A highway tunnel deformation monitoring and forecasting device and system

Abstract

The application discloses a highway tunnel deformation monitoring and forecasting device and system, and belongs to the field of road construction.The system comprises a data acquisition module, a data preprocessing module, a data storage module, a model training and evaluation module, a real-time forecasting module, an early warning mechanism module, a decision support and maintenance module, a feedback and optimization module, and the model training and evaluation module.The device comprises an arched frame and various detection devices.Compared with the traditional monitoring mode, different sensors provide real-time and accurate data basis, and a deep learning model can mine deeper structural health information, which not only improves the monitoring accuracy, but also enhances the intelligent decision-making ability, thereby providing more comprehensive protection for the safe operation of the tunnel.

Description

Technical Field

[0001] This invention belongs to the field of road construction, and more specifically, relates to a device and system for monitoring and predicting deformation of highway tunnels. Background Technology

[0002] Highway tunnels are dedicated passageways for automobile transportation. With the development of society, economy, and production, a large number of expressways have emerged, placing higher standards on road construction technology, requiring straight routes, gentle gradients, and wide road surfaces. Therefore, when roads traverse mountainous areas, the old method of winding mountain roads has often been replaced by tunnels. The construction of tunnels plays a vital role in improving the technical condition of highways, shortening travel distances, increasing transport capacity, and reducing accidents.

[0003] The risks caused by deformation of highway tunnels mainly include the following four aspects:

[0004] Firstly, structural safety hazards: Deformation may cause cracks in the lining and spalling of concrete, which in severe cases may lead to partial or overall collapse, weakening the stability of the tunnel structure and affecting its long-term load-bearing capacity. Deformation cracks may also cause groundwater to seep in, further corroding the steel bars and concrete.

[0005] Secondly, there is the risk of traffic accidents: deformation may reduce the tunnel clearance, causing oversized vehicles to hit the top or side walls. The settlement of the arch or the bulging of the bottom may cause road surface undulation, leading to loss of vehicle control. At the same time, ventilation, lighting, fire protection and other auxiliary facilities may fail due to deformation, affecting emergency functions.

[0006] Third, traffic and economic impact: Severe deformation requires closure for repair, leading to traffic congestion and increased detour costs. The cost of repairing deep structural damage may far exceed the investment in routine maintenance.

[0007] Fourth, secondary disasters: Tunnel deformation may reflect loosening of the surrounding rock, which may induce landslides or rockfalls. It may also cause blockage of drainage ditches, potentially leading to flooding inside the tunnel.

[0008] Meanwhile, my country has also formulated a number of national standards and industry specifications for the design, construction and operation and maintenance of highway tunnels, which clearly define the requirements for deformation control. Through national standard systems such as JTGD70 and JTGH12, deformation is controlled throughout the entire life cycle from design to operation and maintenance.

[0009] Chinese patent CN109556529B discloses a tunnel deformation monitoring and analysis method based on grid projection point cloud processing technology. For a tunnel surface point cloud model obtained by a 3D laser scanner, the method constructs a tunnel design surface model and performs unit grid processing, projects the unit grid onto the tunnel surface point cloud model, fits the point cloud within the projection range into a planar quadrilateral, calculates the distance from the centroid of the planar polygon to the projection center on the tunnel central axis, judges the difference in deformation direction of the point cloud in the projection area corresponding to the unit grid, filters deformation monitoring data, and colors the tunnel surface design model according to the magnitude of the deformation at the corresponding position of the unit grid, thereby realizing the overall analysis and visualization of tunnel deformation monitoring.

[0010] The above technical solution has the following drawbacks:

[0011] The method used for monitoring and analyzing tunnel deformation is based on an ideal state. However, in actual situations, it is still necessary to analyze whether deformation has occurred based on various factors such as the current construction environment and the current climate conditions.

[0012] Currently, there are also methods such as monitoring by installing distributed optical fibers in the tunnel or using Transformer models for prediction. However, traditional distributed optical fiber monitoring has significant limitations. During subsequent dismantling, maintenance, and replacement, it may cause secondary crack damage to tunnels with a long service life. Furthermore, Transformer models usually require a large amount of labeled data for training, resulting in high costs for data collection and labeling. Summary of the Invention

[0013] To address the above deficiencies, this invention provides a highway tunnel deformation monitoring and prediction system, comprising a data acquisition module, a data preprocessing module, a data storage module, a model training and evaluation module, a real-time prediction module, an early warning mechanism module, a decision support and maintenance module, a feedback and optimization module, and a model training and evaluation module.

[0014] The data acquisition module is used to acquire data from fiber optic sensors and external sensors. The data preprocessing module cleans and denoises the data acquired by the data acquisition module and then sends it to the data storage module for integration and storage. The model training and evaluation module trains the model using the labeled dataset obtained from the data storage module. The real-time prediction module receives real-time data and uses the trained model to predict the structural health status. The early warning mechanism module provides intelligent early warnings based on the real-time prediction results and set thresholds. The decision support and maintenance module provides maintenance decision support based on the early warning information and prediction results. The feedback and optimization module analyzes the model performance based on the comparison information between the collected actual monitoring data and the prediction results, and continues to train to improve the prediction accuracy.

[0015] Furthermore, the data acquisition module includes a weather data acquisition unit for acquiring wind speed, humidity, temperature and precipitation data, and a structural health monitoring data acquisition unit for acquiring strain, displacement and vibration data.

[0016] Furthermore, after receiving real-time data, the real-time prediction module divides the data information into three preset condition judgment units:

[0017] When the wind speed is greater than 20 m / s, the strong wind weather unit calculates the average deformation based on the weather data and structural monitoring data collected by the data acquisition module, and outputs the calculated average deformation to the early warning mechanism module.

[0018] When the relative humidity is less than 30% and the temperature is greater than 20℃, the dry weather unit calculates the crack width from the weather data and structural monitoring data collected by the data acquisition module, and outputs the calculated crack width to the early warning mechanism module.

[0019] When precipitation is greater than 5 mm / h and temperature is less than 0℃, the average deformation is calculated by using the rain and snow weather unit to collect weather data and structural monitoring data from the data acquisition module, and the calculated crack width is output to the early warning mechanism module.

[0020] The present invention also discloses an apparatus for the above-mentioned highway tunnel deformation monitoring and prediction system, including an optical fiber sensor, a laser scanner, an ultrasonic sensor and an infrared thermal imager for monitoring the tunnel itself in conjunction with a data acquisition module, and an environmental detection sensor for detecting the current environment.

[0021] The fiber optic sensor includes an embedded fiber optic sensor and an external light sensor. The embedded fiber optic sensor is embedded in the inner wall of the tunnel, while the external light sensor, laser scanner, ultrasonic sensor, infrared thermal imager, and environmental detection sensor are all installed at the arched truss.

[0022] The width and height of the arched truss are adapted to the inner diameter of the tunnel. Its bottom is fixed to the top of the vehicle by a fixing plate, and it can be moved from one end of the tunnel to the other to complete the data collection work.

[0023] The truss includes an arched beam at the top, which is used to adapt to the curved surface of the tunnel;

[0024] Two support columns are connected to the bottom of the arch beam, and the support columns are used to provide vertical support for the arch beam;

[0025] The bottoms of the two support columns are fixed to the fixed base, which is used to fix them to the top of the vehicle;

[0026] A crossbeam is installed between the two ends of the arch beam, and several diagonal web members are connected between the arch beam and the crossbeam. The diagonal web members are used to increase stability.

[0027] Furthermore, the system includes fiber optic sensors, laser scanners, ultrasonic sensors, and infrared thermal imagers that work in conjunction with the data acquisition module to monitor the tunnel itself, and environmental detection sensors used to detect the current environment.

[0028] The fiber optic sensor includes an embedded fiber optic sensor and an external light sensor. The embedded fiber optic sensor is embedded in the inner wall of the tunnel, while the external light sensor, laser scanner, ultrasonic sensor, infrared thermal imager, and environmental detection sensor are all installed at the arched truss.

[0029] The width and height of the arched truss are adapted to the inner diameter of the tunnel. Its bottom is fixed to the top of the vehicle via a mounting plate, allowing it to move from one end of the tunnel to the other to complete data collection.

[0030] The truss includes an arched beam at the top, which is used to adapt to the curved surface of the tunnel;

[0031] Two support columns are connected to the bottom of the arch beam, and the support columns are used to provide vertical support for the arch beam;

[0032] The bottoms of the two support columns are fixed to the fixed base, which is used to fix them to the top of the vehicle;

[0033] A crossbeam is installed between the two ends of the arch beam, and several diagonal web members are connected between the arch beam and the crossbeam. The diagonal web members are used to increase stability.

[0034] Compared with the prior art, the present invention has the following advantages:

[0035] 1. By combining structural and environmental data collected by fiber optic sensors and various other sensors, a multidimensional dataset is formed. At the same time, data cleaning and feature extraction techniques are used to transform the raw data into a format suitable for model input.

[0036] 2. First, use traditional sensor technology for preliminary real-time monitoring to identify obvious structural problems. Then, use historical monitoring data to train Transformer and composite neural network models to extract potential hidden features and patterns. At the same time, based on real-time monitoring, use the trained deep learning model for dynamic prediction to identify potential risks.

[0037] 3. Compared with traditional monitoring methods, different sensors provide real-time and accurate data foundation, while deep learning models can extract deeper structural health information from it, which not only improves monitoring accuracy but also enhances intelligent decision-making capabilities, providing more comprehensive protection for the safe operation of tunnels. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the arched frame in Embodiment 1 of the present invention.

[0039] Figure 2 This is a schematic diagram of the arched frame in Embodiment 2 of the present invention.

[0040] In the diagram: 1. Arched beam; 2. Support column; 3. Fixed base; 31. Base I; 32. Base II; 33. Positioning seat; 34. Limiting column; 35. T-shaped seat; 36. Telescopic rod; 4. Crossbeam; 5. Diagonal web member. Detailed Implementation

[0041] To facilitate understanding of the present invention, the apparatus of the present invention will now be described more fully with reference to the accompanying drawings. Embodiments of the apparatus are shown in the drawings. However, the apparatus can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to make the disclosure of the present invention more thorough and complete.

[0042] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "configuration" should be interpreted broadly. For example, they can refer to a fixed connection or configuration, a detachable connection or configuration, or an integral connection or configuration. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0043] Example 1

[0044] This embodiment provides a highway tunnel deformation monitoring and prediction system for detecting a tunnel on a highway section in central my country. The tunnel is 2228 meters long, 8 meters wide, and 11.20 meters high.

[0045] This highway tunnel deformation monitoring and prediction system includes a data acquisition module, a data preprocessing module, a data storage module, a model training and evaluation module, a real-time prediction module, an early warning mechanism module, a decision support and maintenance module, a feedback and optimization module, and a model training and evaluation module.

[0046] Specifically, the functions and interaction methods of each module are as follows:

[0047] First, the data acquisition module collects data from fiber optic sensors and wireless sensors (wireless sensors are sensors with wireless transmission capabilities) to monitor parameters such as strain, temperature, humidity, and vibration in the tunnel in real time. The collected data is then sent to the data processing module through data transmission interfaces (such as LoRa, Zigbee, etc.).

[0048] In detail, the data acquisition module includes a weather data acquisition unit for collecting wind speed (m / s), humidity (%), and precipitation (mm) data, and a structural health monitoring data acquisition unit for collecting strain, displacement, and vibration data.

[0049] Subsequently, the data processing module cleans and denoises the received data, handles missing values, and then extracts features (such as strain rate, vibration energy, etc.) to form a dataset for use by subsequent models.

[0050] The data storage module is responsible for storing the integrated historical and real-time data and providing data query and access interfaces. The historical data is labeled by date for subsequent analysis modules to call. It should be noted that SQL or NoSQL databases can be selected.

[0051] Subsequently, the model training and evaluation module uses the labeled dataset (historical monitoring data and status) obtained from the data storage module to train the prediction model using a combination of Transformer and composite neural networks to evaluate model performance and accuracy. By combining fiber optic sensing technology and wireless sensor networks with Transformer models and composite neural networks, a comprehensive big data monitoring and prediction platform is formed, which can effectively improve the overall effect of tunnel detection and prediction. Traditional sensors provide a real-time and accurate data foundation, while deep learning models can extract deeper structural health information from it. Therefore, this combination scheme has significant advantages in innovation, effectiveness, and real-time performance. This integration not only improves monitoring accuracy but also enhances intelligent decision-making capabilities, providing more comprehensive protection for the safe operation of tunnels.

[0052] The real-time prediction module receives real-time data and uses a pre-trained model to predict the structural health status. The real-time data comes from the data acquisition module and pre-processed data. The prediction results are then passed to the early warning mechanism module for subsequent decision support.

[0053] The real-time forecasting module, after receiving real-time data, divides the current weather conditions into three preset conditions (which can be set as condition judgment units):

[0054] When the wind speed is greater than 20 m / s, the average deformation is calculated (using the wind load calculation model) based on the weather data and structural monitoring data collected by the data acquisition module using a strong wind weather unit, as follows:

[0055]

[0056] The average deformation is as follows:

[0057]

[0058] Where N is the number of detected points;

[0059] The calculated average deformation is then output to the early warning mechanism module.

[0060] When the relative humidity is <30% and the temperature is >20℃, the crack width is calculated using the dry weather unit based on the weather data and structural monitoring data collected by the data acquisition module (cracks are collected based on laser point cloud data, and cracks are identified by calculating the density change or rate of change of the points; the model uses soil moisture and settlement models), as follows:

[0061] Crack width = maximum crack length - minimum crack length;

[0062] The calculated crack width is then output to the early warning mechanism module.

[0063] When precipitation is greater than 5 mm / h and temperature is less than 0℃, the average deformation is calculated using a rain and snow weather unit based on the weather data and structural monitoring data collected by the data acquisition module (applying water flow and soil deformation models), as follows:

[0064] Calculate the displacement of each point, and combine it with the current humidity and temperature, focusing on the changes in the water-immersed area, and output the calculated crack width to the early warning mechanism module;

[0065] The early warning mechanism module provides intelligent early warnings based on real-time prediction results and set thresholds (and generates early warning information and sends notifications to relevant personnel, which can be achieved using existing technologies, such as downloading an app that works with the system on a mobile phone).

[0066] The threshold is specifically:

[0067] Calculation results for the corresponding strong wind weather unit:

[0068] If the average deformation exceeds 5mm, an early warning will be triggered. If the vibration frequency of the tunnel wall exceeds 1.0Hz, it indicates that the structure is affected and an early warning will also be triggered.

[0069] Calculation results for the corresponding dry weather unit:

[0070] If the crack width exceeds 2mm, an early warning will be triggered. If a local temperature exceeds 30℃ (by using image processing technology to identify abnormal temperature areas through infrared images), it indicates a risk of deformation under dry conditions and will also trigger an early warning.

[0071] Calculation results for the corresponding rain and snow weather units:

[0072] If the average deformation exceeds 5mm and is accompanied by abnormal temperature and humidity, an early warning will be triggered. If an area of ​​moisture accumulation is detected and its temperature is below 0℃, the tunnel may need to be temporarily closed (based on decisions made by staff).

[0073] The decision support and maintenance module provides maintenance decision support based on early warning information and prediction results, and suggests possible maintenance measures and inspection frequencies. At the same time, it is connected to the feedback loop of the data acquisition module to ensure that the monitoring strategy can be dynamically adjusted according to the actual situation.

[0074] The feedback and optimization module is used to collect comparison information between actual monitoring data and prediction results, analyze model performance, and regularly retrain and optimize the model to improve prediction accuracy. Finally, the optimized model is returned to the model training and evaluation module to form a cyclical improvement mechanism.

[0075] like Figure 1 As shown, in order to meet the hardware requirements of the data acquisition module, the present invention also includes the following device:

[0076] The data acquisition module includes fiber optic sensors, a laser scanner (scanning accuracy ±1mm, scanning range 30m, sampling frequency 1000Hz), ultrasonic sensors (detection to a depth of 0.2m, sampling frequency 200Hz), and an infrared thermal imager (resolution 640x480 pixels, detection range 5-150℃) for monitoring the tunnel itself, and environmental sensors (such as traditional meteorological sensors) for detecting the current environment.

[0077] The fiber optic sensors include embedded fiber optic sensors and external light sensors. The embedded fiber optic sensors are embedded in the inner wall of the tunnel, while the external light sensors, laser scanners, ultrasonic sensors, infrared thermal imagers, and environmental detection sensors are all installed at the arched truss.

[0078] The width and height of the arched truss are adapted to the inner diameter of the tunnel. Its bottom is fixed to the top of the vehicle via a fixed plate, and it can be moved from one end of the tunnel to the other to complete the data collection work.

[0079] The truss includes an arched beam at the top, which is used to adapt to the curved surface of the tunnel;

[0080] Two support columns 2 are connected to the bottom of the arch beam 1, and the support columns 2 are used to provide vertical support for the arch beam 1;

[0081] The bottoms of the two support columns 2 are fixed to the fixed base 3, which is used to fix the top of the vehicle;

[0082] A crossbeam 4 is installed between the two ends of the arch beam 1, and several diagonal web members 5 are connected between the arch beam 1 and the crossbeam 4. The diagonal web members 5 are used to increase stability.

[0083] During operation, the vehicle moves at a speed of 0.5 m / s (i.e., 1.8 km / h). This speed balances detection accuracy and efficiency. Under this premise, for laser scanning, 1,000 points are collected per second (including distance and height data for each point). For every 0.5 m advance, the laser scan can generate 1,000 data points, forming a point cloud of a cross-section. For ultrasonic sensors, 200 data points are collected per second to obtain information on the density and internal defects of the concrete. For infrared thermal imaging, each frame of the generated thermal image contains 640x480 = 307,200 pixels of information, enabling real-time monitoring of temperature distribution within the tunnel.

[0084] Example 2

[0085] The difference from Example 1 is that, as Figure 2 As shown, the arch beam 1 can be adjusted in real time for different tunnel sizes, and can be adapted to multiple tunnels of different sizes on the same highway, achieving the effect of one beam for multiple measurements, thus reducing the investment of additional funds.

[0086] Its specific structure is as follows:

[0087] The arch beam 1 is a segmented structure, and the segments are fixed together by hinges. The support column 2 is an electric hydraulic cylinder.

[0088] The fixed base 3 includes base I31, base II32, and positioning seat 33. Both base I31 and base II32 have slots along their bottom centerlines. Positioning seat 33 has screw holes at each of its four corners for fixing to the carrier. Telescopic rods 36 are fixed to both ends of positioning seat 33, with the top ends of the telescopic rods 36 engaging in their top slots. Both base I31 and base II32 are penetrated by two parallel limiting posts 34 to increase stability and ensure their horizontal position. The middle ends of the two limiting posts 34... The T-shaped base 35 is fixed to the middle of the positioning base 33 (the T-shaped base 35 is a T-shaped profile with two holes for the limiting post 34 to pass through). The base I 31 and base II 32 move closer or further apart by a driving device. In this embodiment, the driving device is an electric trolley, that is, two electric trolleys are respectively installed on the top of the two telescopic rods 36 and are engaged with the slide rail in the slot. Alternatively, the two telescopic rods 36 can be fixed by installing electric cylinders on both sides of the T-shaped base 35, which will not be described in detail here.

[0089] It should be noted that the structure described in this invention can be implemented in many different forms and is not limited to the embodiments described. Any equivalent transformations made by those skilled in the art based on the description and drawings of this invention, or direct or indirect applications in other related technical fields, such as the loading and unloading of other items, are included within the protection scope of this invention.

Claims

1. A highway tunnel deformation monitoring and prediction system, characterized in that: It includes a data acquisition module, a data preprocessing module, a data storage module, a model training and evaluation module, a real-time prediction module, an early warning mechanism module, a decision support and maintenance module, a feedback and optimization module, and a model training and evaluation module. The data acquisition module is used to acquire data from fiber optic sensors and external sensors, including a weather data acquisition unit for acquiring data on wind speed, humidity, temperature and precipitation, and a structural health detection data acquisition unit for acquiring data on strain, displacement and vibration. The data preprocessing module cleans and removes noise from the data collected by the data acquisition module before sending it to the data storage module for integration and storage. The model training and evaluation module trains from the labeled dataset obtained from the data storage module; The real-time prediction module is used to receive real-time data, including three condition judgment units set according to wind speed, humidity and precipitation, and uses a trained model to predict the structural health status. The three preset condition judgment units include: When the wind speed is greater than 20 m / s, the strong wind weather unit calculates the average deformation based on the weather data and structural monitoring data collected by the data acquisition module, and outputs the calculated average deformation to the early warning mechanism module. When the relative humidity is <30% and the temperature is >20°C, the dry weather unit calculates the crack width from the weather data and structural monitoring data collected by the data acquisition module, and outputs the calculated crack width to the early warning mechanism module. When precipitation is greater than 5 mm / h and temperature is less than 0°C, the average deformation is calculated by using the rain and snow weather unit to collect weather data and structural monitoring data from the data acquisition module, and the calculated average deformation is output to the early warning mechanism module. The early warning mechanism module provides intelligent early warning based on real-time prediction results and set thresholds; The decision support and maintenance module provides maintenance decision support based on early warning information and prediction results; The feedback and optimization module analyzes the model performance based on the comparison between the collected actual monitoring data and the prediction results, and continues to train to improve the prediction accuracy.

2. A highway tunnel deformation monitoring and prediction device, used in the highway tunnel deformation monitoring and prediction system described in claim 1, characterized in that: This includes fiber optic sensors, laser scanners, ultrasonic sensors, and infrared thermal imagers that work in conjunction with the data acquisition module to monitor the tunnel itself, as well as environmental monitoring sensors used to detect the current environment. The fiber optic sensor includes an embedded fiber optic sensor and an external light sensor. The embedded fiber optic sensor is embedded in the inner wall of the tunnel, while the external light sensor, laser scanner, ultrasonic sensor, infrared thermal imager, and environmental detection sensor are all installed at the arched truss. The width and height of the arched truss are adapted to the inner diameter of the tunnel. Its bottom is fixed to the top of the vehicle by a fixing plate, and it can be moved from one end of the tunnel to the other to complete the data collection work. The truss includes an arched beam at the top, which is used to adapt to the curved surface of the tunnel; Two support columns are connected to the bottom of the arch beam, and the support columns are used to provide vertical support for the arch beam; The bottoms of the two support columns are fixed to the fixed base, which is used to fix them to the top of the vehicle; A crossbeam is installed between the two ends of the arch beam, and several diagonal web members are connected between the arch beam and the crossbeam.

3. The highway tunnel deformation monitoring and prediction device as described in claim 2, characterized in that: The arched beam is a segmented structure, with each segment fixed by hinges, and the support column is an electric hydraulic cylinder. The fixed base includes base I, base II, and positioning seat. The center line of the bottom surface of base I and base II is provided with a slot. The four corners of the positioning seat are provided with screw holes for fixing to the carrier. The two ends of the positioning seat are fixed with telescopic rods. The top of the telescopic rods are inserted into the slots on the top of the telescopic rods. Base I and base II are both penetrated by two limiting posts. The middle ends of the two limiting posts are fixed to the middle of the positioning seat by T-shaped rods. Base I and base II move closer to each other or further away from each other through a driving device.

Images