A tunnel structure whole life cycle damage detection and safety evaluation method

Abstract

This application discloses a method for full life-cycle damage detection and safety assessment of tunnel structures, belonging to the field of damage detection and safety assessment technology. The method includes: installing and configuring a mobile detection platform along the tunnel's axial direction; generating the mobile detection platform's trajectory and detection point distribution scheme; controlling the mobile detection platform to move sequentially to each detection section position according to the detection path, simultaneously activating each detection module for continuous data acquisition during the movement; generating a three-dimensional mathematical model of the tunnel, a lining thickness distribution map, a thermal radiation anomaly distribution map, and a crack damage distribution map; dividing the tunnel's three-dimensional model into N detection segments at preset intervals, and statistically analyzing the average lining thickness, the number of thermal radiation anomaly points, the crack length density, and the damage level of each segment as safety assessment indicators; performing statistical analysis and trend judgment on the safety indicators of each segment, identifying abnormal segments exceeding safety thresholds, and determining their risk levels.

Description

Technical Field

[0001] This application relates to the field of damage detection and safety assessment technology, and more specifically, to a method for damage detection and safety assessment of tunnel structures throughout their entire life cycle. Background Technology

[0002] As a crucial component of transportation infrastructure, the structural safety of tunnels directly impacts operational safety and socio-economic benefits. With increasing tunnel service life, structural materials may age, crack, leak, and the lining may break down. This damage can be accelerated, especially under complex geological conditions and heavy traffic environments. Traditional tunnel safety inspection methods rely heavily on manual inspections and single detection techniques, such as manually measuring crack width, local thickness, or infrared thermal imaging. These methods suffer from drawbacks including low efficiency, discontinuous data, strong subjectivity, and an inability to comprehensively reflect the damage status of the entire tunnel cross-section.

[0003] In recent years, with the development of 3D laser scanning, ground-penetrating radar, infrared thermal imaging, and high-definition image acquisition technologies, multimodal detection methods have been gradually applied to tunnel structural health monitoring. However, existing methods still suffer from problems such as information dispersion, difficulties in data processing and analysis, and a lack of a systematic safety assessment system. Furthermore, existing tunnel inspections are mostly based on single-inspection results, lacking the ability to dynamically track the evolution of structural damage and conduct full life-cycle safety assessments. Especially in cases of long tunnels or complex cross-sections, the deployment and path planning of mobile inspection equipment, the coordination of various inspection modules, and data fusion analysis still face technical challenges, making it difficult for inspection results to accurately reflect the overall structural health status.

[0004] Meanwhile, existing detection methods lack quantitative indicator systems and visualization tools for damage identification and safety assessment. Comprehensive judgments on multiple damage characteristics, such as cracks, voids, lining thickness changes, and thermal anomalies, rely on experience-based analysis, making it difficult to generate scientific, objective, and traceable safety assessment reports. This leaves tunnel operators without reliable data for developing maintenance plans, implementing risk prevention and control, and optimizing operational management, increasing potential safety hazards.

[0005] In summary, how to achieve full-section damage identification, dynamic safety assessment, and full life-cycle risk management of tunnel structures based on multimodal data has become an urgent technical problem to be solved. Summary of the Invention

[0006] To overcome a series of shortcomings in existing technologies, the purpose of this application is to provide a method for full life-cycle damage detection and safety assessment of tunnel structures, applicable to a tunnel safety monitoring system. The tunnel safety monitoring system includes a mobile detection platform, a three-dimensional laser scanning module, a ground-penetrating radar detection module, an infrared thermal imaging module, a high-definition image acquisition module, and a data processing center. The method includes the following steps:

[0007] Obtain the tunnel foundation information and inspection requirements after tunnel completion, determine the initial deployment location of the mobile inspection platform, and complete the installation and configuration of the mobile inspection platform along the tunnel axis.

[0008] Based on the tunnel length, cross-sectional shape, and detection accuracy requirements, a detection path plan is developed, generating the mobile detection platform's movement trajectory and detection point distribution scheme.

[0009] The mobile detection platform is controlled to move sequentially to each detection section position according to the detection path, and each detection module is started simultaneously to collect data continuously during the movement.

[0010] Whenever the mobile detection platform reaches a detection section, it marks the current detection section as the target detection area and configures the working parameters of each detection module based on the geological conditions and structural characteristics of the area.

[0011] The three-dimensional laser scanning module is used to acquire three-dimensional point cloud data of the tunnel, the geological radar detection module is used to collect medium characteristic data of the lining structure, the infrared thermal imaging module is used to acquire thermal image data, and the high-definition image acquisition module is used to acquire visible light image data.

[0012] During the data acquisition process, the data processing center receives and preprocesses multimodal detection data in real time, and performs information fusion analysis by combining point cloud data, medium characteristic data, thermal image data and high-definition image data.

[0013] Based on the fused multimodal data, a three-dimensional mathematical model of the tunnel, a thickness distribution map of the lining structure, a thermal radiation anomaly distribution map, and a crack damage distribution map are generated.

[0014] The tunnel 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density and damage level of each segment are statistically analyzed as safety assessment indicators.

[0015] Perform statistical analysis and trend judgment on the safety indicators of each segment, identify abnormal segments that exceed the safety threshold and determine their risk level;

[0016] If all safety indicators for the tested segments are within the normal range, a qualified safety assessment report will be generated.

[0017] If there are sections with abnormal safety indicators, a detailed risk assessment plan will be developed for the abnormal areas, generating a comprehensive safety assessment report that includes basic tunnel information, detection data, safety indicator analysis, abnormal location identification, risk level classification, and handling suggestions.

[0018] Establish a regular inspection mechanism, repeat the above inspection process according to a set cycle, and monitor the development and evolution trend of tunnel structural damage by comparing the data from previous inspections, so as to achieve dynamic tracking and assessment of the safety status of the tunnel throughout its entire life cycle.

[0019] Furthermore, the mobile inspection platform includes a chassis walking mechanism, a lifting and adjusting mechanism, a rotary positioning mechanism, and a stabilizing support mechanism. The chassis walking mechanism adopts a tracked drive system, possessing omnidirectional movement capabilities (forward, backward, left turn, right turn), with a continuously adjustable speed within the range of 0.1–1.0 m / s to adapt to inspection needs in different tunnel environments. The lifting and adjusting mechanism achieves vertical position adjustment of the inspection equipment via hydraulic drive, with an adjustment height range of 0.5–5.0 meters, ensuring that the inspection equipment can accurately align with various inspection target areas on the top, sidewalls, and bottom of the tunnel. The rotary positioning mechanism has a 360-degree continuous rotation function with a rotation angle accuracy of no less than 0.5 degrees, supporting precise orientation of the inspection equipment in the horizontal plane and meeting the omnidirectional inspection requirements of complex tunnel cross-section shapes. The stabilizing support mechanism eliminates the influence of uneven ground on inspection accuracy through an automatic horizontal adjustment function.

[0020] Furthermore, the three-dimensional laser scanning module is equipped with a high-precision laser emitter and receiver, with a laser wavelength of 905 nanometers, a ranging accuracy of ±5 millimeters, and a scanning range covering a horizontal 360-degree and a vertical 270-degree field of view. The scanning resolution reaches 5 millimeters × 5 millimeters at a distance of 10 meters from the target, enabling it to capture the geometric changes and deformation features of the tunnel surface. During the scanning process, the three-dimensional coordinates, reflection intensity, and timestamp information of each measurement point are recorded simultaneously to generate high-density point cloud data containing color textures.

[0021] The ground-penetrating radar detection module transmits high-frequency electromagnetic pulse waves and receives reflected signals from different media interfaces. It analyzes the amplitude, phase, and propagation time characteristics of the reflected waves to identify cavities, cracks, steel reinforcement distribution, and thickness variations inside the lining. The operating frequency range covers four frequency bands: 200MHz, 400MHz, 800MHz, and 1000MHz.

[0022] The infrared thermal imaging module identifies potential structural defects, such as areas of water leakage, hollow areas, and locations of material property changes, by detecting abnormal temperature distribution on the tunnel surface. It uses an uncooled thermal imaging sensor with a spectral response range of 8-14 micrometers, a thermal resolution of 0.1 degrees Celsius, an image resolution of 640×480 pixels, and a frame rate of 25 frames per second.

[0023] Furthermore, the process of acquiring basic tunnel information specifically includes:

[0024] Key testing parameters and quality control requirements are determined by analyzing the tunnel's design standards, structural form, material specifications, and construction quality.

[0025] Based on the geological environment, hydrological conditions, climate characteristics, and traffic load of the tunnel, assess potential risk factors and formulate targeted detection strategies;

[0026] Taking into account the tunnel's service life, operational status, maintenance history, and management requirements, the key areas, critical indicators, and accuracy requirements for testing are clearly defined.

[0027] The tunnel is divided into high-risk, medium-risk, and low-risk areas, and differentiated detection plans and frequencies are developed for different areas.

[0028] Furthermore, the real-time reception and preprocessing of the multimodal detection data includes:

[0029] A timestamp mechanism is used to ensure that data from different detection modules correspond in the time dimension;

[0030] The raw data from each detection module is converted into standardized data formats, including the general format for point cloud data, the standard format for radar data, the TIFF format for thermal images, and the JPEG format for visible light images.

[0031] Abnormal data is identified and marked by setting data integrity, accuracy range and signal-to-noise ratio thresholds. Data with insufficient signal strength, out-of-tolerance accuracy or noise is labeled with quality level.

[0032] The raw data is filtered according to the detection objectives and analysis requirements to remove redundant and invalid information.

[0033] Furthermore, the information fusion analysis specifically includes:

[0034] By using coordinate system one and geometric registration, the original data of different modalities are mapped to a unified three-dimensional spatial coordinate system, achieving pixel-level precise correspondence;

[0035] Extract key feature information from each modality of data and form a multidimensional feature descriptor by combining feature vectors;

[0036] By combining the reliability and complementarity of the detection results from each modality, a final fusion judgment result is generated.

[0037] Furthermore, the construction of the three-dimensional mathematical model of the tunnel specifically includes:

[0038] Based on laser scanning point cloud data, the precise three-dimensional geometry of the tunnel is generated by surface reconstruction.

[0039] Based on the results of ground-penetrating radar detection, a three-dimensional distribution model of the lining material is established, including spatial distribution information of material type, thickness distribution, density variation and strength characteristics;

[0040] By integrating multimodal detection data, a three-dimensional distribution model of damage defects is constructed, including information on the spatial location, geometric shape, development direction, and degree of damage of cracks;

[0041] The tunnel structure is discretized into hexahedral elements, and the element size is adaptively adjusted according to the detection accuracy requirements. The accuracy and completeness of the model are verified by comparing the design drawings and measured data to ensure that the three-dimensional model can truly reflect the current state of the tunnel structure.

[0042] Furthermore, the process of generating the crack damage distribution map specifically includes:

[0043] Based on laser scanning point cloud data, surface roughness analysis and geometric feature extraction methods are used to identify the three-dimensional geometric features of surface cracks;

[0044] Based on high-definition visible light images, the planar features of cracks are automatically identified;

[0045] Based on infrared thermal imaging data, potential internal cracks and developing microcracks are identified through temperature gradient analysis.

[0046] Construct a crack distribution map that includes information on the spatial location, geometric parameters, development direction, and damage level of the cracks;

[0047] Based on their causes, morphology, and development characteristics, cracks are classified into four types: shrinkage cracks, load cracks, temperature cracks, and settlement cracks, and are represented by different colors and line types.

[0048] Furthermore, the tunnel's 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density, and damage level of each segment are statistically analyzed as safety assessment indicators, including the following steps:

[0049] Based on the total length and structural features of the tunnel 3D model, the model is divided into N detection segments in the axial direction at preset intervals to ensure that each segment has a clear spatial boundary and number.

[0050] Within each segment, the lining thickness value calculated by the fusion of laser point cloud and ground-penetrating radar is extracted. After removing data that exceeds the accuracy threshold, the mean, maximum, minimum and standard deviation of the thickness of that segment are statistically analyzed to reflect the lining construction quality and uniformity.

[0051] By combining thermal imaging data, pixels in each detection segment whose temperature exceeds the set normal range are identified and counted, and the number of abnormal thermal radiation points is counted as potential indicators of leakage, hollowing or material degradation.

[0052] Within each segment, crack lines extracted from the image are identified, and geometric verification is performed by combining the point cloud roughness change results. The total crack length is calculated and normalized to the unit area to form a crack length density index, which is used to reflect the degree of local damage concentration.

[0053] Based on a comprehensive scoring standard that includes thickness deviation, thermal anomaly density, and crack length density, the damage level of each segment is determined and labeled in the three-dimensional model using a level label.

[0054] Furthermore, statistical analysis and trend judgment are performed on the safety indicators of each segment to identify abnormal segments that exceed the safety threshold and determine their risk level, including the following steps:

[0055] For each detection segment, the deviation between each indicator and its corresponding safety threshold is calculated, and a scoring mechanism is used to quantify the overall safety status of each segment, generating a preliminary risk scoring table.

[0056] Filter out detection segments that exceed the safety threshold on any key indicator, mark the segment as an abnormal segment, label the abnormality type and corresponding indicator item, and form a visual abnormality distribution map;

[0057] By comparing current test results with historical data, we can analyze the trend of indicator changes in abnormal segments and identify whether the indicators are continuously deteriorating, stabilizing, or improving.

[0058] Based on the number of abnormal indicators, the extent of exceeding limits, and the trend of change, each abnormal segment is assigned a corresponding risk level;

[0059] The risk level results are mapped to the tunnel's 3D model, and the model's color gradation is updated. At the same time, a risk distribution report is generated, listing the location, abnormality of indicators, and risk level of each abnormal segment.

[0060] Compared with the prior art, the beneficial effects of this application are as follows.

[0061] This application integrates multimodal detection technology and a mobile detection platform, and through high-precision 3D modeling and segmented index evaluation, it realizes dynamic damage identification and intelligent safety level assessment of tunnel structures throughout their entire life cycle. Attached Figure Description

[0062] Figure 1 This is a flowchart of a method for full life-cycle damage detection and safety assessment of tunnel structures disclosed in an embodiment of this application. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some embodiments of this invention, but not all embodiments.

[0064] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0066] like Figure 1 As shown, a method for full life-cycle damage detection and safety assessment of tunnel structures is applied to a tunnel safety monitoring system. The tunnel safety monitoring system includes a mobile detection platform, a three-dimensional laser scanning module, a ground-penetrating radar detection module, an infrared thermal imaging module, a high-definition image acquisition module, and a data processing center. The method includes the following steps:

[0067] Obtain the tunnel foundation information and inspection requirements after tunnel completion, determine the initial deployment location of the mobile inspection platform, and complete the installation and configuration of the mobile inspection platform along the tunnel axis.

[0068] Based on the tunnel length, cross-sectional shape, and detection accuracy requirements, a detection path plan is developed, generating the mobile detection platform's movement trajectory and detection point distribution scheme.

[0069] The mobile detection platform is controlled to move sequentially to each detection section position according to the detection path, and each detection module is started simultaneously to collect data continuously during the movement.

[0070] Whenever the mobile detection platform reaches a detection section, it marks the current detection section as the target detection area and configures the working parameters of each detection module based on the geological conditions and structural characteristics of the area.

[0071] The three-dimensional laser scanning module is used to acquire three-dimensional point cloud data of the tunnel, the geological radar detection module is used to collect medium characteristic data of the lining structure, the infrared thermal imaging module is used to acquire thermal image data, and the high-definition image acquisition module is used to acquire visible light image data.

[0072] During the data acquisition process, the data processing center receives and preprocesses multimodal detection data in real time, and performs information fusion analysis by combining point cloud data, medium characteristic data, thermal image data and high-definition image data.

[0073] Based on the fused multimodal data, a three-dimensional mathematical model of the tunnel, a thickness distribution map of the lining structure, a thermal radiation anomaly distribution map, and a crack damage distribution map are generated.

[0074] The tunnel 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density and damage level of each segment are statistically analyzed as safety assessment indicators.

[0075] Perform statistical analysis and trend judgment on the safety indicators of each segment, identify abnormal segments that exceed the safety threshold and determine their risk level;

[0076] If all safety indicators for the tested segments are within the normal range, a qualified safety assessment report will be generated.

[0077] If there are sections with abnormal safety indicators, a detailed risk assessment plan will be developed for the abnormal areas, generating a comprehensive safety assessment report that includes basic tunnel information, detection data, safety indicator analysis, abnormal location identification, risk level classification, and handling suggestions.

[0078] Establish a regular inspection mechanism, repeat the above inspection process according to a set cycle, and monitor the development and evolution trend of tunnel structural damage by comparing the data from previous inspections, so as to achieve dynamic tracking and assessment of the safety status of the tunnel throughout its entire life cycle.

[0079] This method for full life-cycle damage detection and safety assessment of tunnel structures enables comprehensive and refined monitoring and scientific evaluation of tunnel structural conditions. First, a mobile detection platform equipped with a multimodal detection module continuously acquires high-precision 3D point cloud, medium properties, thermal imaging, and visible light image data along the tunnel axis, ensuring spatial and temporal consistency and richness of the data. Second, by fusing and analyzing the acquired multi-source heterogeneous data, the information advantages of different detection methods are effectively integrated, improving the accuracy and reliability of damage identification. The 3D tunnel model and multidimensional damage distribution map constructed based on the fused data achieve spatial localization and quantitative description of tunnel structural damage, providing a scientific basis for safety assessment. Third, through segmented statistics and trend analysis, the method can keenly capture damage evolution patterns, promptly identify potential risks, and support hierarchical management and precise maintenance. Furthermore, the periodic repeated detection and dynamic tracking mechanism helps to comprehensively grasp changes in tunnel safety status, promotes dynamic adjustments in early warning and decision-making, and extends the tunnel's service life.

[0080] Furthermore, the mobile inspection platform includes a chassis walking mechanism, a lifting and adjusting mechanism, a rotary positioning mechanism, and a stabilizing support mechanism. The chassis walking mechanism adopts a tracked drive system, possessing omnidirectional movement capabilities (forward, backward, left turn, right turn), with a continuously adjustable speed within the range of 0.1–1.0 m / s to adapt to inspection needs in different tunnel environments. The lifting and adjusting mechanism achieves vertical position adjustment of the inspection equipment via hydraulic drive, with an adjustment height range of 0.5–5.0 meters, ensuring that the inspection equipment can accurately align with various inspection target areas on the top, sidewalls, and bottom of the tunnel. The rotary positioning mechanism has a 360-degree continuous rotation function with a rotation angle accuracy of no less than 0.5 degrees, supporting precise orientation of the inspection equipment in the horizontal plane and meeting the omnidirectional inspection requirements of complex tunnel cross-section shapes. The stabilizing support mechanism eliminates the influence of uneven ground on inspection accuracy through an automatic horizontal adjustment function.

[0081] In summary, this mobile inspection platform achieves high flexibility and high precision in tunnel environments through multi-mechanism collaboration. The tracked chassis ensures controllable movement in all directions and continuous speed adjustment, adapting to different tunnel conditions; the lifting and adjusting mechanism precisely adjusts the height of the inspection equipment, covering the tunnel top, sidewalls, and bottom; the rotary positioning mechanism supports high-precision horizontal rotation, enabling all-around cross-sectional inspection; and the stabilizing support mechanism automatically levels, improving data acquisition accuracy under complex ground conditions. Overall, this enhances the automation, reliability, and operational efficiency of tunnel inspection.

[0082] Furthermore, the three-dimensional laser scanning module is equipped with a high-precision laser emitter and receiver, with a laser wavelength of 905 nanometers, a ranging accuracy of ±5 millimeters, and a scanning range covering a horizontal 360-degree and a vertical 270-degree field of view. The scanning resolution reaches 5 millimeters × 5 millimeters at a distance of 10 meters from the target, enabling it to capture the geometric changes and deformation features of the tunnel surface. During the scanning process, the three-dimensional coordinates, reflection intensity, and timestamp information of each measurement point are recorded simultaneously to generate high-density point cloud data containing color textures.

[0083] The ground-penetrating radar detection module transmits high-frequency electromagnetic pulse waves and receives reflected signals from different medium interfaces. It analyzes the amplitude, phase, and propagation time characteristics of the reflected waves to identify cavities, cracks, steel reinforcement distribution, and thickness variations inside the lining. The operating frequency range covers four frequency bands: 200MHz, 400MHz, 800MHz, and 1000MHz.

[0084] The infrared thermal imaging module identifies potential structural defects, such as areas of water leakage, hollow areas, and locations of material property changes, by detecting abnormal temperature distribution on the tunnel surface. It uses an uncooled thermal imaging sensor with a spectral response range of 8-14 micrometers, a thermal resolution of 0.1 degrees Celsius, an image resolution of 640×480 pixels, and a frame rate of 25 frames per second.

[0085] In summary, this multimodal detection system enables refined monitoring of tunnel structures. The 3D laser scanning module provides high-precision, high-density point cloud data, accurately reflecting the geometric features and minute deformations of the tunnel surface; the ground-penetrating radar module can penetrate deep into the lining to identify voids, cracks, and reinforcement placement, enhancing structural health diagnostic capabilities; and the infrared thermal imaging module detects potential defects through temperature anomalies, enabling non-contact and rapid identification of leaks, voids, and material deterioration locations. Overall, this system improves the accuracy, comprehensiveness, and real-time performance of tunnel damage detection, providing reliable data support for subsequent safety assessments.

[0086] Furthermore, the data processing center includes a data acquisition server, an image processing server, a computing server, and a data storage server. The data acquisition server is equipped with multiple data interfaces, enabling it to simultaneously receive raw data from four detection modules. The image processing server is equipped with a GPU computing unit, possessing parallel image processing capabilities, and can handle denoising, enhancement, and feature extraction operations on multiple image data streams. The computing server has analytical capabilities for defect identification, damage assessment, and risk level judgment. The data storage server employs disk array technology, with a storage capacity of 50TB, supporting long-term storage and retrieval of detection data.

[0087] In summary, this data processing center achieves efficient management and intelligent analysis of tunnel inspection data through the collaborative work of multiple servers. The data acquisition server can simultaneously receive multimodal raw data, ensuring information integrity; the image processing server utilizes GPU acceleration to complete image denoising, enhancement, and feature extraction, improving processing speed and accuracy; the computing server can quickly identify defects and assess the degree of damage and risk level; and the data storage server provides a large-capacity, secure, and reliable storage environment, supporting long-term storage and easy retrieval of historical data. Overall, this significantly improves the automation level, data processing efficiency, and analytical accuracy of tunnel monitoring.

[0088] Furthermore, the process of acquiring basic tunnel information specifically includes:

[0089] Key testing parameters and quality control requirements are determined by analyzing the tunnel's design standards, structural form, material specifications, and construction quality.

[0090] Based on the geological environment, hydrological conditions, climate characteristics, and traffic load of the tunnel, assess potential risk factors and formulate targeted detection strategies;

[0091] Taking into account the tunnel's service life, operational status, maintenance history, and management requirements, the key areas, critical indicators, and accuracy requirements for testing are clearly defined.

[0092] The tunnel is divided into high-risk, medium-risk, and low-risk areas, and differentiated detection plans and frequencies are developed for different areas.

[0093] In summary, the process of acquiring basic tunnel information provides a scientific basis and targeted strategies for subsequent inspections. By analyzing design standards, structural forms, and construction materials, key inspection parameters and quality requirements are clarified; by combining environmental factors such as geology, hydrology, climate, and traffic loads, potential risks are assessed and inspection strategies are formulated; by considering the service life, operation, and maintenance history, key inspection areas and indicator accuracy are determined; and by risk zoning, differentiated inspections and reasonable inspection frequency arrangements for different areas are achieved, thereby improving the scientific rigor, accuracy, and management efficiency of tunnel safety assessments.

[0094] Furthermore, the real-time reception and preprocessing of the multimodal detection data includes:

[0095] A timestamp mechanism is used to ensure that data from different detection modules correspond in the time dimension;

[0096] The raw data from each detection module is converted into standardized data formats, including the general format for point cloud data, the standard format for radar data, the TIFF format for thermal images, and the JPEG format for visible light images.

[0097] Abnormal data is identified and marked by setting data integrity, accuracy range and signal-to-noise ratio thresholds. Data with insufficient signal strength, out-of-tolerance accuracy or noise is labeled with quality level.

[0098] The raw data is filtered according to the detection objectives and analysis requirements to remove redundant and invalid information.

[0099] In summary, this real-time reception and preprocessing process for multimodal detection data ensures data consistency, reliability, and availability. The timestamp mechanism enables synchronized alignment of data from different modules, ensuring spatiotemporal accuracy of the analysis; standardized data formats facilitate subsequent processing and multi-source fusion; anomaly identification and quality labeling enhance data reliability; and data filtering removes redundant and invalid information, improving processing efficiency and analytical accuracy, thus providing a high-quality data foundation for tunnel damage identification and safety assessment.

[0100] Furthermore, the information fusion analysis specifically includes:

[0101] By using coordinate system one and geometric registration, the original data of different modalities are mapped to a unified three-dimensional spatial coordinate system, achieving pixel-level precise correspondence;

[0102] Extract key feature information from each modality of data and form a multidimensional feature descriptor by combining feature vectors;

[0103] By combining the reliability and complementarity of the detection results from each modality, a final fusion judgment result is generated.

[0104] In summary, this information fusion analysis process enables high-precision integration and comprehensive judgment of multimodal data. By unifying coordinates and geometric registration, data from different sources are mapped to the same three-dimensional space, achieving pixel-level correspondence and ensuring spatial information consistency. Key feature extraction and multi-dimensional feature descriptor construction effectively integrate the details and features of each modality. Combining the reliability and complementarity of the results from each modality to generate a fusion judgment improves the accuracy and comprehensiveness of damage identification, providing scientific and reliable data support for tunnel safety assessment.

[0105] Furthermore, the construction of the three-dimensional mathematical model of the tunnel specifically includes:

[0106] Based on laser scanning point cloud data, the precise three-dimensional geometry of the tunnel is generated by surface reconstruction.

[0107] Based on the results of ground-penetrating radar detection, a three-dimensional distribution model of the lining material is established, including spatial distribution information of material type, thickness distribution, density variation and strength characteristics;

[0108] By integrating multimodal detection data, a three-dimensional distribution model of damage defects is constructed, including information on the spatial location, geometric shape, development direction, and degree of damage of cracks;

[0109] The tunnel structure is discretized into hexahedral elements, and the element size is adaptively adjusted according to the detection accuracy requirements. The accuracy and completeness of the model are verified by comparing the design drawings and measured data to ensure that the three-dimensional model can truly reflect the current state of the tunnel structure.

[0110] In summary, this method for constructing a three-dimensional mathematical model of a tunnel can accurately reflect the geometric shape, material distribution, and damage state of the tunnel structure. A high-precision curved surface model is generated based on laser point clouds, and a spatial distribution model of the lining material is established by combining ground-penetrating radar data, enabling visualization of thickness, density, and strength characteristics. A three-dimensional distribution model of cracks and other damage defects is constructed by integrating multimodal detection results, clarifying their spatial location, shape, and development trend. The model's accuracy and completeness are ensured through hexahedral element discretization and comparison with design drawings, providing a reliable foundation for refined safety assessment and risk analysis.

[0111] Furthermore, the process of generating the lining structure thickness distribution map specifically includes:

[0112] The actual thickness distribution of the tunnel lining is calculated by combining the results of laser scanning of the outer surface contour and the internal detection results of ground-penetrating radar.

[0113] By establishing the three-dimensional coordinate correspondence between the outer and inner surfaces of the lining, the thickness value of each measuring point is calculated along the normal direction;

[0114] The thickness values ​​are mapped to different colors to visually display the spatial variation of the thickness.

[0115] The acceptable thickness range is set to ±15% of the design thickness; areas exceeding the acceptable range are highlighted with special markings.

[0116] Calculate the mean, standard deviation, maximum and minimum thickness of each section of the tunnel, and generate a thickness distribution histogram and statistical report.

[0117] In summary, the process of generating this lining thickness distribution map can intuitively reflect the actual thickness and spatial variation characteristics of the tunnel lining. By combining laser scanning of the outer contour with data from ground-penetrating radar, accurate calculation of the lining thickness is achieved; a three-dimensional correspondence between the outer and inner surfaces is established to ensure the spatial accuracy of thickness measurement; thickness values ​​are visualized as color distributions to highlight abnormal areas exceeding the design range; and the mean, standard deviation, and extreme values ​​of thickness in each section are statistically analyzed, generating histograms and reports to provide quantitative basis and intuitive reference for structural health assessment and maintenance decisions.

[0118] Furthermore, the process of generating the crack damage distribution map specifically includes:

[0119] Based on laser scanning point cloud data, surface roughness analysis and geometric feature extraction methods are used to identify the three-dimensional geometric features of surface cracks;

[0120] Based on high-definition visible light images, the planar features of cracks are automatically identified;

[0121] Based on infrared thermal imaging data, potential internal cracks and developing microcracks are identified through temperature gradient analysis.

[0122] Construct a crack distribution map that includes information on the spatial location, geometric parameters, development direction, and damage level of the cracks;

[0123] Based on their causes, morphology, and development characteristics, cracks are classified into four types: shrinkage cracks, load cracks, temperature cracks, and settlement cracks, and are represented by different colors and line types.

[0124] In summary, this crack damage distribution map generation process can comprehensively and accurately reflect the spatial characteristics and development status of tunnel cracks. It extracts three-dimensional geometric features from surface data using laser point cloud data, identifies planar cracks using high-definition visible light images, and simultaneously utilizes infrared thermal imaging to discover potential internal cracks, achieving multi-dimensional crack detection. It constructs an infographic containing spatial location, geometric parameters, development direction, and damage level; and classifies and visualizes cracks according to their causes and morphologies, intuitively displaying crack types and distribution patterns, providing a reliable basis for tunnel structural health assessment and risk prevention.

[0125] Furthermore, the safety assessment report comprises six parts: an executive summary, an inspection overview, data analysis, results evaluation, risk assessment, and recommended measures. The executive summary concisely summarizes the overall safety status of the tunnel, key findings, and critical recommendations, providing a rapid reference for management decisions. The inspection overview details the time, location, scope, methods, equipment, and personnel information of the inspection, ensuring the traceability of the inspection process. The data analysis section includes raw data statistics, processing instructions, and analysis results presentation, comprehensively showcasing the inspection findings through charts, images, and numerical tables. The results evaluation assesses the safety status of various parts of the tunnel according to preset evaluation standards, clearly identifying qualified and unqualified areas. The risk assessment identifies potential risk sources, analyzes the likelihood of risk occurrence and the severity of consequences, and formulates risk control measures. The recommended measures propose specific maintenance plans, monitoring suggestions, and management recommendations for the identified problems, specifying implementation priorities and time requirements.

[0126] In summary, this safety assessment report systematically and comprehensively reflects the safety status of the tunnel structure. The executive summary provides an overall overview and key recommendations for rapid decision-making; the inspection overview records information throughout the process, ensuring data traceability; the data analysis section presents inspection results with charts and numerical values, improving information comprehensibility; the results evaluation clarifies the safety level of each component, providing a basis for maintenance and management; the risk assessment identifies potential hazards and analyzes possible consequences, supporting scientific prevention and control; and the recommended measures propose targeted maintenance, monitoring, and management plans, specifying priorities and implementation deadlines, providing effective assurance for tunnel operational safety.

[0127] Furthermore, the tunnel's 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density, and damage level of each segment are statistically analyzed as safety assessment indicators, including the following steps:

[0128] Based on the total length and structural features of the tunnel 3D model, the model is divided into N detection segments in the axial direction at preset intervals to ensure that each segment has a clear spatial boundary and number.

[0129] Within each segment, the lining thickness value calculated by the fusion of laser point cloud and ground-penetrating radar is extracted. After removing data that exceeds the accuracy threshold, the mean, maximum, minimum and standard deviation of the thickness of that segment are statistically analyzed to reflect the lining construction quality and uniformity.

[0130] By combining thermal imaging data, pixels in each detection segment whose temperature exceeds the set normal range are identified and counted, and the number of abnormal thermal radiation points is counted as potential indicators of leakage, hollowing or material degradation.

[0131] Within each segment, crack lines extracted from the image are identified, and geometric verification is performed by combining the point cloud roughness change results. The total crack length is calculated and normalized to the unit area to form a crack length density index, which is used to reflect the degree of local damage concentration.

[0132] Based on a comprehensive scoring standard that includes thickness deviation, thermal anomaly density, and crack length density, the damage level of each segment is determined and labeled in the three-dimensional model using a level label.

[0133] In summary, this segmented safety assessment method enables refined and quantitative analysis of tunnel structural conditions. By dividing the 3D model into N equally axially oriented detection segments, each segment has a clearly defined spatial boundary and is numbered, achieving localized management of structural information. The average and extreme values ​​of the lining thickness reflect construction quality and uniformity. Thermal imaging data is used to calculate the number of anomalies, identifying potential leakage or material degradation areas. Crack length density indicators are used to quantify the concentration of local damage. Finally, the segment damage level is determined based on a comprehensive multi-indicator score, and visualized in the 3D model, providing a scientific basis for accurate risk identification, local maintenance decisions, and dynamic safety management.

[0134] Furthermore, statistical analysis and trend judgment are performed on the safety indicators of each segment to identify abnormal segments that exceed the safety threshold and determine their risk level, including the following steps:

[0135] For each detection segment, the deviation between each indicator and its corresponding safety threshold is calculated, and a scoring mechanism is used to quantify the overall safety status of each segment, generating a preliminary risk scoring table.

[0136] Filter out detection segments that exceed the safety threshold on any key indicator, mark the segment as an abnormal segment, label the abnormality type and corresponding indicator item, and form a visual abnormality distribution map;

[0137] By comparing current test results with historical data, we can analyze the trend of indicator changes in abnormal segments and identify whether the indicators are continuously deteriorating, stabilizing, or improving.

[0138] Based on the number of abnormal indicators, the extent of exceeding limits, and the trend of change, each abnormal segment is assigned a corresponding risk level;

[0139] The risk level results are mapped to the tunnel's 3D model, and the model's color gradation is updated. At the same time, a risk distribution report is generated, listing the location, abnormality of indicators, and risk level of each abnormal segment.

[0140] In summary, this segmented safety indicator statistical and trend analysis method enables dynamic identification and quantitative management of tunnel structural risks. By calculating the deviation of indicators for each segment and generating risk scores, the overall safety status is quantified; segments exceeding limits are screened and anomaly types are marked, visually presenting potential risk areas; by analyzing the trend of indicator changes in historical data, the damage development situation can be identified, providing a basis for early warning; based on the extent and trend of exceeding limits, segment risk levels are assigned and visualized in a 3D model, while generating a detailed risk distribution report, providing a scientific reference for tunnel maintenance decisions, prioritization, and long-term monitoring.

Claims

1. A method for full life-cycle damage detection and safety assessment of tunnel structures, applied to tunnel safety monitoring systems, characterized in that, The tunnel safety monitoring system includes a mobile detection platform, a three-dimensional laser scanning module, a ground-penetrating radar detection module, an infrared thermal imaging module, a high-definition image acquisition module, and a data processing center. The method includes the following steps: Obtain the tunnel foundation information and inspection requirements after tunnel completion, determine the initial deployment location of the mobile inspection platform, and complete the installation and configuration of the mobile inspection platform along the tunnel axis. Based on the tunnel length, cross-sectional shape, and detection accuracy requirements, a detection path plan is developed, generating the mobile detection platform's movement trajectory and detection point distribution scheme. The mobile detection platform is controlled to move sequentially to each detection section position according to the detection path, and each detection module is started simultaneously to collect data continuously during the movement. Whenever the mobile detection platform reaches a detection section, it marks the current detection section as the target detection area and configures the working parameters of each detection module based on the geological conditions and structural characteristics of the area. The three-dimensional laser scanning module is used to acquire three-dimensional point cloud data of the tunnel, the geological radar detection module is used to collect medium characteristic data of the lining structure, the infrared thermal imaging module is used to acquire thermal image data, and the high-definition image acquisition module is used to acquire visible light image data. During the data acquisition process, the data processing center receives and preprocesses multimodal detection data in real time, and performs information fusion analysis by combining point cloud data, medium characteristic data, thermal image data and high-definition image data. Based on the fused multimodal data, a three-dimensional mathematical model of the tunnel, a thickness distribution map of the lining structure, a thermal radiation anomaly distribution map, and a crack damage distribution map are generated. The tunnel 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density and damage level of each segment are statistically analyzed as safety assessment indicators. Perform statistical analysis and trend judgment on the safety indicators of each segment, identify abnormal segments that exceed the safety threshold and determine their risk level; If all safety indicators for the tested segments are within the normal range, a qualified safety assessment report will be generated. If there are sections with abnormal safety indicators, a detailed risk assessment plan will be developed for the abnormal areas, generating a comprehensive safety assessment report that includes basic tunnel information, detection data, safety indicator analysis, abnormal location identification, risk level classification, and handling suggestions. Establish a regular inspection mechanism, repeat the above steps according to a set cycle, and monitor the development and evolution trend of tunnel structural damage by comparing the data from previous inspections, so as to achieve dynamic tracking and assessment of the safety status of the tunnel throughout its entire life cycle. The mobile inspection platform includes a chassis walking mechanism, a lifting and adjusting mechanism, a rotary positioning mechanism, and a stabilizing support mechanism. The chassis walking mechanism adopts a tracked drive system, possessing omnidirectional movement capabilities (forward, backward, left turn, right turn), and its movement speed is continuously adjustable within the range of 0.1 to 1.0 meters per second to adapt to the inspection needs of different tunnel environments. The lifting and adjusting mechanism achieves vertical position adjustment of the inspection equipment through hydraulic drive, with an adjustment height range of 0.5 to 5.0 meters, ensuring that the inspection equipment can accurately align with various inspection target areas on the top, sidewalls, and bottom of the tunnel. The rotary positioning mechanism has a 360-degree continuous rotation function with a rotation angle accuracy of not less than 0.5 degrees, supporting precise orientation of the inspection equipment in the horizontal plane and meeting the all-round inspection requirements of complex tunnel cross-section shapes. The stabilizing support mechanism eliminates the impact of uneven ground on detection accuracy through an automatic leveling adjustment function. The construction of the three-dimensional mathematical model of the tunnel specifically includes: Based on laser scanning point cloud data, the precise three-dimensional geometry of the tunnel is generated by surface reconstruction. Based on the results of ground-penetrating radar detection, a three-dimensional distribution model of the lining material is established, including spatial distribution information of material type, thickness distribution, density variation and strength characteristics; By integrating multimodal detection data, a three-dimensional distribution model of damage defects is constructed, including information on the spatial location, geometric shape, development direction, and degree of damage of cracks; The tunnel structure is discretized into hexahedral elements, and the element size is adaptively adjusted according to the detection accuracy requirements. The accuracy and completeness of the model are verified by comparing the design drawings and measured data to ensure that the three-dimensional model can truly reflect the current state of the tunnel structure. The tunnel's 3D model is divided into N detection segments at preset intervals. The average lining thickness, number of thermal radiation anomalies, crack length density, and damage level of each segment are statistically analyzed as safety assessment indicators, including the following steps: Based on the total length and structural features of the tunnel 3D model, the model is divided into N detection segments in the axial direction at preset intervals to ensure that each segment has a clear spatial boundary and number. Within each segment, the lining thickness value calculated by the fusion of laser point cloud and ground-penetrating radar is extracted. After removing data that exceeds the accuracy threshold, the mean, maximum, minimum and standard deviation of the thickness of that segment are statistically analyzed to reflect the lining construction quality and uniformity. By combining thermal imaging data, pixels in each detection segment whose temperature exceeds the set normal range are identified and counted, and the number of abnormal thermal radiation points is counted as potential indicators of leakage, hollowing or material degradation. Within each segment, crack lines extracted from the image are identified, and geometric verification is performed by combining the point cloud roughness change results. The total crack length is calculated and normalized to the unit area to form a crack length density index, which is used to reflect the degree of local damage concentration. Based on a comprehensive scoring standard that considers thickness deviation, thermal anomaly density, and crack length density, the damage level of each segment is determined and labeled in the 3D model using a level label.

2. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, The three-dimensional laser scanning module is equipped with a high-precision laser transmitter and receiver, with a laser wavelength of 905 nanometers, a ranging accuracy of ±5 millimeters, and a scanning range covering a horizontal 360-degree and a vertical 270-degree field of view. The scanning resolution reaches 5 millimeters × 5 millimeters at a distance of 10 meters from the target, enabling it to capture the geometric changes and deformation features of the tunnel surface. During the scanning process, the three-dimensional coordinates, reflection intensity, and timestamp information of each measurement point are recorded simultaneously to generate high-density point cloud data containing color texture. The ground-penetrating radar detection module transmits high-frequency electromagnetic pulse waves and receives reflected signals from different media interfaces. It analyzes the amplitude, phase, and propagation time characteristics of the reflected waves to identify cavities, cracks, steel reinforcement distribution, and thickness variations inside the lining. The operating frequency range covers four frequency bands: 200MHz, 400MHz, 800MHz, and 1000MHz. The infrared thermal imaging module identifies potential structural defects, such as areas of water leakage, hollow areas, and locations of material property changes, by detecting abnormal temperature distribution on the tunnel surface. It uses an uncooled thermal imaging sensor with a spectral response range of 8-14 micrometers, a thermal resolution of 0.1 degrees Celsius, an image resolution of 640×480 pixels, and a frame rate of 25 frames per second.

3. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, The process of acquiring basic tunnel information specifically includes: Key testing parameters and quality control requirements are determined by analyzing the tunnel's design standards, structural form, material specifications, and construction quality. Based on the geological environment, hydrological conditions, climate characteristics, and traffic load of the tunnel, assess potential risk factors and formulate targeted detection strategies; Taking into account the tunnel's service life, operational status, maintenance history, and management requirements, the key areas, critical indicators, and accuracy requirements for testing are clearly defined. The tunnel is divided into high-risk, medium-risk, and low-risk areas, and differentiated detection plans and frequencies are developed for different areas.

4. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, The real-time reception and preprocessing of the multimodal detection data includes: A timestamp mechanism is used to ensure that data from different detection modules correspond in the time dimension; The raw data from each detection module is converted into standardized data formats, including the general format for point cloud data, the standard format for radar data, the TIFF format for thermal images, and the JPEG format for visible light images. Abnormal data is identified and marked by setting data integrity, accuracy range and signal-to-noise ratio thresholds. Data with insufficient signal strength, out-of-tolerance accuracy or noise is labeled with quality level. The raw data is filtered according to the detection objectives and analysis requirements to remove redundant and invalid information.

5. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, The information fusion analysis specifically includes: By using coordinate system one and geometric registration, the original data of different modalities are mapped to a unified three-dimensional spatial coordinate system, achieving pixel-level precise correspondence; Extract key feature information from each modality of data and form a multidimensional feature descriptor by combining feature vectors; By combining the reliability and complementarity of the detection results from each modality, a final fusion judgment result is generated.

6. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, The process of generating the crack damage distribution map specifically includes: Based on laser scanning point cloud data, surface roughness analysis and geometric feature extraction methods are used to identify the three-dimensional geometric features of surface cracks; Based on high-definition visible light images, the planar features of cracks are automatically identified; Based on infrared thermal imaging data, potential internal cracks and developing microcracks are identified through temperature gradient analysis. Construct a crack distribution map that includes information on the spatial location, geometric parameters, development direction, and damage level of the cracks; Based on their causes, morphology, and development characteristics, cracks are classified into four types: shrinkage cracks, load cracks, temperature cracks, and settlement cracks, and are represented by different colors and line types.

7. The method for full life-cycle damage detection and safety assessment of tunnel structures according to claim 1, characterized in that, Statistical analysis and trend assessment of safety indicators for each segment are performed to identify abnormal segments exceeding safety thresholds and determine their risk levels, including the following steps: For each detection segment, the deviation between each indicator and its corresponding safety threshold is calculated, and a scoring mechanism is used to quantify the overall safety status of each segment, generating a preliminary risk scoring table. Filter out detection segments that exceed the safety threshold on any key indicator, mark the segment as an abnormal segment, label the abnormality type and corresponding indicator item, and form a visual abnormality distribution map; By comparing current test results with historical data, we can analyze the trend of indicator changes in abnormal segments and identify whether the indicators are continuously deteriorating, stabilizing, or improving. Based on the number of abnormal indicators, the extent of exceeding limits, and the trend of change, each abnormal segment is assigned a corresponding risk level; The risk level results are mapped to the tunnel's 3D model, and the model's color gradation is updated. At the same time, a risk distribution report is generated, listing the location, abnormality of indicators, and risk level of each abnormal segment.

Images