A method and system for monitoring tunnel damage

Abstract

This application discloses a tunnel damage monitoring method and system, belonging to the field of data processing technology. The method includes: acquiring tunnel operation data and tunnel damage data, and transmitting them to a control center; creating a scoring dataset; creating a damage dataset; constructing a tunnel damage monitoring model, training the damage monitoring model until a standard is met; continuously collecting real-time tunnel operation data, inputting the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score; if the real-time damage score exceeds a preset damage score threshold, the real-time tunnel damage data is determined through the tunnel damage monitoring model. This solution can promptly detect potential damage or anomalies, ensuring tunnel safety. Automated data processing using a model improves data processing efficiency. By analyzing historical and real-time data, the model captures patterns and trends in real-world situations, ensuring that the results match the actual situation and improving the reliability of the analysis results.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and system for monitoring tunnel damage. Background Technology

[0002] Tunnels, as a crucial component of highway transportation, play a vital role in ensuring smooth traffic flow and protecting people's lives and property. However, due to the complex geological environment in which tunnels are located, they may be affected by various factors during construction and operation, leading to structural damage. Therefore, tunnel damage monitoring is an essential means to ensure the safe operation of tunnels.

[0003] Currently, tunnel damage monitoring is conducted manually. Inspectors, relying on their professional knowledge and practical experience, use tools such as crack microscopes and measuring rulers to conduct a comprehensive inspection of tunnel lining cracks, quality, and water leakage, recording detailed data such as crack location, length, width, and lining quality. By compiling and analyzing the inspection results, the safety and stability of the tunnel structure are assessed, and repair and reinforcement plans are formulated accordingly.

[0004] However, manual inspection requires a significant amount of time and manpower, especially in long tunnels or tunnel complexes, where the inspection speed is slow and affects normal road traffic. Due to the complex environment and poor lighting conditions inside tunnels, manual inspection is easily limited by visual constraints, resulting in low inspection accuracy and the possibility of missed detections or misjudgments. Summary of the Invention

[0005] This application provides a tunnel damage monitoring method and system, which solves the technical problems in the prior art where manual detection of tunnel damage requires a lot of time and manpower, especially in the case of long tunnels or tunnel groups, the detection speed is slow and affects the normal passage of roads; and due to the complex internal environment of the tunnel and poor lighting conditions, manual detection is easily limited by vision, resulting in low detection accuracy and possible missed detections or misjudgments.

[0006] In a first aspect, embodiments of this application provide a tunnel damage monitoring method, the method comprising:

[0007] In the initial stage of tunnel operation, tunnel operation data and tunnel damage data are acquired at each preset collection time point, and the tunnel operation data and tunnel damage data are transmitted to the control center. Among them, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the staff's handheld terminal.

[0008] The system receives damage scores for tunnel operation data sent by the control center, labels the tunnel operation data with damage score tags based on the damage scores, and creates a score dataset based on the tunnel operation data and the damage score tags.

[0009] Obtain a preset damage score threshold, and divide the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data based on the damage score and the damage score threshold;

[0010] Based on the tunnel damage data, determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags.

[0011] Construct a tunnel damage monitoring model, and train the damage monitoring model based on the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard;

[0012] Real-time tunnel operation data is continuously collected and input into the tunnel damage monitoring model to obtain a real-time damage score. If the real-time damage score exceeds a preset damage score threshold, the real-time tunnel damage data is determined through the tunnel damage monitoring model.

[0013] Furthermore, after continuously collecting real-time tunnel operation data and inputting the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, the method further includes:

[0014] If the real-time damage score is lower than the preset damage score threshold, the tunnel operation data is input into the preset maintenance detection model to obtain the real-time maintenance score.

[0015] If the real-time maintenance score exceeds the preset maintenance score threshold, the maintenance type is determined by the preset maintenance detection model and the maintenance type is transmitted to the control center.

[0016] Furthermore, after determining real-time tunnel damage data through the tunnel damage monitoring model, the method further includes:

[0017] Acquire tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data, and create a three-dimensional tunnel model based on the tunnel geometry data;

[0018] The material properties of the tunnel 3D model are set according to the tunnel material data, and the boundary data of the tunnel 3D model is set according to the tunnel boundary data.

[0019] The tunnel mesh is divided into three-dimensional models of the tunnel based on the tunnel geometry data and tunnel material data, and the tunnel load data is input into the three-dimensional models of the tunnel.

[0020] Furthermore, after meshing the tunnel 3D model based on the tunnel geometry data and tunnel material data, and inputting tunnel load data into the tunnel 3D model, the method further includes:

[0021] Obtain historical monitoring results, and determine historical tunnel damage data, causes of historical damage, and impacts of historical damage based on the historical monitoring results;

[0022] Obtain simulation requirements, set simulation parameters for the tunnel 3D model according to the simulation requirements, input the historical tunnel damage data into the tunnel 3D model for simulation testing, and obtain tunnel simulation results;

[0023] Based on the tunnel simulation results, historical damage causes, and historical damage impacts, determine whether the tunnel 3D model meets the preset simulation standards.

[0024] If the preset simulation standards are met, the tunnel 3D model is deemed to have passed the simulation test.

[0025] Furthermore, after determining whether the tunnel 3D model meets the preset simulation standards based on the tunnel simulation results and historical tunnel damage data, the method further includes:

[0026] If the preset simulation standard is not met, the adjustment range of material properties, boundary data, and tunnel mesh is determined based on tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data.

[0027] Within the adjustment ranges of material properties, boundary data, and tunnel mesh, the preset adjustment step sizes for material properties, boundary data, and tunnel mesh are used to continuously adjust the material properties, boundary data, and tunnel mesh. After each adjustment, a simulation test is performed again until the tunnel 3D model passes the simulation test.

[0028] Furthermore, after the tunnel 3D model passes simulation testing, the method also includes:

[0029] Real-time tunnel damage data is input into the tunnel's 3D model to obtain the real-time damage causes and effects, and then the real-time damage causes and effects are sent to the control center.

[0030] Furthermore, the training process for the pre-defined maintenance detection model includes:

[0031] Obtain historical maintenance rating records, and determine historical tunnel operation data and historical maintenance ratings based on the historical maintenance records;

[0032] The maintenance rating labels are labeled with the maintenance ratings of the historical tunnel operation data according to the historical maintenance ratings, and the maintenance rating dataset is determined according to the historical tunnel operation data and the maintenance rating labels.

[0033] Based on the historical tunnel operation data and the preset maintenance scoring threshold, abnormal historical tunnel operation data is determined, and the historical maintenance type of the abnormal historical tunnel operation data is obtained.

[0034] A maintenance type dataset is created by labeling the abnormal historical tunnel operation data with the maintenance type labels based on the historical maintenance type;

[0035] Construct a maintenance detection model, and train the maintenance detection model based on the maintenance score dataset and the maintenance type dataset until the maintenance detection model reaches the preset maintenance detection model training standard.

[0036] Furthermore, after determining real-time tunnel damage data through the tunnel damage monitoring model, the method further includes:

[0037] The real-time tunnel damage data, tunnel operation data, and real-time damage score are hashed to obtain the hash values ​​corresponding to the real-time tunnel damage data, tunnel operation data, and real-time damage score.

[0038] Real-time tunnel damage data, tunnel operation data, and the hash values ​​corresponding to real-time damage scores are stored on the blockchain according to a pre-set smart contract.

[0039] Furthermore, after obtaining real-time tunnel damage data, tunnel operation data, and the hash value corresponding to the real-time damage score, the method further includes:

[0040] Zero-knowledge proofs are generated based on real-time tunnel damage data, tunnel operation data, real-time damage scores, and the hash values ​​corresponding to the real-time tunnel damage data, tunnel operation data, and real-time damage scores.

[0041] Accordingly, real-time tunnel damage data, tunnel operation data, and the hash values ​​corresponding to real-time damage scores are stored on the blockchain according to a pre-set smart contract, including:

[0042] The zero-knowledge proof, along with the hash values ​​corresponding to real-time tunnel damage data, tunnel operation data, and real-time damage scores, are stored in the blockchain according to a preset smart contract.

[0043] According to a second aspect of this application, a tunnel damage monitoring system is provided, the system comprising:

[0044] The data acquisition module is used to acquire tunnel operation data and tunnel damage data at preset acquisition time points during the initial stage of tunnel operation, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the workers' handheld terminals.

[0045] The scoring dataset creation module is used to receive damage scores of tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage scores, and create a scoring dataset based on the tunnel operation data and the damage score tags.

[0046] The data segmentation module is used to obtain a preset damage score threshold and to segment the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data based on the damage score and the damage score threshold.

[0047] The damage dataset creation module is used to determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data based on the tunnel damage data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags.

[0048] The model building module is used to build a tunnel damage monitoring model and train the damage monitoring model based on the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard.

[0049] The real-time data determination module is used to continuously collect real-time tunnel operation data, input the real-time tunnel operation data into the tunnel damage monitoring model, obtain a real-time damage score, and if the real-time damage score exceeds a preset damage score threshold, determine the real-time tunnel damage data through the tunnel damage monitoring model.

[0050] Compared with the prior art, this application has the following beneficial effects:

[0051] This application's embodiment utilizes the aforementioned tunnel damage monitoring method. By automating the collection and analysis of tunnel operation data, it avoids the tedious and time-consuming nature of manual inspection, significantly improving monitoring efficiency, especially in long tunnels or tunnel groups. Employing various high-precision sensors such as piezoelectric ceramic sensors, fiber optic sensors, and accelerometers, combined with a tunnel damage monitoring model, it can capture minute changes in the tunnel structure, improving monitoring accuracy and reducing missed detections and misjudgments. Real-time collection and analysis of tunnel operation data allows for the timely detection of potential damage or anomalies, automatically triggering early warning mechanisms to ensure tunnel safety and prevent accidents. Automated data processing using a model improves data processing efficiency. By analyzing historical and real-time data, the model captures patterns and trends in the actual situation, ensuring results match reality and providing a scientific basis for formulating maintenance and reinforcement plans, thus improving the accuracy and reliability of decision-making. Hash processing and blockchain storage technology ensure the integrity and immutability of tunnel damage data, improving data security and providing a reliable foundation for subsequent data analysis and decision-making. Real-time monitoring and analysis enable the early identification of potential safety hazards, preventing accidents and reducing maintenance costs caused by accidents. Meanwhile, data-driven maintenance and inspection models can guide reasonable maintenance plans, avoid over-maintenance, and further reduce maintenance costs. Storing data using blockchain technology and zero-knowledge proofs improves data transparency and traceability, enhancing public trust in tunnel safety. Attached Figure Description

[0052] Figure 1 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 1 of this application;

[0053] Figure 2 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 2 of this application;

[0054] Figure 3 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 3 of this application;

[0055] Figure 4 This is a schematic diagram of the tunnel damage monitoring system provided in Embodiment 4 of this application; Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0057] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0058] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0059] The tunnel damage monitoring method provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0060] Example 1

[0061] Figure 1 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following:

[0062] S101, during the initial stage of tunnel operation, acquire tunnel operation data and tunnel damage data at each preset acquisition time point, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the staff's handheld terminal.

[0063] Firstly, this solution can be used in scenarios where data is collected during the initial stage of tunnel operation to establish a damage monitoring model, and then the model is deployed to collect data in real time after it meets the standards, in order to determine the damage score and damage data.

[0064] Based on the above application scenarios, it is understandable that the implementing entity of this application can be a tunnel damage monitoring system, and no further restrictions are imposed here.

[0065] The initial stage of tunnel operation can refer to the early period after a tunnel has just been put into use or after significant construction work has been completed. During this stage, monitoring data is particularly important to ensure structural safety and identify potential damage.

[0066] Each preset data collection time point refers to a time point pre-set during the monitoring process. These time points are used to collect data on tunnel operation and damage. For example, data collection can be set to occur hourly, daily, or after a specific event (such as rainfall or an earthquake).

[0067] Tunnel operation data can include data reflecting the tunnel's structural state and environmental conditions. Specifically, tunnel stress data describes the distribution of forces within the tunnel; tunnel strain data describes the degree of deformation of tunnel materials under external forces; tunnel temperature data records temperature changes within the tunnel, which may affect material properties and structural safety; and tunnel vibration data reflects the vibration of the tunnel structure during operation or under the influence of external factors.

[0068] Tunnel damage data can describe the damage to the tunnel structure. Tunnel displacement data can reflect the amount of displacement of the tunnel in a specific direction. Tunnel deformation data can describe changes in the tunnel's shape, including both lateral and longitudinal variations. Tunnel crack data can record the occurrence, location, and size of cracks in the tunnel structure.

[0069] The control center can be a centralized monitoring and data analysis facility responsible for receiving, storing, and analyzing data collected from the tunnel, and for real-time monitoring and early warning.

[0070] Piezoelectric ceramic sensors are sensors that operate using the piezoelectric effect and can measure changes in stress and strain. They generate electrical signals when subjected to mechanical stress, and therefore can be used to monitor minute deformations in tunnel structures.

[0071] Fiber optic sensors are sensors that use fiber optic technology to measure physical quantities (such as temperature, strain, etc.). Fiber optic sensors have high sensitivity and anti-interference capabilities.

[0072] An accelerometer can be a sensor used to measure vibrations and motion inside a tunnel, and can detect changes in acceleration of the tunnel under the action of external forces (such as earthquakes or traffic).

[0073] The handheld terminal can be a portable device that workers use to collect and record tunnel damage data in real time, such as displacement, deformation, and crack conditions.

[0074] The frequency and timing of data acquisition can be set according to the tunnel's operation and environmental conditions. For example, data can be collected hourly, daily, or after specific events (such as rainfall or increased traffic load). Piezoelectric ceramic sensors are installed at critical locations in the tunnel structure to monitor stress and strain in real time. Fiber optic sensors are deployed inside the tunnel to measure temperature changes and can be used in conjunction with strain sensors to improve data accuracy. Accelerometers are placed at different locations in the tunnel to monitor vibration and assess external influences. The sensor system then collects data automatically and periodically to ensure real-time data accuracy. Workers inside the tunnel use handheld devices to regularly inspect and record displacement, deformation, and crack conditions. This can be done manually or through image recognition. The real-time acquired data is transmitted to the control center using wireless communication technologies such as Wi-Fi, LTE, or a dedicated wireless network.

[0075] S102, Receive the damage score of the tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage score, and create a score dataset according to the tunnel operation data and the damage score tags.

[0076] A damage score is a numerical value derived from the analysis of tunnel operation data, representing the health status or degree of damage to the tunnel structure under specific conditions. The score is typically calculated based on a series of parameters (such as stress, strain, and temperature); a higher score indicates a greater risk of damage.

[0077] Damage score labels can directly use the numerical values ​​of the damage score to represent the health status of the tunnel. For example, the damage score can be used directly as a label (such as 0.2, 0.5, 0.8, etc.), and these values ​​can be used in subsequent regression analysis or other models that require numerical input.

[0078] A rating dataset can be a database that aggregates all tunnel operation data and their corresponding damage rating labels, and is typically used to train and validate damage monitoring models. Each record includes tunnel operation data (such as stress, strain, temperature, etc.) and the corresponding damage rating label.

[0079] The system can receive real-time tunnel operation data and calculated damage scores from the control center. The received damage scores are then correlated with corresponding tunnel operation data (such as stress, strain, and temperature). A corresponding damage score label is assigned to each piece of tunnel operation data. Finally, the tunnel operation data labeled with damage scores are compiled into a score dataset.

[0080] S103, obtain a preset damage score threshold, and divide the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data according to the damage score and the damage score threshold.

[0081] The preset damage scoring threshold can be a specific numerical value used to distinguish between normal and abnormal tunnel operating states. For example, if the damage score is higher than this threshold, it is considered that there is potential damage.

[0082] Normal tunnel operation data can be data with a damage score lower than or equal to a preset threshold, indicating that the tunnel is in a safe operating state.

[0083] Abnormal tunnel operation data can be data with a damage score higher than a preset threshold, indicating that there is potential damage or risk to the tunnel.

[0084] Preset damage scoring thresholds can be obtained from system configuration or database. Each tunnel's operational data is iterated through, and the damage score is compared to the threshold. Tunnel operational data below or equal to the preset damage scoring threshold is considered normal tunnel operational data, while tunnel operational data above the preset damage scoring threshold is considered abnormal tunnel operational data.

[0085] S104, determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data based on the tunnel damage data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags.

[0086] Abnormal tunnel damage data can refer to damage information corresponding to abnormal tunnel operation data. It can include detailed data on the actual damage to the tunnel, such as displacement, deformation, cracks, etc.

[0087] Tunnel damage labels are specific numerical values ​​that mark tunnel damage, quantifying its degree, type, or location. For example, damage labels could be crack width (in millimeters), displacement (in millimeters), or damage classification (e.g., minor, moderate, severe). These labels play a crucial role in analysis and modeling, helping to build effective damage monitoring models.

[0088] Damage datasets can be collections containing abnormal tunnel operation data and their corresponding damage labels, and are typically used for subsequent data analysis, model training, or decision support.

[0089] Damage scoring analysis of tunnel operation data can identify abnormal tunnel operation data. These abnormal data are typically records whose damage scores exceed preset thresholds. The process involves determining which tunnel damage data (such as displacement, deformation, cracks, etc.) are associated with the identified abnormal operation data. This can be done by using timestamps, sensor locations, or other identifiers to map the abnormal operation data to the damage data. Then, based on the identified abnormal tunnel damage data, each abnormal tunnel operation data point is labeled with a corresponding tunnel damage tag. Finally, all abnormal tunnel operation data and their corresponding tunnel damage tags are aggregated to create a damage dataset. This dataset will provide the foundation for subsequent analysis and modeling, helping to build more effective damage monitoring models.

[0090] S105, Construct a tunnel damage monitoring model, and train the damage monitoring model based on the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard.

[0091] A tunnel damage monitoring model can be a machine learning or deep learning-based model used to identify and assess tunnel damage. This model predicts the tunnel's health status and potential damage by analyzing input tunnel operation data (such as stress, displacement, and temperature) and corresponding damage data (such as crack width and displacement).

[0092] Pre-defined training metrics for damage monitoring models can refer to indicators used to evaluate model performance during training. Specifically, these may include: Accuracy: the proportion of correctly predicted damages; Precision and Recall: measuring the model's accuracy and false negative rate in damage detection, respectively; F1 Score: the harmonic mean of precision and recall, used to comprehensively evaluate model performance; and Loss Function: a metric monitored during training, typically aimed at minimizing.

[0093] The scoring and damage datasets can be formatted to suit the model input. Data preprocessing includes standardization, noise reduction, and handling of missing values. A suitable model architecture, such as support vector machines, decision trees, or neural networks, should be selected based on the data characteristics. The model should be trained using the scoring and damage datasets, with parameters adjusted to improve performance. During training, the model's performance on the validation set should be validated periodically to ensure it is not overfitting. The model's performance should be evaluated on the test set, calculating metrics such as accuracy, precision, and recall. The model should be compared to predefined damage monitoring model training standards to determine if it meets the requirements. If the model performance does not meet the predefined standards, optimization can be achieved by adjusting model parameters, improving data quality, or increasing the amount of training data. Once the model meets the predefined standards, it should be saved and deployed in a real-world application for real-time tunnel damage monitoring.

[0094] S106, continuously collect real-time tunnel operation data, input the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, and if the real-time damage score exceeds a preset damage score threshold, determine the real-time tunnel damage data through the tunnel damage monitoring model.

[0095] Real-time tunnel operation data refers to the operational status data collected in real time by various sensors and monitoring equipment during tunnel operation. Specifically, this can include stress data (the stress state of the tunnel structure), strain data (the deformation of the tunnel materials), temperature data (temperature changes inside and outside the tunnel), and vibration data (the dynamic response of the tunnel, such as vibrations caused by vehicle traffic or other external factors).

[0096] Real-time damage score is a numerical indicator generated by analyzing real-time tunnel operation data using a tunnel damage monitoring model. This score quantifies the health status of the tunnel, typically calculating damage-related characteristics based on input data to reflect the potential degree of damage. A higher score indicates a greater risk of damage.

[0097] Real-time tunnel damage data refers to the specific damage information determined by the model when the real-time damage score exceeds a preset threshold. Specifically, this can include damage type (e.g., cracks, displacement, deformation), damage location (the exact location where the damage occurred), and damage severity (quantitative information such as crack width and displacement).

[0098] Real-time tunnel operation data can be continuously collected via sensors to ensure data accuracy and completeness. The collected data is cleaned and preprocessed to remove noise and outliers, ensuring input data quality. The processed real-time tunnel operation data is then input into a pre-trained tunnel damage monitoring model. The model infers from the input real-time tunnel operation data and calculates a real-time damage score. The real-time damage score reflects the current health status of the tunnel and is typically a numerical value. The real-time damage score is compared with a preset damage score threshold. If the real-time damage score exceeds the preset threshold, the tunnel damage monitoring model analyzes the input data to determine the corresponding real-time tunnel damage data.

[0099] The technical solution provided in this embodiment acquires tunnel operation data and tunnel damage data at preset collection time points during the initial stage of tunnel operation, and transmits the tunnel operation data and tunnel damage data to the control center. The tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers. The tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by workers' handheld terminals. The system receives damage scores for the tunnel operation data sent by the control center, labels the tunnel operation data with damage score tags based on the damage scores, and creates a score dataset based on the tunnel operation data and damage score tags. A preset damage score threshold is obtained, and the data is then processed according to the damage scores and damage... The tunnel damage scoring threshold divides tunnel operation data into normal and abnormal tunnel operation data. Based on the tunnel damage data, abnormal tunnel damage data corresponding to the abnormal tunnel operation data is determined. Tunnel damage labels are then labeled for the abnormal tunnel damage data, and a damage dataset is created based on the abnormal tunnel operation data and the tunnel damage labels. A tunnel damage monitoring model is constructed and trained using the scoring dataset and the damage dataset until the damage monitoring model reaches a preset training standard. Real-time tunnel operation data is continuously collected and input into the tunnel damage monitoring model to obtain a real-time damage score. If the real-time damage score exceeds a preset damage scoring threshold, the real-time tunnel damage data is determined through the tunnel damage monitoring model. Through the above tunnel damage monitoring method, potential damage or abnormalities can be detected in a timely manner by continuously collecting and analyzing tunnel operation data, ensuring tunnel safety. The model automates data processing, improving data processing efficiency. By analyzing historical and real-time data, the model captures patterns and trends in real-world situations, ensuring that the results match the actual situation and improving the reliability of the analysis results.

[0100] Based on the above technical solution, optionally, after inputting the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, the method further includes:

[0101] If the real-time damage score is lower than the preset damage score threshold, the tunnel operation data is input into the preset maintenance detection model to obtain the real-time maintenance score.

[0102] If the real-time maintenance score exceeds the preset maintenance score threshold, the maintenance type is determined by the preset maintenance detection model and the maintenance type is transmitted to the control center.

[0103] In this solution, the pre-defined maintenance detection model can be an algorithm or model developed based on historical data and expert knowledge, used to assess the maintenance needs of equipment or structures (such as tunnels). This model may include machine learning algorithms, rule engines, or physics-based models to analyze data and generate maintenance recommendations.

[0104] Real-time maintenance score is a quantitative value generated by a preset maintenance detection model, reflecting the current maintenance needs or status of the tunnel.

[0105] The preset maintenance score threshold can be a set value. When the real-time maintenance score exceeds this value, the system will consider that the tunnel needs maintenance.

[0106] Maintenance type refers to the specific maintenance measures or operations determined based on real-time maintenance scores and analysis results, which may include regular inspections, cleaning, maintenance, and replacement of parts.

[0107] The system continuously acquires tunnel operational data (such as stress, temperature, and displacement) and inputs this data into a maintenance monitoring model. The model analyzes the real-time operational data using a preset maintenance monitoring model to generate a real-time maintenance score. This score is then compared to a preset maintenance score threshold. If the score is below the threshold, maintenance is not required; if it exceeds the threshold, the maintenance type needs further determination. If the real-time maintenance score exceeds the threshold, the preset maintenance monitoring model analyzes the current status to determine the specific maintenance type. The determined maintenance type information is then transmitted to the control center for subsequent maintenance scheduling and execution.

[0108] This solution utilizes real-time monitoring and analysis to quickly identify potential maintenance needs, reduce accident risks, and ensure the safe operation of the tunnel. Analysis based on real-time data can reduce unnecessary maintenance and avoid over-maintenance, thereby lowering maintenance costs.

[0109] Based on the above technical solution, the optional, preset training process of the maintenance detection model includes:

[0110] Obtain historical maintenance rating records, and determine historical tunnel operation data and historical maintenance ratings based on the historical maintenance records;

[0111] The maintenance rating labels are labeled with the maintenance ratings of the historical tunnel operation data according to the historical maintenance ratings, and the maintenance rating dataset is determined according to the historical tunnel operation data and the maintenance rating labels.

[0112] Based on the historical tunnel operation data and the preset maintenance scoring threshold, abnormal historical tunnel operation data is determined, and the historical maintenance type of the abnormal historical tunnel operation data is obtained.

[0113] A maintenance type dataset is created by labeling the abnormal historical tunnel operation data with the maintenance type labels based on the historical maintenance type;

[0114] Construct a maintenance detection model, and train the maintenance detection model based on the maintenance score dataset and the maintenance type dataset until the maintenance detection model reaches the preset maintenance detection model training standard.

[0115] In this plan, the historical maintenance rating record can record the rating information of past tunnel maintenance, including the effectiveness, quality and implementation of the maintenance.

[0116] Historical tunnel operation data can include monitoring data of the tunnel during historical periods, such as operating parameters like stress, displacement, and temperature, reflecting the tunnel's condition and performance.

[0117] Historical maintenance ratings can be ratings of maintenance effectiveness based on historical tunnel operation data, typically derived from evaluations of operational data after actual maintenance.

[0118] Maintenance rating labels can be used to mark rating values ​​for historical tunnel operation data, in order to facilitate subsequent data analysis and model training.

[0119] The maintenance rating dataset can be a collection of data containing historical tunnel operation data and their corresponding maintenance rating labels, used to train and evaluate maintenance detection models.

[0120] Abnormal historical tunnel operation data refers to tunnel operation data that exhibits abnormal characteristics in historical records, usually caused by operating conditions or other factors that exceed the normal range.

[0121] Historical maintenance type can refer to the maintenance category performed in response to specific abnormal situations in the historical record, such as routine maintenance, emergency repairs, preventive maintenance, etc.

[0122] Maintenance type labels can be used to mark the maintenance type corresponding to abnormal historical tunnel operation data, so as to facilitate subsequent analysis and model training.

[0123] The maintenance type dataset can be a collection of data containing abnormal historical tunnel operation data and their corresponding maintenance type labels, which can be used to train a maintenance detection model.

[0124] The pre-defined training criteria for the maintenance detection model can be the performance metrics that the model needs to achieve during the training process, such as accuracy, recall, or other key performance metrics, to ensure the effectiveness of the model.

[0125] Historical maintenance records can be extracted from the maintenance management system or database. These records typically include information such as the date, type, personnel, content, and score of each maintenance session. Ensuring the completeness and accuracy of these records may require confirmation with relevant management or maintenance personnel. By associating the dates and times in the maintenance records, tunnel operation data (such as stress, displacement, and temperature) for the corresponding time period can be extracted from the monitoring system. A maintenance score is calculated for each maintenance record based on the corresponding tunnel operation data, ensuring standardization and consistency. Historical maintenance scores are then labeled with their corresponding historical tunnel operation data, assigning a maintenance score label to each historical data point. The historical tunnel operation data and their corresponding maintenance score labels are then combined to form a maintenance score dataset. The dataset must be cleaned and organized for subsequent model training. Using a preset maintenance score threshold, the historical tunnel operation data is analyzed to identify abnormal historical data with scores below the threshold. The timestamps and relevant characteristics of these abnormal data are recorded for subsequent analysis. By finding maintenance records corresponding to the abnormal historical tunnel operation data, the historical maintenance type (such as emergency maintenance or periodic inspection) of this data is extracted. Maintenance type labels are assigned to the abnormal historical tunnel operation data based on the historical maintenance type for subsequent data analysis and modeling. Then, the abnormal historical tunnel operation data and their corresponding maintenance type labels are combined to form a maintenance type dataset. Ensure the dataset format is consistent and perform appropriate data cleaning. Select an appropriate machine learning algorithm (such as decision tree, random forest, or neural network) to build the maintenance detection model. Train the model using the maintenance score dataset and the maintenance type dataset, adjusting hyperparameters to optimize model performance. Continue training until the model reaches the preset maintenance detection model training standards (such as accuracy, recall, etc.) to ensure its effectiveness in predicting maintenance needs.

[0126] This solution trains a more accurate maintenance detection model, which can improve its ability to predict future maintenance needs, thereby achieving more effective resource allocation.

[0127] Example 2

[0128] Figure 2 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 2 of this application, as shown below. Figure 2 As shown, the specific method includes the following steps:

[0129] S201, acquire tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data, and create a three-dimensional tunnel model based on the tunnel geometry data.

[0130] Tunnel geometry data can include the tunnel's shape and dimensions, such as: Length: the overall length of the tunnel; Width: the cross-sectional width of the tunnel; Height: the longitudinal height of the tunnel; Shape: such as circular, rectangular, or other special shapes; Gradient: the tunnel's slope or gradient information.

[0131] Tunnel material data can include the physical and mechanical properties of the various materials that make up the tunnel, such as: concrete: elastic modulus, compressive strength, density, etc.; reinforcing steel: tensile strength, yield strength, etc.; lining materials: performance data such as epoxy resin or other special materials.

[0132] Tunnel boundary data defines the tunnel's constraints and boundary conditions, such as fixed points (fixed support points at both ends of the tunnel or specific locations), free ends (open ends or unfixed locations of the tunnel), and contact conditions (such as the contact relationship between the soil and the tunnel lining).

[0133] Tunnel load data can refer to various external forces and influences applied to the tunnel, such as traffic load: the weight of vehicles and their distribution within the tunnel; earth pressure: the lateral pressure exerted by the surrounding soil on the tunnel structure; wind pressure: the force exerted by external airflow on the tunnel surface; and seismic force: the force exerted on the tunnel during an earthquake.

[0134] A tunnel 3D model can be a computer-generated 3D representation used to simulate and analyze the physical structure of a tunnel. It includes the aforementioned geometric, material, boundary, and load data, and is constructed using computer-aided design (CAD) software or finite element analysis (FEA) tools.

[0135] Laser rangefinders, total stations, and other tools can be used for on-site measurements to obtain the tunnel's length, width, height, and shape. Design and construction drawings are reviewed to extract relevant geometric information. Geological survey data is obtained to understand the topographical changes at the tunnel's location, and the above data is combined into geometric data. On-site sampling and laboratory testing are conducted to determine the physical and mechanical properties of materials such as concrete and steel reinforcement. Property data of commonly used materials is obtained by referring to relevant material standards and specifications. Material selection records from the design phase are reviewed to obtain tunnel material data. The fixed and free boundary locations of the tunnel are determined based on the design drawings, and the tunnel's support structure is inspected on-site, recording support points and contact conditions to obtain tunnel boundary data. Traffic monitoring is used to obtain data on the number and weight of passing vehicles. The physical properties of the surrounding soil are obtained to calculate earth pressure. Load test data and seismic records from similar tunnels are referenced to obtain tunnel load data.

[0136] The acquired geometric data can be input using computer-aided design (CAD) software (such as AutoCAD and Revit) or 3D modeling tools (such as SketchUp and 3ds Max). In the modeling software, specify the corresponding material properties for the model, including mechanical properties. Define fixed points, free ends, and contact conditions based on the acquired boundary data. Apply traffic loads, earth pressures, etc., to the model, setting the load location and magnitude according to calculation requirements.

[0137] S202, set the material properties of the tunnel three-dimensional model according to the tunnel material data, and set the boundary data of the tunnel three-dimensional model according to the tunnel boundary data.

[0138] Material properties refer to physical and mechanical parameters that describe the characteristics of a material. Specifically, these may include: elastic modulus: the stress-to-strain ratio within the elastic range of a material; Poisson's ratio: the deformation perpendicular to a material when it is stretched or compressed in one direction; density: the mass-to-volume ratio of a material; compressive strength and tensile strength: the maximum compressive and tensile forces a material can withstand under stress; and shear strength: the material's ability to resist shear forces.

[0139] Boundary data can define the boundary conditions of a model. Specifically, it can include fixed boundaries: certain parts of the model are fixed and cannot move; free boundaries: certain parts of the model can move freely and are affected by external forces; and contact conditions: the contact behavior between different materials, such as friction and adhesion.

[0140] In the modeling software, select the parts whose material properties need to be set (such as concrete, steel reinforcement, etc.). Input the acquired material data, ensuring that parameters such as elastic modulus, Poisson's ratio, and density are accurate. If the software supports this, select the corresponding material from the material library, and the relevant properties will be automatically filled in. Based on the design drawings and site survey results, determine the location of each boundary in the modeling software. For fixed boundaries, select the corresponding face or edge in the model and set it to "fixed". For free boundaries, ensure that no constraints are applied, allowing them to deform freely. If there are contact conditions, set the contact surface and friction coefficient to ensure realistic simulation of the interaction.

[0141] S203, based on the tunnel geometry data and tunnel material data, divide the tunnel three-dimensional model into a tunnel mesh, and input the tunnel load data into the tunnel three-dimensional model.

[0142] Tunnel meshing is the process of dividing a 3D tunnel model into multiple small units (mesh). These units can be triangles, quadrilaterals, hexahedrons, etc., and the purpose is to enable more accurate calculation of stress, strain, and other physical quantities within the model during numerical simulation. The quality of mesh generation directly affects the accuracy and computational efficiency of the simulation.

[0143] Mesh generation tools can be found in the modeling or simulation software used. Set the mesh size according to the tunnel's geometric complexity and the required simulation accuracy. Smaller meshes improve accuracy but increase computational cost. Choose a suitable mesh type (e.g., triangular, quadrilateral, hexahedral, etc.) based on the model's geometry. Use the software's automatic meshing function or manually select specific areas for meshing. Ensure the mesh has sufficient fineness in critical tunnel areas (e.g., supports, joints) to accurately capture stress changes. Check the generated mesh for distortion, overlap, or irregular shapes to ensure good mesh quality. Mesh quality check tools in the software can be used to evaluate mesh quality. After meshing, apply tunnel load data (e.g., static loads, dynamic loads, etc.) to the model according to design requirements and actual conditions. Ensure the load application location and direction are correct, and set time-varying patterns as needed.

[0144] In this embodiment, by acquiring detailed geometry, material, boundary, and load data, a more accurate three-dimensional model can be constructed, enhancing the realism and reliability of the simulation.

[0145] Based on the above technical solution, optionally, after meshing the tunnel three-dimensional model according to the tunnel geometric data and tunnel material data, and inputting tunnel load data into the tunnel three-dimensional model, the method further includes:

[0146] Obtain historical monitoring results, and determine historical tunnel damage data, causes of historical damage, and impacts of historical damage based on the historical monitoring results;

[0147] Obtain simulation requirements, set simulation parameters for the tunnel 3D model according to the simulation requirements, input the historical tunnel damage data into the tunnel 3D model for simulation testing, and obtain tunnel simulation results;

[0148] Based on the tunnel simulation results, historical damage causes, and historical damage impacts, determine whether the tunnel 3D model meets the preset simulation standards.

[0149] If the preset simulation standards are met, the tunnel 3D model is deemed to have passed the simulation test.

[0150] In this scheme, historical monitoring results can be data collected from past tunnel monitoring, including information such as stress, deformation, and displacement recorded by various sensors.

[0151] Historical tunnel damage data can be a record of tunnel damage determined based on historical monitoring results, and may include specific values ​​such as crack width and displacement.

[0152] Historical damage causes can be known causes of tunnel damage, such as soil liquefaction, construction defects, geological changes, etc.

[0153] Historical damage impact can be an assessment of the effects of past tunnel damage on structural safety, function, stability, and other aspects.

[0154] Simulation requirements can be the requirements for simulation analysis needed for a specific purpose or problem, such as the working conditions to be tested, the environmental conditions to be simulated, etc.

[0155] Simulation parameters can be parameters required when setting up a simulation model, including material properties, load conditions, boundary conditions, time steps, etc.

[0156] The tunnel simulation results should include specific data related to the causes and effects of historical damage. For example, this may include indicators such as stress distribution, degree of deformation, and crack width.

[0157] The preset simulation criteria can be used to judge whether the simulation test results are qualified, and may include thresholds for similarity with historical damage causes and effects.

[0158] Monitoring data from a past period can be collected and organized, potentially from various sensors, inspection records, and monitoring reports. Specific tunnel damage data, including parameters such as crack width, displacement, and deformation, can be extracted from historical monitoring results. This data provides the foundation for analyzing the tunnel's condition. Historical damage data is analyzed to determine the causes of the damage, which may include factors such as construction defects, geological changes, and load effects. The actual impact of historical damage on the tunnel is assessed, including its effect on structural stability, increased maintenance costs, or safety hazards. Then, based on engineering needs, expected analysis objectives, and design requirements, the specific content and direction of the simulation are clarified. Simulation parameters for the tunnel's 3D model are configured according to simulation requirements, including material properties, load conditions, and boundary conditions. The identified historical tunnel damage data is input into the tunnel's 3D model for simulation testing based on actual damage conditions. The simulation model is run to generate tunnel simulation results, including analytical data on stress, deformation, and crack development. The tunnel simulation results are compared with the historical damage causes and effects to assess their similarity. A similarity threshold is set to determine whether the model results meet expectations. If the simulation results reach the preset similarity standard, the tunnel's 3D model is confirmed to have passed the simulation test.

[0159] In this approach, a simulation model is used to verify the tunnel's performance under different damage conditions, ensuring the model's accuracy and reliability, thereby enhancing the credibility of the engineering design. By comparing the similarity of the simulation results, the tunnel's risks can be effectively assessed, potential safety hazards can be identified in advance, and necessary preventative measures can be taken.

[0160] Based on the above technical solution, optionally, after determining whether the tunnel 3D model meets the preset simulation standard according to the tunnel simulation results and historical tunnel damage data, the method further includes:

[0161] If the preset simulation standard is not met, the adjustment range of material properties, boundary data, and tunnel mesh is determined based on tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data.

[0162] Within the adjustment ranges of material properties, boundary data, and tunnel mesh, the preset adjustment step sizes for material properties, boundary data, and tunnel mesh are used to continuously adjust the material properties, boundary data, and tunnel mesh. After each adjustment, a simulation test is performed again until the tunnel 3D model passes the simulation test.

[0163] In this scheme, the material property adjustment range refers to the range within which material properties (such as elastic modulus, strength, density, etc.) can be adjusted during the simulation process.

[0164] The range of boundary data adjustment can be the range of boundary conditions that can be adjusted in the simulation.

[0165] The range of tunnel mesh adjustment can be the range of mesh refinement in the tunnel model.

[0166] The preset material property adjustment step size can be the increment each time the material property is adjusted.

[0167] The preset boundary data adjustment step size can be the amount of change when adjusting boundary conditions.

[0168] The preset tunnel mesh adjustment step size can be the step size when adjusting mesh refinement.

[0169] Based on existing tunnel geometry, material, boundary, and load data, the adjustment ranges for each item can be defined. The specific increment for each adjustment should be determined to ensure that parameter changes during the adjustment process are reasonable and controllable. Adjustments can be made within the material property adjustment range, using the set step size. Adjustments can also be made within the boundary data adjustment range and the tunnel mesh adjustment range. After each adjustment, a simulation test should be performed again, and the results recorded. The simulation results should be checked to see if they meet the preset standards. If they do not meet the standards, the next round of adjustments should be performed until the preset simulation standards are met.

[0170] In this scheme, through meticulous parameter adjustments, the actual situation of the tunnel can be reflected more accurately, ensuring a higher degree of matching between the model and the real environment, thereby improving the credibility of the simulation results.

[0171] Based on the above technical solution, optionally, after the tunnel 3D model passes simulation testing, the method further includes:

[0172] Real-time tunnel damage data is input into the tunnel's 3D model to obtain the real-time damage causes and effects, and then the real-time damage causes and effects are sent to the control center.

[0173] In this scheme, real-time damage causes can refer to specific factors or events that lead to tunnel damage, such as: overload: vehicles or other objects exceeding the design load passing through the tunnel; environmental factors: such as the effects of high temperature, humidity, and soil changes; construction defects: damage caused by improper construction or material quality problems; and long-term use: natural aging or fatigue damage caused by long-term use of the tunnel.

[0174] Real-time damage impacts can refer to the specific effects of damage on tunnel safety, stability, and performance. For example, decreased structural stability may increase the risk of tunnel collapse. Operational safety hazards may affect normal traffic and increase the risk of accidents. Increased maintenance costs may necessitate earlier repairs or reinforcement, increasing maintenance expenses.

[0175] Real-time damage data, including displacement, crack width, and strain, can be acquired from sensors and input into a 3D tunnel model. The model is then used to analyze the real-time damage data and identify potential causes of damage. The model may combine historical data, simulation results, and real-time monitoring data for inference. Algorithms calculate the real-time damage impact, assessing the degree of influence of current damage on the tunnel structure. Based on the analysis results, a report on the causes and impacts of real-time damage is generated, including a detailed description of each damage and its potential consequences. This information is then transmitted to the control center via a data transmission system (such as a wireless or wired network) for evaluation and decision-making by relevant personnel.

[0176] This solution utilizes real-time monitoring and analysis to quickly identify potential safety hazards, enabling timely intervention to prevent accidents. Understanding the causes and impacts of damage helps in developing more effective maintenance and repair plans, reducing maintenance costs.

[0177] Example 3

[0178] Figure 3 This is a schematic flowchart of the tunnel damage monitoring method provided in Embodiment 3 of this application, as shown below. Figure 3 As shown, the specific method includes the following steps:

[0179] S301, during the initial stage of tunnel operation, acquire tunnel operation data and tunnel damage data at each preset acquisition time point, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the staff's handheld terminal.

[0180] S302, Receive damage scores for tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage scores, and create a score dataset based on the tunnel operation data and the damage score tags.

[0181] S303, obtain a preset damage score threshold, and divide the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data according to the damage score and the damage score threshold.

[0182] S304, determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data based on the tunnel damage data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags.

[0183] S305, Construct a tunnel damage monitoring model, and train the damage monitoring model based on the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard.

[0184] S306, continuously collect real-time tunnel operation data, input the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, and if the real-time damage score exceeds a preset damage score threshold, determine the real-time tunnel damage data through the tunnel damage monitoring model.

[0185] S307, the real-time tunnel damage data, tunnel operation data and real-time damage score are hashed respectively to obtain the hash values ​​corresponding to the real-time tunnel damage data, tunnel operation data and real-time damage score.

[0186] A hash value is a string (usually a combination of numbers and letters) that transforms input data (such as real-time damage data, runtime data, and damage scores) into a fixed length using a hash function, and is used to uniquely identify the original data.

[0187] You can choose a suitable hash algorithm, such as SHA-256 or MD5. Generally, a more secure and widely used hash algorithm is chosen. Convert the real-time tunnel damage data, tunnel operation data, and real-time damage scores into string format. If the data is numerical or in another format, it needs to be converted to a string first. Input these strings into the selected hash algorithm, and the algorithm will generate the corresponding hash value.

[0188] S308 stores real-time tunnel damage data, tunnel operation data, and the hash value corresponding to the real-time damage score on the blockchain according to a preset smart contract.

[0189] A pre-defined smart contract can be a piece of code that executes automatically on the blockchain, defining the rules for data storage and access. Smart contracts can ensure data integrity and immutability, and execute automatically when specific conditions are met. For example, a smart contract can define the structure for storing hash values, the method for verifying data, and the permissions for authorizing access.

[0190] A smart contract can be written, including: a data structure defining the data type to be stored (e.g., hash value, timestamp, source, etc.); a storage function implementing a function to receive the hash value and store it in the blockchain; and access control defining who can call the storage function, ensuring only authorized users or systems can operate on it. The smart contract is then deployed to the blockchain network. This process uploads the smart contract's code to the blockchain, making it immutable. During real-time data processing, real-time tunnel damage data, tunnel operation data, and real-time damage scores are hashed to obtain corresponding hash values. The hash value is sent to the smart contract using the storage function provided by the contract: a blockchain network client (e.g., Web3.js or Ethers.js) is used to connect to the blockchain. The storage function in the smart contract is called, passing the hash value as a parameter. After submitting the storage request, the transaction is awaited confirmation from the blockchain network. Upon confirmation, the hash value is permanently stored on the blockchain and recorded along with the timestamp and other relevant information, ensuring the data's immutability and traceability.

[0191] In this solution, blockchain storage ensures data security and transparency, preventing tampering. Smart contracts can execute automatically, improving efficiency and reducing human intervention.

[0192] Based on the above technical solution, optionally, after obtaining real-time tunnel damage data, tunnel operation data, and the hash value corresponding to the real-time damage score, the method further includes:

[0193] Zero-knowledge proofs are generated based on real-time tunnel damage data, tunnel operation data, real-time damage scores, and the hash values ​​corresponding to the real-time tunnel damage data, tunnel operation data, and real-time damage scores.

[0194] Accordingly, real-time tunnel damage data, tunnel operation data, and the hash values ​​corresponding to real-time damage scores are stored on the blockchain according to a pre-set smart contract, including:

[0195] The zero-knowledge proof, along with the hash values ​​corresponding to real-time tunnel damage data, tunnel operation data, and real-time damage scores, are stored in the blockchain according to a preset smart contract.

[0196] In this scheme, zero-knowledge proof can be an encryption protocol that allows one party (the prover) to prove the truth of a statement to another party (the verifier) ​​without revealing any specific information about that statement. Simply put, the prover can prove they know a secret (such as the truth of certain data) without needing to show that data itself.

[0197] The content to be proven can be determined. For example, proving that "real-time tunnel damage data is valid" without revealing the specific data. A suitable zero-knowledge proof scheme, such as zk-SNARKs or zk-STARKs, is chosen. These schemes each have their own characteristics, and the specific choice depends on security, efficiency, and implementation complexity. A proof is generated using real-time tunnel damage data, tunnel operation data, and real-time damage scores. This process typically involves complex mathematical calculations. The generated proof will contain some encrypted information but not the original data. The proof and its associated hash value are received by a verifier (possibly a blockchain node or other user). The validity of the proof is confirmed using a verification algorithm based on the zero-knowledge proof protocol, ensuring that the prover's statement is true. Then, it is determined how to handle the zero-knowledge proof in a smart contract. The smart contract needs to be able to receive the proof and hash value. The smart contract is deployed to the blockchain, enabling it to accept the zero-knowledge proof and hash value. A blockchain client connects to the network. The smart contract's storage function is called, passing the zero-knowledge proof and its corresponding hash value as parameters. After submitting the request, the transaction is awaited to be confirmed by the blockchain network. Once confirmed, the zero-knowledge proof and hash value are stored on the blockchain, ensuring data security and immutability.

[0198] In this scheme, zero-knowledge proofs allow for the verification of data authenticity without exposing the data content, thus protecting privacy. By storing the proofs and hash values ​​on the blockchain, the immutability and traceability of the data can be ensured.

[0199] Example 4

[0200] Figure 4 This is a schematic diagram of the structure of a tunnel damage monitoring system provided in Embodiment 4 of this application, as shown below. Figure 4 As shown, this system is used to implement the tunnel damage monitoring system method provided in Embodiments 1, 2, and 3. The system specifically includes the following:

[0201] The data acquisition module 401 is used to acquire tunnel operation data and tunnel damage data at preset acquisition time points during the initial stage of tunnel operation, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the handheld terminal of the staff.

[0202] The scoring dataset creation module 402 is used to receive the damage score of the tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage score, and create a scoring dataset according to the tunnel operation data and the damage score tags.

[0203] The data segmentation module 403 is used to obtain a preset damage score threshold and to segment the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data according to the damage score and the damage score threshold.

[0204] The damage dataset creation module 404 is used to determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data based on the tunnel damage data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags.

[0205] The model building module 405 is used to build a tunnel damage monitoring model and train the damage monitoring model according to the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard.

[0206] The real-time data determination module 406 is used to continuously collect real-time tunnel operation data, input the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, and if the real-time damage score exceeds a preset damage score threshold, determine the real-time tunnel damage data through the tunnel damage monitoring model.

[0207] In this embodiment, the data acquisition module is used to acquire tunnel operation data and tunnel damage data at preset acquisition time points during the initial stage of tunnel operation, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by workers' handheld terminals; the scoring dataset creation module is used to receive damage scores of the tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage scores, and create a scoring dataset according to the tunnel operation data and damage score tags; the data segmentation module is used to obtain a preset damage score threshold, and segment the data according to the damage scores and damage score threshold. The system categorizes tunnel operation data into normal and abnormal tunnel operation data. A damage dataset creation module identifies abnormal tunnel damage data corresponding to the abnormal tunnel operation data, labels the abnormal tunnel damage data with tunnel damage tags, and creates a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags. A model building module constructs a tunnel damage monitoring model, trains the model based on the scoring dataset and the damage dataset until the model reaches a preset training standard. A real-time data determination module continuously collects real-time tunnel operation data, inputs it into the tunnel damage monitoring model, and obtains a real-time damage score. If the real-time damage score exceeds a preset damage score threshold, the real-time tunnel damage data is determined by the tunnel damage monitoring model. Through this tunnel damage monitoring system, potential damage or anomalies can be detected promptly by continuously collecting and analyzing tunnel operation data, ensuring tunnel safety. The model automates data processing, improving data processing efficiency. By analyzing historical and real-time data, the model captures patterns and trends in real-world situations, ensuring results match actual conditions and improving the reliability of the analysis results.

[0208] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A method for monitoring tunnel damage, characterized in that, The method includes: In the initial stage of tunnel operation, tunnel operation data and tunnel damage data are acquired at each preset collection time point, and the tunnel operation data and tunnel damage data are transmitted to the control center. Among them, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the staff's handheld terminal. The system receives damage scores for tunnel operation data sent by the control center, labels the tunnel operation data with damage score tags based on the damage scores, and creates a score dataset based on the tunnel operation data and the damage score tags. Obtain a preset damage score threshold, and divide the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data based on the damage score and the damage score threshold; Based on the tunnel damage data, determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags. Construct a tunnel damage monitoring model, and train the tunnel damage monitoring model based on the scoring dataset and the damage dataset until the tunnel damage monitoring model reaches the preset tunnel damage monitoring model training standard; Real-time tunnel operation data is continuously collected and input into the tunnel damage monitoring model to obtain a real-time damage score. If the real-time damage score exceeds a preset damage score threshold, the real-time tunnel damage data is determined through the tunnel damage monitoring model.

2. The tunnel damage monitoring method according to claim 1, characterized in that, After continuously collecting real-time tunnel operation data and inputting the real-time tunnel operation data into the tunnel damage monitoring model to obtain a real-time damage score, the method further includes: If the real-time damage score is lower than the preset damage score threshold, the real-time tunnel operation data is input into the preset maintenance detection model to obtain the real-time maintenance score. If the real-time maintenance score exceeds the preset maintenance score threshold, the maintenance type is determined by the preset maintenance detection model and the maintenance type is transmitted to the control center.

3. The tunnel damage monitoring method according to claim 1, characterized in that, After determining real-time tunnel damage data using the tunnel damage monitoring model, the method further includes: Acquire tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data, and create a three-dimensional tunnel model based on the tunnel geometry data; The material properties of the tunnel 3D model are set according to the tunnel material data, and the boundary data of the tunnel 3D model is set according to the tunnel boundary data. The tunnel mesh is divided into three-dimensional models of the tunnel based on the tunnel geometry data and tunnel material data, and the tunnel load data is input into the three-dimensional models of the tunnel.

4. The tunnel damage monitoring method according to claim 3, characterized in that, After dividing the tunnel into a tunnel mesh based on the tunnel geometry data and tunnel material data, and inputting tunnel load data into the tunnel 3D model, the method further includes: Obtain historical monitoring results, and determine historical tunnel damage data, causes of historical damage, and impacts of historical damage based on the historical monitoring results; Obtain simulation requirements, set simulation parameters for the tunnel 3D model according to the simulation requirements, input the historical tunnel damage data into the tunnel 3D model for simulation testing, and obtain tunnel simulation results; Based on the tunnel simulation results, historical damage causes, and historical damage impacts, determine whether the tunnel 3D model meets the preset simulation standards. If the preset simulation standards are met, the tunnel 3D model is deemed to have passed the simulation test.

5. The tunnel damage monitoring method according to claim 4, characterized in that, After determining whether the tunnel 3D model meets the preset simulation standards based on the tunnel simulation results, historical damage causes, and historical damage impacts, the method further includes: If the preset simulation standard is not met, the adjustment range of material properties, boundary data, and tunnel mesh is determined based on tunnel geometry data, tunnel material data, tunnel boundary data, and tunnel load data. Within the adjustment ranges of material properties, boundary data, and tunnel mesh, the preset adjustment step sizes for material properties, boundary data, and tunnel mesh are used to continuously adjust the material properties, boundary data, and tunnel mesh. After each adjustment, a simulation test is performed again until the tunnel 3D model passes the simulation test.

6. The tunnel damage monitoring method according to claim 4 or 5, characterized in that, After the 3D tunnel model passes simulation testing, the method further includes: Real-time tunnel damage data is input into the tunnel's 3D model to obtain the real-time damage causes and effects, and then the real-time damage causes and effects are sent to the control center.

7. The tunnel damage monitoring method according to claim 2, characterized in that, The training process for the pre-defined maintenance detection model includes: Obtain historical maintenance rating records, and determine historical tunnel operation data and historical maintenance ratings based on the historical maintenance rating records; The maintenance rating labels are labeled with the maintenance ratings of the historical tunnel operation data according to the historical maintenance ratings, and the maintenance rating dataset is determined according to the historical tunnel operation data and the maintenance rating labels. Based on the historical tunnel operation data and the preset maintenance scoring threshold, abnormal historical tunnel operation data is determined, and the historical maintenance type of the abnormal historical tunnel operation data is obtained. A maintenance type dataset is created by labeling the abnormal historical tunnel operation data with the maintenance type labels based on the historical maintenance type; Construct a maintenance detection model, and train the maintenance detection model based on the maintenance score dataset and the maintenance type dataset until the maintenance detection model reaches the preset maintenance detection model training standard.

8. The tunnel damage monitoring method according to claim 1, characterized in that, After determining real-time tunnel damage data using the tunnel damage monitoring model, the method further includes: The real-time tunnel damage data, real-time tunnel operation data, and real-time damage score are hashed respectively to obtain the hash values ​​corresponding to the real-time tunnel damage data, real-time tunnel operation data, and real-time damage score. Real-time tunnel damage data, real-time tunnel operation data, and the hash values ​​corresponding to real-time damage scores are stored on the blockchain according to a pre-set smart contract.

9. The tunnel damage monitoring method according to claim 8, characterized in that, After obtaining real-time tunnel damage data, real-time tunnel operation data, and the hash value corresponding to the real-time damage score, the method further includes: Zero-knowledge proofs are generated based on real-time tunnel damage data, real-time tunnel operation data, real-time damage scores, and the hash values ​​corresponding to the real-time tunnel damage data, real-time tunnel operation data, and real-time damage scores. Accordingly, real-time tunnel damage data, real-time tunnel operation data, and the hash values ​​corresponding to real-time damage scores are stored on the blockchain according to a pre-set smart contract, including: The zero-knowledge proof, real-time tunnel damage data, real-time tunnel operation data, and hash values ​​corresponding to real-time damage scores are stored in the blockchain according to a preset smart contract.

10. A tunnel damage monitoring system, characterized in that, The system includes: The data acquisition module is used to acquire tunnel operation data and tunnel damage data at preset acquisition time points during the initial stage of tunnel operation, and transmit the tunnel operation data and tunnel damage data to the control center; wherein, the tunnel operation data includes tunnel stress data and tunnel strain data transmitted by piezoelectric ceramic sensors, tunnel temperature data transmitted by fiber optic sensors, and tunnel vibration data transmitted by accelerometers; the tunnel damage data includes tunnel displacement data, tunnel deformation data, and tunnel crack data transmitted by the workers' handheld terminals. The scoring dataset creation module is used to receive damage scores of tunnel operation data sent by the control center, label the tunnel operation data with damage score tags according to the damage scores, and create a scoring dataset based on the tunnel operation data and the damage score tags. The data segmentation module is used to obtain a preset damage score threshold and to segment the tunnel operation data into normal tunnel operation data and abnormal tunnel operation data based on the damage score and the damage score threshold. The damage dataset creation module is used to determine the abnormal tunnel damage data corresponding to the abnormal tunnel operation data based on the tunnel damage data, label the abnormal tunnel operation data with tunnel damage tags based on the abnormal tunnel damage data, and create a damage dataset based on the abnormal tunnel operation data and the tunnel damage tags. The model building module is used to build a tunnel damage monitoring model and train the damage monitoring model based on the scoring dataset and the damage dataset until the damage monitoring model reaches the preset damage monitoring model training standard. The real-time data determination module is used to continuously collect real-time tunnel operation data, input the real-time tunnel operation data into the tunnel damage monitoring model, obtain a real-time damage score, and if the real-time damage score exceeds a preset damage score threshold, determine the real-time tunnel damage data through the tunnel damage monitoring model.

Images