quantify the stress correlation and mutual influence between different measuring points or different time periods, so as to deeply understand the overall response characteristics and systematic changes of the pavement structure. First, the pavement stress input vector is a data set arranged according to the time dimension. Each vector represents the pavement stress data at a specific time point or a specific measuring point, and contains multi-dimensional information. By multiplying each pavement stress input vector by its transpose, a square matrix, namely the pavement stress correlation matrix, can be obtained. In this matrix, each element represents the linear correlation between the corresponding two time points or measuring points. For example, if the stress change trends between two time points or measuring points are similar or there is a certain correlation, then the matrix elements at the corresponding positions will reflect the strength and direction of this relationship. Multiplying the pavement stress input vector by its transpose to obtain the pavement stress correlation matrix is ​​to deeply understand the stress interaction relationship between the internal parts of the pavement structure, and provide an efficient and accurate data analysis tool for the intelligent pavement monitoring and management system, so as to optimize the operation and maintenance efficiency of transportation infrastructure.

[0040] In a specific embodiment of the present application, the pavement stress feature encoding unit 123 includes: passing the pavement stress correlation matrix through a pavement stress spatial convolution coding model to obtain a pavement stress spatial correlation feature map; performing maximum pooling on the pavement stress spatial correlation feature map to obtain a pavement stress spatial correlation feature vector; based on the tunnel and bridge pavement texture enhancement feature vector, performing inverse error perception based on multi-view feature relationship quantization on the pavement stress spatial correlation feature vector to obtain an optimized pavement stress spatial correlation feature vector.

[0041] It should be understood that the pavement stress spatial convolutional coding model is usually based on the deep learning convolutional neural network (CNN) architecture. Through multiple layers of convolutional layers, pooling layers and encoder structures, it can effectively extract and encode the spatial features in the pavement stress data. When processing the pavement stress correlation matrix, the model first performs a spatial convolution operation on the matrix, that is, extracts local features on the matrix by means of a sliding window. These convolution operations enable the model to capture the spatial layout and structural relationships between different measuring points, such as the distance, position and relative stress distribution differences between the measuring points. After being processed by the encoding module, the model further encodes the feature maps obtained by convolution into higher-level representations, such as local patterns in the pavement structure, regional stress change trends, etc. These encoded feature maps have higher abstraction and representation capabilities, and can better describe and understand the complex stress distribution characteristics of the pavement in space. Specifically, each layer of the pavement stress spatial convolution coding model is used to perform the following operations on the input data in the forward transfer of the layer: convolution processing is performed on the input data based on the convolution kernel to generate a convolution feature map; global mean pooling processing is performed on the convolution feature map based on the feature matrix to generate a pooled feature map; nonlinear activation is performed on the eigenvalues ​​of each position in the pooled feature map to generate an activated feature addition map; wherein the output of the last layer of the pavement stress spatial convolution coding model is the pavement stress spatial correlation feature map, the input from the second layer to the last layer of the pavement stress spatial convolution coding model is the output of the previous layer, and the input of the pavement stress spatial convolution coding model is the pavement stress correlation matrix.

[0042] Furthermore, the pavement stress spatial correlation feature map is subjected to maximum pooling, aiming to extract the most significant and representative feature points from the complex spatial feature representation to form the pavement stress spatial correlation feature vector. The maximum pooling operation is implemented by selecting the maximum value in each sub-region of the feature map, which effectively reduces the dimension and data volume of the feature map while retaining the key structural information and spatial correlation characteristics. Before the maximum pooling, the pavement stress spatial correlation feature map usually contains complex spatial distribution information of different regions of the pavement, such as local stress concentration, uneven load distribution, etc. This information reflects the stress changes of the pavement structure at different locations and the interaction between different regions. Through the maximum pooling operation, the maximum value in each feature map sub-region is extracted, which can effectively reduce the complexity and calculation of the data while maintaining the sensitivity to the key features of the pavement structure. These maximum values, as the summarized feature vectors, have a higher level of abstraction and a more compact representation, which can effectively characterize the spatial stress distribution characteristics of the entire pavement structure.

[0043] Furthermore, in the technical solution of the present application, it is considered that the pavement stress data is quantitative data obtained based on physical measurement, while the tunnel bridge pavement texture enhancement feature is qualitative data obtained based on image processing. These two data types essentially reflect different physical phenomena, which may lead to differences in their feature expressions. The pavement stress data extracts features through a spatial convolutional coding model, while the pavement texture features are extracted through a convolutional neural network. Different feature extraction methods may lead to semantic inconsistencies in feature vectors. The pavement stress data is dynamic data that changes over time, while the pavement texture features are extracted from static images. This difference in the time dimension may cause the pavement stress feature vector to be unable to fully reflect the dynamic changes of the pavement texture features. When fusing and optimizing the pavement stress spatial correlation feature vector and the tunnel bridge pavement texture enhancement feature vector, if the fusion strategy does not fully consider the temporal and spatial characteristics of the two, it may lead to the pavement stress spatial correlation feature vector relative to the tunnel bridge pavement texture enhancement feature vector. There is a lack of a posteriori-priori causal association. In order to solve this problem, in the technical solution of the present application, based on the tunnel bridge pavement texture enhancement feature vector, the pavement stress spatial correlation feature vector is subjected to inverse error perception based on multi-view feature relationship quantification.

[0044] Specifically, based on the tunnel and bridge pavement texture enhancement feature vector, the pavement stress space correlation feature vector is subjected to inverse error perception based on multi-view feature relationship quantization to obtain an optimized pavement stress space correlation feature vector, including: calculating the position difference between the pavement stress space correlation feature vector and the tunnel and bridge pavement texture enhancement feature vector to obtain a first viewpoint metric representation vector; calculating the position point multiplication between the pavement stress space correlation feature vector and the tunnel and bridge pavement texture enhancement feature vector to obtain a second viewpoint metric representation vector; calculating the position sum between the pavement stress space correlation feature vector and the tunnel and bridge pavement texture enhancement feature vector to obtain a first viewpoint metric representation vector. Three viewpoint measurement representation vectors; calculating the gradient amplitude of each position in the first viewpoint measurement representation vector, the second viewpoint measurement representation vector and the third viewpoint measurement representation vector to obtain the first viewpoint measurement perception vector, the second viewpoint measurement perception vector and the third viewpoint measurement perception vector; inputting the first viewpoint measurement perception vector, the second viewpoint measurement perception vector and the third viewpoint measurement perception vector into the backward machine learning reverse multi-viewpoint perception optimization module to obtain the reverse multi-viewpoint perception optimization representation vector; calculating the position point multiplication between the reverse multi-viewpoint perception optimization representation vector and the pavement stress space correlation feature vector to obtain the optimized pavement stress space correlation feature vector.

[0045] The backward machine learning reverse multi-view perception optimization module is expressed as follows:

[0046]

[0047] in, and represents the eigenvalues ​​of the i-th position of the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector, respectively, and μ represents the three eigenvalues ​​of the corresponding positions in the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector , σ is the standard deviation of the three eigenvalues ​​at corresponding positions in the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector, represents the maximum value of the three eigenvalues ​​at the corresponding positions in the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector, ReLU represents a linear rectification function, softmax represents a normalized exponential function, and v ri Represents the eigenvalue of the i-th position of the inverse multi-view perceptual optimization representation vector.

[0048] More specifically, in the technical solution of the present application, based on the tunnel bridge pavement texture enhancement feature vector, the pavement stress space correlation feature vector is optimized based on the reverse error perception based on the quantification of multi-view feature relationship, and the process includes firstly calculating the position difference, dot product and sum between the optimized pavement stress space correlation feature vector and the tunnel bridge pavement texture enhancement feature vector through multi-view analysis (Multi-View Analysis) to obtain three viewpoint metric representation vectors, and quantify the relationship between features from different angles. Then, the gradient amplitude of each position in these viewpoint metric representation vectors is calculated through reverse error perception (Backward Error Perception) to identify key change points to provide a basis for feature optimization.

[0049] Then, the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector are input into the backward machine learning reverse multi-viewpoint perception optimization module, wherein the backward machine learning reverse multi-viewpoint perception optimization module captures the significant comprehensive ability factor of the feature importance perception (Feature Importance Perception) of the feature sub-dimension distribution of the corresponding position in the first to third viewpoint metric perception vectors by adopting a squeeze-incentive mechanism composed of -functions. Then, the position point multiplication between the reverse multi-viewpoint perception optimization representation vector and the pavement stress spatial correlation feature vector is calculated to perform backward optimization perception (Backward Optimization Perception) on the optimized pavement stress spatial correlation feature vector to adjust the pavement stress spatial correlation feature vector so as to better associate it with the tunnel bridge pavement texture enhancement feature vector, thereby enhancing the causal relationship between the optimized pavement stress spatial correlation feature vector and the tunnel bridge pavement texture enhancement feature vector.

[0050] In the above-mentioned tunnel and bridge monitoring and management system 100, the tunnel and bridge pavement early warning judgment module 130 is used to determine whether to issue a tunnel and bridge pavement early warning based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector. It should be understood that the texture enhancement feature vector of the tunnel and bridge pavement usually reflects the roughness, uniformity and possible wear degree of the pavement surface. These feature vectors can be obtained by optical image processing or laser scanning and other technologies to describe the actual situation of the pavement surface, such as cracks, unevenness or wear degree. The changes in these feature vectors can indicate possible surface problems of the pavement structure, such as the location and degree of pavement damage. The optimized pavement stress space correlation feature vector provides a comprehensive analysis of the internal stress distribution of the pavement structure. Through the spatial convolution coding model and maximum pooling operation mentioned above, the generated optimized pavement stress space correlation feature vector can accurately reflect the stress distribution characteristics of the pavement in different areas or different time periods. These feature vectors can be used to detect whether there is abnormal stress concentration or uneven load distribution in the pavement structure, which is an important basis for evaluating the health status of the pavement structure. Comprehensively considering the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector, a comprehensive evaluation model can be established. The model can perform real-time assessment and prediction of the health of pavement structures based on historical data and real-time monitoring information. For example, if the pavement texture enhancement feature vector shows obvious signs of wear or damage, combined with the optimized pavement stress spatial correlation feature vector indicating an abnormal stress distribution pattern, the system can automatically issue a pavement warning signal.

[0051] Figure 4FIG. 1 is a block diagram of a tunnel bridge pavement early warning judgment module in a tunnel bridge monitoring and management system according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the tunnel bridge pavement warning judgment module 130 includes: a tunnel bridge pavement feature fusion unit 131, which is used to fuse the tunnel bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector to obtain a tunnel bridge pavement warning judgment feature vector; a pavement warning judgment classification unit 132, which is used to pass the tunnel bridge pavement warning judgment feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a tunnel bridge pavement warning.

[0052] It should be understood that the fusion of tunnel and bridge pavement texture enhancement feature vectors and the optimization of pavement stress spatial correlation feature vectors to obtain tunnel and bridge pavement early warning judgment feature vectors aims to combine the multi-dimensional information of pavement surface status and internal stress distribution, so as to comprehensively evaluate and predict the health status and possible problems of the pavement structure. It is an efficient and reliable pavement monitoring and early warning method that can effectively ensure the safety and sustainable operation of transportation infrastructure.

[0053] Furthermore, the tunnel bridge pavement early warning judgment feature vector combines the pavement texture enhancement characteristics and stress space correlation characteristics, and has the ability to grasp the pavement structure status in many aspects. These feature vectors reflect the key physical properties of the pavement surface and interior, such as the degree of wear, the location of the damaged area, and the unevenness of stress distribution. The classification result directly determines whether to issue a tunnel bridge pavement early warning. After the classifier classifies a specific feature vector, if it is determined that there are potential problems or a high possibility of damage to the pavement, the system will automatically trigger the early warning mechanism and notify the relevant maintenance personnel or monitoring center for further inspection and maintenance. This automated early warning system not only improves the timeliness and accuracy of the early warning, but also effectively reduces the errors and delays caused by human judgment, and ensures the safety and reliability of the tunnel bridge structure.

[0054] In summary, the embodiment of the present application first obtains the tunnel bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the two. Finally, the classification result is obtained through the classifier to determine whether to issue a tunnel bridge pavement warning, so as to promptly discover abnormal conditions and potential problems in the tunnel bridge pavement and achieve immediate response, thereby achieving early warning and preventive maintenance, thereby reducing accident risks and maintenance costs.

[0055] As described above, the tunnel and bridge monitoring and management system 100 according to the embodiment of the present application can be implemented in various terminal devices. In one example, the tunnel and bridge monitoring and management system 100 can be integrated into the terminal device as a software module and / or a hardware module. For example, the tunnel and bridge monitoring and management system 100 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the tunnel and bridge monitoring and management system 100 can also be one of the many hardware modules of the terminal device.

[0056] Alternatively, in another example, the tunnel bridge monitoring and management system 100 and the terminal device may also be separate devices, and the tunnel bridge monitoring and management system 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0057] Figure 5 Flow chart of the tunnel and bridge monitoring management method according to an embodiment of the present application. Figure 5 As shown, according to the tunnel and bridge monitoring and management method of the embodiment of the present application, it includes: S110, acquiring the tunnel and bridge pavement monitoring image collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; S120, extracting the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector from the tunnel and bridge pavement monitoring image collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; S130, judging whether to issue a tunnel and bridge pavement warning based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector.

[0058] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned tunnel and bridge monitoring management method have been described in detail above. Figures 1 to 4 The description of the tunnel and bridge monitoring and management system has been described in detail, and therefore, its repeated description will be omitted.

[0059] Below, reference Figure 6 To describe an electronic device according to an embodiment of the present application.

[0060] Figure 6 A block diagram of an electronic device according to an embodiment of the present application is illustrated.

[0061] like Figure 6As shown, the electronic device 10 includes an input device 11, an input interface 12, a central processing unit 13, a memory 14, an output interface 15, an output device 16, and a bus 17. The input interface 12, the central processing unit 13, the memory 14, and the output interface 15 are connected to each other through the bus 17, and the input device 11 and the output device 16 are connected to the bus 17 through the input interface 14 and the output interface 15, respectively, and then connected to other components of the electronic device 10.

[0062] Specifically, the input device 11 receives input information from the outside and transmits the input information to the central processing unit 13 through the input interface 12; the central processing unit 13 processes the input information based on the computer executable instructions stored in the memory 14 to generate output information, temporarily or permanently stores the output information in the memory 14, and then transmits the output information to the output device 16 through the output interface 15; the output device 16 outputs the output information to the outside of the electronic device 10 for user use.

[0063] In one embodiment, Figure 6 The electronic device 10 shown can be implemented as a network device, which may include: a memory configured to store programs; a processor configured to run the programs stored in the memory to execute any one of the tunnel and bridge monitoring and management methods described in the above embodiments.

[0064] According to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network, and / or installed from a removable storage medium.

[0065] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0066] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present application, but the present application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of the present application, and these modifications and improvements are also considered to be within the scope of protection of the present application.

Claims

1. A tunnel and bridge monitoring and management system, characterized in that: include: A tunnel and bridge monitoring data acquisition module is used to acquire tunnel and bridge pavement monitoring images acquired by cameras and pavement stress data at multiple time points acquired by fiber grating sensors; A tunnel and bridge monitoring data extraction module, used to extract tunnel and bridge pavement texture enhancement feature vectors and optimized pavement stress space correlation feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; The tunnel and bridge pavement warning judgment module is used to judge whether to issue a tunnel and bridge pavement warning based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector.

2. The tunnel and bridge monitoring and management system according to claim 1 is characterized in that: The tunnel and bridge monitoring data extraction module includes: A tunnel bridge pavement image feature extraction unit, used for extracting features from the tunnel bridge pavement monitoring image collected by the camera to obtain a tunnel bridge pavement texture enhancement feature vector; A road surface stress data preprocessing unit, used for preprocessing the road surface stress data at multiple time points collected by the fiber grating sensor to obtain a road surface stress correlation matrix; The pavement stress characteristic encoding unit is used to perform characteristic encoding on the pavement stress correlation matrix to obtain the optimized pavement stress spatial correlation characteristic vector.

3. The tunnel and bridge monitoring and management system according to claim 2 is characterized in that: The tunnel bridge pavement image feature extraction unit comprises: A road surface monitoring image texture feature encoding subunit is used to perform texture feature encoding on the tunnel and bridge road surface monitoring image collected by the camera to obtain a tunnel and bridge road surface texture enhancement feature map; The road surface monitoring image texture feature pooling subunit is used to pool the tunnel bridge road surface texture enhancement feature map to obtain the tunnel bridge road surface texture enhancement feature vector.

4. The tunnel and bridge monitoring and management system according to claim 3 is characterized in that: The road surface monitoring image texture feature encoding subunit includes: The tunnel bridge pavement monitoring image collected by the camera is passed through a gray level co-occurrence matrix to obtain a tunnel bridge pavement texture enhancement feature map; The tunnel bridge pavement texture enhancement feature map is passed through a tunnel bridge pavement texture enhancement convolutional neural network based on a feature encoder to obtain the tunnel bridge pavement texture enhancement feature map.

5. The tunnel and bridge monitoring and management system according to claim 4 is characterized in that: The pavement stress data preprocessing unit comprises: Arranging the pavement stress data at multiple time points collected by the fiber grating sensor according to the time dimension to obtain a pavement stress input vector; The pavement stress input vector is multiplied by its transpose to obtain the pavement stress correlation matrix.

6. The tunnel and bridge monitoring and management system according to claim 5, characterized in that: The pavement stress characteristic encoding unit comprises: Passing the pavement stress correlation matrix through a pavement stress spatial convolution coding model to obtain a pavement stress spatial correlation feature map; Performing maximum value pooling on the pavement stress spatial correlation feature map to obtain a pavement stress spatial correlation feature vector; Based on the tunnel and bridge pavement texture enhancement feature vector, the pavement stress spatial correlation feature vector is subjected to inverse error perception based on multi-view feature relationship quantification to obtain the optimized pavement stress spatial correlation feature vector.

7. The tunnel and bridge monitoring and management system according to claim 6, characterized in that: Based on the tunnel and bridge pavement texture enhancement feature vector, the pavement stress spatial correlation feature vector is subjected to inverse error perception based on multi-view feature relationship quantification to obtain the optimized pavement stress spatial correlation feature vector, including: Calculating the position difference between the road surface stress spatial correlation feature vector and the tunnel bridge road surface texture enhancement feature vector to obtain a first viewpoint metric representation vector; Calculate the point-by-point multiplication between the pavement stress spatial correlation feature vector and the tunnel bridge pavement texture enhancement feature vector to obtain a second viewpoint metric representation vector; Calculating the positional sum of the pavement stress spatial correlation feature vector and the tunnel bridge pavement texture enhancement feature vector to obtain a third viewpoint metric representation vector; Calculating the gradient magnitude of each position in the first viewpoint metric representation vector, the second viewpoint metric representation vector, and the third viewpoint metric representation vector to obtain a first viewpoint metric perception vector, a second viewpoint metric perception vector, and a third viewpoint metric perception vector; Inputting the first viewpoint metric perception vector, the second viewpoint metric perception vector and the third viewpoint metric perception vector into a backward machine learning inverse multi-view perception optimization module to obtain a reverse multi-view perception optimization representation vector; The optimized road surface stress space correlation feature vector is obtained by calculating the point-by-point multiplication between the inverse multi-view perception optimization representation vector and the road surface stress space correlation feature vector.

8. The tunnel and bridge monitoring and management system according to claim 7, characterized in that: The tunnel bridge road surface early warning judgment module includes: A tunnel and bridge pavement feature fusion unit, used to fuse the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector to obtain a tunnel and bridge pavement early warning judgment feature vector; The road surface warning judgment classification unit is used to pass the tunnel bridge road surface warning judgment feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a tunnel bridge road surface warning.

9. A tunnel and bridge monitoring and management method, characterized in that: include: Acquire monitoring images of tunnel and bridge pavement collected by cameras and pavement stress data at multiple time points collected by fiber grating sensors; Extracting tunnel and bridge pavement texture enhancement feature vectors and optimized pavement stress space correlation feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor; Based on the tunnel and bridge pavement texture enhancement feature vector and the optimized pavement stress space correlation feature vector, it is determined whether to issue a tunnel and bridge pavement warning.

10. The tunnel and bridge monitoring and management method according to claim 9, characterized in that: Extracting tunnel and bridge pavement texture enhancement feature vectors and optimizing pavement stress space correlation feature vectors from the tunnel and bridge pavement monitoring images collected by the camera and the pavement stress data at multiple time points collected by the fiber grating sensor, including: Performing feature extraction on the tunnel bridge pavement monitoring image collected by the camera to obtain the tunnel bridge pavement texture enhancement feature vector; Preprocessing the pavement stress data at multiple time points collected by the fiber grating sensor to obtain a pavement stress correlation matrix; The pavement stress correlation matrix is ​​feature encoded to obtain the optimized pavement stress spatial correlation feature vector.

Images