A deformation monitoring method and system for high-altitude frozen soil roads

Abstract

The present invention relates to highway safety monitoring, and discloses a deformation monitoring method and system for high-altitude frozen soil highways, the method comprising: laying sensors and environmental sensors on the frozen soil highway to obtain sensor data and environmental data; using unmanned aerial vehicles equipped with LiDAR sensors at key points of the frozen soil highway to obtain road deformation data; uploading sensor data, road deformation data and environmental data through wireless communication and satellite communication, preprocessing the sensor data, road deformation data and environmental data, and then using the Kalman filter method to perform data fusion to obtain monitoring data; obtaining highway deformation results and highway deformation risk trends by using a machine learning model for the monitoring data; and judging whether to alarm according to the highway deformation results and highway deformation risk trends. The present invention can solve the problems of being unable to monitor in real time, poor accuracy and insufficient early warning capability in the current technology.

Description

Technical Field

[0001] The present invention relates to highway safety monitoring, and in particular to a deformation monitoring method and system for a high-altitude frozen soil highway. Background Art

[0002] With the impact of climate change, the permafrost in high-altitude areas around the world is undergoing significant changes. In particular, in high-altitude permafrost highway areas, the melting and frost heaving of permafrost are gradually intensifying, leading to unstable roadbeds, which in turn cause safety hazards such as road deformation, cracking, and even collapse. Due to the complex road traffic conditions and the difficulty of construction and maintenance in these areas, traditional highway deformation monitoring methods have been difficult to meet the accuracy requirements, especially in high-altitude, low-temperature, and permafrost environments. Traditional monitoring methods such as manual inspections, regular inspections, and visual observations have significant lags. Manual inspections often fail to detect potential deformation problems in a timely manner, especially under complex climatic conditions, and cannot fully monitor the deformation status of the entire highway. In addition, the results of manual monitoring are easily affected by factors such as weather conditions and manual experience, and there are large uncertainties and errors.

[0003] Currently, many frozen soil road monitoring systems rely mainly on a single type of sensor (such as strain gauges, displacement gauges, etc.), which can only monitor limited geographical areas or specific depth levels. Due to the complex structure of frozen soil roads, there are different deformation modes and dynamic changes. The data collection of a single sensor cannot fully reflect the overall deformation status of frozen soil roads. In addition, the failure of a single sensor may cause the failure of the entire monitoring system, affecting the reliability of the data.

[0004] Therefore, there is an urgent need for a method that can effectively integrate multi-source data and accurately assess highway deformation and its potential risks through data fusion and intelligent analysis. Summary of the invention

[0005] In view of this, the present invention proposes a deformation monitoring method and system for high-altitude frozen soil roads, aiming to solve the problems of the current technology such as the inability to monitor in real time, poor accuracy, and insufficient early warning capability.

[0006] The present invention provides a deformation monitoring method for a high-altitude frozen soil highway, comprising:

[0007] Distributed optical fiber sensors are laid out at equal intervals on the roadbed and pavement of the frozen soil highway, and multi-point displacement meters and soil temperature and humidity sensors are evenly buried deep in the frozen soil highway to obtain sensor data.

[0008] Drones equipped with LiDAR sensors were used at key points on frozen roads to obtain road deformation data;

[0009] Environmental sensors are deployed at equal intervals on frozen roads to obtain environmental data;

[0010] Uploading the sensor data, the road surface deformation data and the environmental data to the central data processing system through wireless communication and satellite communication, pre-processing the sensor data, the road surface deformation data and the environmental data in the central data processing system, obtaining the first sensor data, the first road surface deformation data and the first environmental data and storing them;

[0011] Using Kalman filtering method to perform data fusion on the first sensor data, the first road deformation data and the first environment data to obtain monitoring data;

[0012] The highway deformation results and highway deformation risk trends are obtained by applying machine learning models to monitoring data;

[0013] Determine whether to issue an alarm based on the highway deformation results and highway deformation risk trends.

[0014] Furthermore, the sensor data includes road surface strain data, road surface displacement data, roadbed strain data, roadbed displacement data, highway deep displacement data, and highway deep temperature and humidity data.

[0015] Furthermore, the environmental sensor includes a temperature and humidity sensor, an air pressure sensor and a wind speed sensor;

[0016] The environmental data includes temperature and humidity data, air pressure data and wind speed data.

[0017] Furthermore, the central data processing system pre-processes the sensor data, road deformation data and environmental data. Specifically, the central data processing system first continues to denoise the sensor data, road deformation data and environmental data to obtain denoised sensor data, denoised road deformation data and denoised environmental data, then performs data consistency processing, time synchronizes and spatially aligns the denoised sensor data, denoised road deformation data and denoised environmental data to obtain aligned sensor data, aligned road deformation data and aligned environmental data, and finally performs normalization processing to obtain first sensor data, first road deformation data and first environmental data.

[0018] Furthermore, the central data processing system first continues to denoise the sensor data, the road deformation data and the environmental data to obtain denoised sensor data, denoised road deformation data and denoised environmental data, which are expressed as:

[0019] ;

[0020] ;

[0021] ;

[0022] in, is the denoised sensor data, is the denoised road deformation data, is the denoised environmental data, is the sensor data, is the road deformation data, For environmental data, is the filter function.

[0023] Furthermore, the data consistency processing is performed again, and the denoised sensor data, the denoised road deformation data and the denoised environmental data are time synchronized and space aligned to obtain aligned sensor data, aligned road deformation data and aligned environmental data, which are expressed as:

[0024] ;

[0025] ;

[0026] ;

[0027] in, is the synchronized sensor data, is the synchronized road deformation data, is the synchronized environment data. is the interpolation function, is a unified time sequence point;

[0028] ;

[0029] ;

[0030] ;

[0031] in, is the spatial coordinate of the synchronized sensor data, is the spatial coordinate of the synchronized road deformation data, is the spatial coordinate of the synchronized environment data, is the spatial transformation parameter, is the coordinate transformation function, is the aligned sensor data, is the road deformation data after alignment, is the aligned environmental data.

[0032] Furthermore, the above method finally performs normalization processing to obtain the first sensor data, the first road surface deformation data and the first environment data, which are expressed as:

[0033] ;

[0034] ;

[0035] ;

[0036] in, is the first sensor data, is the first road deformation data, is the first environment data, is the minimum value of sensor data, is the maximum value of sensor data, is the minimum value of road deformation data, is the maximum value of the road deformation data, is the minimum value of environmental data, It is the maximum value of environmental data.

[0037] Furthermore, the highway deformation results and highway deformation risk trends are obtained by using a machine learning model on the monitoring data. Specifically, the support vector machine model and the long short-term memory network model are trained and verified respectively using historical deformation data to obtain a deformation result classification model and a deformation risk prediction model. The monitoring data is input into the deformation result classification model to obtain the highway deformation results, including normal state, slight deformation and severe deformation. The monitoring data is input into the deformation risk prediction model to obtain the highway deformation trend, including low risk, medium risk and high risk.

[0038] Furthermore, the determination of whether to issue an alarm is made based on the highway deformation result and the highway deformation risk trend. Specifically, if the highway deformation result is normal and the highway deformation trend is low risk, no alarm is issued.

[0039] If the road deformation result is normal, and the road deformation trend is medium risk or high risk, a first-level alarm will be issued;

[0040] If the road deformation result is slight deformation and the road deformation trend is low risk, a first-level alarm will be issued;

[0041] If the road deformation result is slight deformation, and the road deformation trend is medium risk or high risk, a second-level alarm will be issued;

[0042] If the road deformation result is severe deformation, and the road deformation trend is any of low risk, medium risk or high risk, a third-level alarm will be issued.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows: the deformation monitoring method of a high-altitude frozen soil highway of the present invention effectively covers all levels and key parts of the frozen soil highway by laying distributed optical fiber sensors, multi-point displacement meters and soil moisture temperature sensors at key positions such as the roadbed, road surface and deep layer of the frozen soil highway, and collects data in combination with the LiDAR sensor carried by the drone, thereby ensuring the comprehensiveness and high accuracy of the data. At the same time, the introduction of environmental sensors provides an important basis for the comprehensive evaluation of the external environmental impact of highway deformation; in terms of data processing, the Kalman filter method is used to denoise and fuse sensor data, road surface deformation data and environmental data, effectively reducing the noise impact of the sensor and ensuring the high quality of the monitoring data. Data consistency processing, time synchronization and spatial alignment technology ensure the uniformity and timeliness of data from different sources, providing a reliable basis for subsequent analysis; by introducing machine learning models, especially models such as support vector machines (SVM) and long short-term memory networks (LSTM), it is possible to accurately analyze monitoring data and predict future trends and potential risks of highway deformation. The machine learning model can not only classify deformation states according to historical data, but also predict the risk level in the future in real time, provide scientific decision-making basis for road management personnel, reduce manual intervention and prediction errors; determine whether to trigger the alarm mechanism according to the deformation analysis results and risk prediction values, so as to issue early warning information in time when the highway is deformed or at risk. The alarm level is dynamically adjusted according to the deformation results and risk prediction values, which can effectively distinguish deformations of different levels, reasonably allocate resources, and ensure the safe operation of the highway; it combines high-precision sensors, advanced data processing technology and intelligent analysis methods, has strong adaptability, can operate efficiently in complex environments such as high altitude, low temperature, and frozen soil, and ensures stable monitoring and accurate assessment of road deformation risks under extreme climatic conditions; it can monitor highway deformation in real time and warn of risks in time, avoiding the lag and limitations of traditional inspection methods, helping to quickly discover potential safety hazards, reduce the probability of road accidents, reduce the cost of highway maintenance and repair, and improve the safety and service life of high-altitude frozen soil highways.

[0044] On the other hand, the present invention also proposes a deformation monitoring system for a high-altitude frozen soil highway, the system comprising:

[0045] Sensor systems and central data processing systems;

[0046] The sensor system comprises:

[0047] a data acquisition module configured to acquire sensor data, road deformation data, and environmental data;

[0048] a data transmission module configured to transmit sensor data, road deformation data, and environmental data to a central data processing system;

[0049] The central data processing system comprises:

[0050] A preprocessing module is configured to preprocess the sensor data, the road surface deformation data and the environmental data to obtain first sensor data, first road surface deformation data and first environmental data;

[0051] A storage module configured to store first sensor data, first road deformation data, first environment data, and historical deformation data;

[0052] A data fusion module is configured to fuse the first sensor data, the first road deformation data, and the first environment data to obtain monitoring data;

[0053] An analysis module configured to use a machine learning model on the monitoring data to obtain a highway deformation result and a highway deformation risk trend;

[0054] The alarm module is configured to determine whether to issue an alarm based on the highway deformation result and the highway deformation risk trend.

[0055] It is understandable that the above-mentioned deformation monitoring system for high-altitude frozen soil roads has the same beneficial effects, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0057] Figure 1 This is a flow chart of a deformation monitoring method for a high-altitude frozen soil highway according to an embodiment of the present invention;

[0058] Figure 2 The present invention is a structural block diagram of a deformation monitoring system for a high-altitude frozen soil highway according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0060] See also Figure 1 As shown, an embodiment of the present invention provides a deformation monitoring method for a high-altitude frozen soil highway, comprising:

[0061] S1: Distributed optical fiber sensors are laid out at equal intervals on the roadbed and pavement of the frozen soil highway, and multi-point displacement meters and soil temperature and humidity sensors are evenly buried in the deep layer of the frozen soil highway to obtain sensor data;

[0062] S2: Use drones equipped with LiDAR (laser radar) sensors at key points on frozen roads to obtain road deformation data;

[0063] S3: Deploy environmental sensors at equal intervals on the frozen soil road to obtain environmental data;

[0064] S4: uploading the sensor data, the road surface deformation data and the environmental data to the central data processing system through wireless communication and satellite communication, preprocessing the sensor data, the road surface deformation data and the environmental data in the central data processing system, obtaining the first sensor data, the first road surface deformation data and the first environmental data and storing them;

[0065] S5: using Kalman filtering to fuse the first sensor data, the first road deformation data and the first environment data to obtain monitoring data;

[0066] S6: Obtain highway deformation results and highway deformation risk trends by applying machine learning models to monitoring data;

[0067] S7: Determine whether to issue an alarm based on the highway deformation results and highway deformation risk trend.

[0068] Specifically, key points of frozen soil roads include complex mountainous areas and sections that are difficult to access.

[0069] Furthermore, the sensor data includes road surface strain data, road surface displacement data, roadbed strain data, roadbed displacement data, highway deep displacement data, and highway deep temperature and humidity data.

[0070] Specifically, the strain data of the road surface can monitor in real time the micro stress and deformation caused by freeze-thaw cycles, traffic loads or temperature changes. By obtaining strain data, it is possible to determine the stress condition of the road surface material, identify the formation process of potential cracks or cracks, detect local damage or destruction of the road surface in advance, and prevent serious damage from occurring;

[0071] The displacement data of the road surface can help identify settlement, warping or deformation caused by frozen soil activity, ice thawing and other reasons. Real-time monitoring of displacement can help capture uneven settlement or uplift, and take timely measures to avoid traffic safety risks caused by serious road deformation.

[0072] The strain data of the roadbed directly affects the stability and durability of the highway. The strain sensor can detect the stress changes caused by moisture changes in the roadbed during freezing and thawing, as well as possible soil compression or expansion problems, thereby revealing the impact of frozen soil on the highway structure and providing decision support for roadbed reinforcement or reconstruction;

[0073] The displacement data of the roadbed can monitor the displacement of the roadbed in real time. These data help to determine the slippage, settlement or expansion in the frozen soil layer, and then evaluate the stability of the roadbed. If the roadbed has a large displacement, it may indicate that the frozen soil highway has serious structural problems during use. Timely discovery and repair of these problems can effectively extend the service life of the road;

[0074] Highway deep displacement data can monitor the deformation of the deep layer of highway structure, including displacement, settlement or expansion of soil layer. Road deformation in frozen soil areas is usually the result of deep freeze-thaw cycles. Deep displacement data can help identify structural changes caused by frozen soil, thus providing a scientific basis for subsequent maintenance and reinforcement.

[0075] The deep temperature and humidity data of highways can obtain the temperature changes of each layer of the road in real time, especially the temperature fluctuations of the frozen soil layer. The temperature changes in the frozen soil area directly affect the freezing and thawing process of the road. The temperature data can reveal the periodic changes of the freeze-thaw cycle, which is of great significance for the prediction and repair of highway deformation. By monitoring the temperature data, preventive measures can be taken before the road surface or road base causes deformation due to temperature changes to reduce the damage caused by freeze-thaw to the road surface.

[0076] Furthermore, the environmental sensors include temperature and humidity sensors, air pressure sensors, and wind speed sensors;

[0077] Environmental data includes temperature and humidity data, air pressure data, and wind speed data.

[0078] Specifically, by monitoring temperature and humidity data, the impact of seasonal and weather changes on highway structures can be identified. For example, in cold seasons, a sudden drop in temperature can cause the soil to freeze, while a rise in temperature can cause melting, which can lead to highway subsidence or cracks. An increase in humidity can cause frozen soil to expand, while a decrease in humidity can cause the soil to shrink. Temperature and humidity data can help predict the impact of these natural environmental changes on the road surface and roadbed in advance, thereby providing early warnings for engineering personnel to carry out maintenance and repair work in advance.

[0079] Fluctuations in atmospheric pressure are often closely related to climate change. In particular, extreme weather such as snowstorms and heavy rains can cause drastic changes in air pressure, which may affect the freeze-thaw process of permafrost and the stability of highway structures. For example, changes in air pressure may be accompanied by strong winds or precipitation, causing the structure of the permafrost layer to become loose or unstable, which in turn affects the long-term service life of the road. By monitoring air pressure data, it can help predict upcoming extreme weather conditions and make road condition inspections and protective measures in advance;

[0080] Changes in wind speed may affect the soil surface state in permafrost areas, especially strong winds may cause greater soil drying effects or lead to accelerated freezing / thawing of the permafrost layer. In addition, strong winds may blow snow or dust, affecting the adhesion of the road surface, thereby accelerating the wear or freeze-thaw cycle of the road surface and roadbed. Monitoring wind speed data can help predict the potential impact of these meteorological factors on permafrost roads and provide a reference for whether wind and snow protection measures are needed.

[0081] Furthermore, to ensure the long-term operation of the equipment, sensors and data transmission equipment are equipped with solar panels or wind power generation systems to solve the problem of insufficient traditional power supply.

[0082] Furthermore, the central data processing system preprocesses the sensor data, road deformation data and environmental data. Specifically, the central data processing system first continues to denoise the sensor data, road deformation data and environmental data to obtain denoised sensor data, denoised road deformation data and denoised environmental data, then performs data consistency processing, time synchronizes and spatially aligns the denoised sensor data, denoised road deformation data and denoised environmental data to obtain aligned sensor data, aligned road deformation data and aligned environmental data, and finally performs normalization processing to obtain first sensor data, first road deformation data and first environmental data.

[0083] Specifically, first, the central data processing system uses algorithms (such as Kalman filtering, mean filtering, wavelet transform, etc.) to remove irrelevant or erroneous data from the original sensor data, road deformation data, and environmental data, retaining valid information, thereby obtaining denoised sensor data, denoised road deformation data, and denoised environmental data. The denoised data is more accurate and can provide a reliable basis for subsequent analysis to avoid misjudgment and calculation bias caused by noise. The denoised data also needs to be processed for consistency to ensure that sensor data, road deformation data, and environmental data from different sources can be correctly aligned in time and space. Finally, all data (sensor data, road deformation data, and environmental data) are normalized. Normalization refers to mapping data to a unified range (for example, between 0 and 1) so that all types of data are processed at the same scale. The dimensions and magnitudes of different types of monitoring data (such as temperature, humidity, displacement, strain, etc.) may vary greatly, and direct comparison or operation may lead to uneven effects. Through normalization, the dimensions of all data are unified, making subsequent machine learning, pattern recognition, or data analysis more stable and effective. Finally, the normalized data are the first sensor data, the first road deformation data, and the first environment data, ready for further analysis and risk prediction.

[0084] It should be noted that denoising can effectively eliminate erroneous and interfering data, ensuring that the remaining data reflects the actual road deformation state and environmental changes. Data consistency processing enables data collected from different sources and at different times to be effectively combined, thereby avoiding inaccurate results caused by time and space errors and improving data reliability. Normalization processing allows different types of data to be in the same dimensional range, avoiding analytical bias caused by inconsistent data scales, which is particularly important in multivariate data analysis and machine learning models.

[0085] Furthermore, the central data processing system first continues to denoise the sensor data, road deformation data and environmental data to obtain denoised sensor data, denoised road deformation data and denoised environmental data, which are expressed as:

[0086] ;

[0087] ;

[0088] ;

[0089] in, is the denoised sensor data, is the denoised road deformation data, is the denoised environmental data, is the sensor data, is the road deformation data, For environmental data, is the filter function.

[0090] Specifically, the filter function includes Gaussian filtering, wavelet transform or Kalman filtering. This embodiment uses Gaussian filtering for denoising.

[0091] It should be noted that Gaussian filtering can effectively reduce random noise caused by environmental changes, equipment failures or external factors (such as temperature fluctuations, electromagnetic interference, etc.). This is particularly important for frozen road deformation monitoring, because under extreme climatic conditions, sensor data is susceptible to irregular fluctuations. Through Gaussian filtering, more stable monitoring data can be obtained, thereby improving the reliability and accuracy of the data. When monitoring frozen road deformation, occasional abnormal sensor readings may mislead subsequent analysis. Gaussian filtering helps to remove these unnecessary fluctuations and ensure data stability. For complex geological changes such as foundation settlement and freeze-thaw effects that may occur in frozen road monitoring, noise suppression can help more clearly identify the true pattern of deformation and reduce the probability of misdiagnosis and missed diagnosis. Especially in the comprehensive analysis of multi-point monitoring data and environmental data, the denoised data can more effectively provide accurate deformation trends and risk warnings.

[0092] Furthermore, data consistency processing is performed to synchronize the denoised sensor data, denoised road deformation data and denoised environmental data in time and space, and the aligned sensor data, aligned road deformation data and aligned environmental data are obtained, which are expressed as:

[0093] ;

[0094] ;

[0095] ;

[0096] in, is the synchronized sensor data, is the synchronized road deformation data, is the synchronized environment data. is the interpolation function, is a unified time sequence point;

[0097] ;

[0098] ;

[0099] ;

[0100] in, is the spatial coordinate of the synchronized sensor data, is the spatial coordinate of the synchronized road deformation data, is the spatial coordinate of the synchronized environment data, is the spatial transformation parameter, is the coordinate transformation function, is the aligned sensor data, is the road deformation data after alignment, is the aligned environmental data.

[0101] Specifically, the interpolation function includes linear interpolation or spline interpolation. This embodiment uses linear interpolation to synchronize data. The coordinate transformation function includes a rotation matrix or rigid body transformation. This embodiment uses a rotation matrix.

[0102] It should be noted that in the process of time synchronization, linear interpolation can infer missing or irregular time data based on existing time points to ensure the time continuity of the data set. For example, some sensor records may be missing data points due to failures or transmission delays. Linear interpolation can interpolate at these missing time nodes to fill in the data gaps and ensure the continuity and integrity of the data. Linear interpolation can effectively smooth data transitions, so that the originally scattered and irregular sampling data can be connected continuously and consistently in time. It is suitable for processing most frozen soil road deformation monitoring data. Linear interpolation has a fast calculation speed and is suitable for real-time data processing.

[0103] In a multi-source sensor monitoring system, there may be certain coordinate deviations due to the different installation angles and positions of each sensor. The rotation matrix can effectively eliminate these deviations, ensure data consistency, and thus improve the accuracy of the analysis results.

[0104] Furthermore, a normalization process is finally performed to obtain the first sensor data, the first road deformation data and the first environment data, which are expressed as:

[0105] ;

[0106] ;

[0107] ;

[0108] in, is the first sensor data, is the first road surface deformation data, is the first environment data, is the minimum value of sensor data, is the maximum value of sensor data, is the minimum value of road deformation data, is the maximum value of the road deformation data, is the minimum value of environmental data, It is the maximum value of environmental data.

[0109] Furthermore, by using machine learning models on the monitoring data, the highway deformation results and highway deformation risk trends are obtained. The specific contents are as follows: first, the support vector machine model and the long short-term memory network model are trained and verified respectively using historical deformation data to obtain a deformation result classification model and a deformation risk prediction model. The monitoring data are input into the deformation result classification model to obtain the highway deformation results, including normal state, slight deformation and severe deformation. The monitoring data are input into the deformation risk prediction model to obtain the highway deformation trend, including low risk, medium risk and high risk.

[0110] Specifically, SVM (Support Vector Machine) is a common classification algorithm that can efficiently process high-dimensional data and perform well in multi-classification problems. Here, the SVM model can classify the current monitoring data and output specific deformation status results by training historical data. LSTM (Long Short-Term Memory Network) is a deep learning model commonly used for time series data prediction, which can capture the temporal relationship in the data. By training historical data with the LSTM model, the system can better predict the future deformation trend of the highway, analyze the time, location and change rate of deformation, and thus obtain the predicted value of deformation risk.

[0111] Furthermore, it is determined whether to issue an alarm based on the highway deformation result and the highway deformation risk trend. Specifically, if the highway deformation result is normal and the highway deformation trend is low risk, no alarm is issued.

[0112] Specifically, if the road deformation result is normal and the road deformation trend is low risk, it means that the road has not undergone significant deformation or damage, and from the perspective of environmental conditions, the possibility of future road deformation is small, so no immediate intervention is required. Not giving an alarm can avoid meaningless responses, reduce the waste of system resources, and ensure that staff can focus on dealing with emergencies that really require attention;

[0113] If the road deformation result is normal, and the road deformation trend is medium risk or high risk, a first-level alarm will be issued;

[0114] Specifically, if the highway deformation result is normal, or the highway deformation trend is medium risk or high risk, it means that although the current deformation state is normal, according to the prediction of environmental and historical data, the possibility of deformation in the future is high, or it is about to enter a dangerous state, triggering a first-level alarm, which can give early warning to ensure that relevant departments pay attention to and investigate potential deformation problems, and reduce the serious consequences caused by sudden deformation;

[0115] If the road deformation result is slight deformation and the road deformation trend is low risk, a first-level alarm will be issued;

[0116] Specifically, if the highway deformation result is slight deformation and the highway deformation trend is low risk, it means that although the highway has slight deformation, the deformation trend is relatively stable and the future deformation risk is low. The highway can continue to operate without affecting safety. The first-level alarm reminds management personnel to conduct routine inspections and repair measures to avoid the slight deformation from gradually worsening over time, thereby delaying the development of the problem.

[0117] If the road deformation result is slight deformation, and the road deformation trend is medium risk or high risk, a second-level alarm will be issued;

[0118] Specifically, if the highway deformation result is slight deformation and the highway deformation trend is medium risk or high risk, it means that the highway has already undergone slight deformation, and the risk forecast shows that its deformation trend will intensify in the short term and may develop into serious structural problems, triggering a secondary alarm and requiring immediate further preventive measures, such as repair or local reinforcement, to prevent the slight deformation from further deteriorating and causing greater damage;

[0119] If the road deformation result is severe deformation, and the road deformation trend is any of low risk, medium risk or high risk, a level 3 alarm will be issued;

[0120] Specifically, if the highway deformation result is severe deformation, and the highway deformation trend is any of low risk, medium risk or high risk, it means that the highway is already in a severe deformation state. Regardless of the risk prediction result, the deformation is severe enough to affect the highway's carrying capacity or traffic safety. The third-level alarm is triggered and emergency measures must be taken immediately, including traffic control, emergency repairs, and even partial road closures to ensure driving safety and personnel safety.

[0121] See also Figure 2 As shown, an embodiment of the present invention further provides a deformation monitoring system for a high-altitude frozen soil highway, the system comprising:

[0122] Sensor systems and central data processing systems;

[0123] The sensor system includes:

[0124] a data acquisition module configured to acquire sensor data, road deformation data, and environmental data;

[0125] a data transmission module configured to transmit sensor data, road deformation data, and environmental data to a central data processing system;

[0126] The central data processing system includes:

[0127] A preprocessing module is configured to preprocess the sensor data, the road surface deformation data and the environmental data to obtain first sensor data, first road surface deformation data and first environmental data;

[0128] A storage module configured to store first sensor data, first road deformation data, first environment data, and historical deformation data;

[0129] A data fusion module is configured to fuse the first sensor data, the first road deformation data, and the first environment data to obtain monitoring data;

[0130] An analysis module configured to use a machine learning model on the monitoring data to obtain a highway deformation result and a highway deformation risk trend;

[0131] The alarm module is configured to determine whether to issue an alarm based on the highway deformation result and the highway deformation risk trend.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A deformation monitoring method for high-altitude frozen soil roads, characterized in that: include: Distributed optical fiber sensors are laid out at equal intervals on the roadbed and pavement of the frozen soil highway, and multi-point displacement meters and soil temperature and humidity sensors are evenly buried deep in the frozen soil highway to obtain sensor data. Drones equipped with LiDAR sensors were used at key points on frozen roads to obtain road deformation data; Environmental sensors are deployed at equal intervals on frozen roads to obtain environmental data; Uploading the sensor data, the road surface deformation data and the environmental data to the central data processing system through wireless communication and satellite communication, pre-processing the sensor data, the road surface deformation data and the environmental data in the central data processing system, obtaining the first sensor data, the first road surface deformation data and the first environmental data and storing them; Using Kalman filtering method to perform data fusion on the first sensor data, the first road deformation data and the first environment data to obtain monitoring data; The highway deformation results and highway deformation risk trends are obtained by applying machine learning models to monitoring data; Determine whether to issue an alarm based on the highway deformation results and highway deformation risk trends; The central data processing system pre-processes the sensor data, the road surface deformation data and the environmental data, and the specific content is: the central data processing system first continues to denoise the sensor data, the road surface deformation data and the environmental data to obtain denoised sensor data, denoised road surface deformation data and denoised environmental data, then performs data consistency processing, performs time synchronization and spatial alignment on the denoised sensor data, denoised road surface deformation data and denoised environmental data to obtain aligned sensor data, aligned road surface deformation data and aligned environmental data, and finally performs normalization processing to obtain first sensor data, first road surface deformation data and first environmental data; The central data processing system first continues to denoise the sensor data, road deformation data and environmental data to obtain denoised sensor data, denoised road deformation data and denoised environmental data, which are expressed as: ； ； ； in, is the denoised sensor data, is the denoised road deformation data, is the denoised environmental data, is the sensor data, is the road deformation data, For environmental data, is the filter function; The data consistency processing is then performed to synchronize the denoised sensor data, the denoised road deformation data and the denoised environmental data in time and space to obtain aligned sensor data, aligned road deformation data and aligned environmental data, which are expressed as: ； ； ； in, is the synchronized sensor data, is the synchronized road deformation data, is the synchronized environment data. is the interpolation function, is a unified time sequence point; ； ； ； in, is the spatial coordinate of the synchronized sensor data, is the spatial coordinate of the synchronized road deformation data, is the spatial coordinate of the synchronized environment data, is the spatial transformation parameter, is the coordinate transformation function, is the aligned sensor data, is the road deformation data after alignment, is the aligned environmental data; The first sensor data, the first road deformation data and the first environment data are finally obtained by normalizing, which are expressed as: ； ； ； in, is the first sensor data, is the first road surface deformation data, is the first environment data, is the minimum value of sensor data, is the maximum value of sensor data, is the minimum value of road deformation data, is the maximum value of the road deformation data, is the minimum value of environmental data, is the maximum value of environmental data; The method of obtaining highway deformation results and highway deformation risk trends by using a machine learning model on monitoring data includes: firstly, using historical deformation data to train and verify a support vector machine model and a long short-term memory network model, respectively, to obtain a deformation result classification model and a deformation risk prediction model; inputting the monitoring data into the deformation result classification model to obtain highway deformation results, including normal state, slight deformation and severe deformation; and inputting the monitoring data into the deformation risk prediction model to obtain highway deformation trends, including low risk, medium risk and high risk; The determination of whether to issue an alarm is based on the highway deformation result and the highway deformation risk trend. Specifically, if the highway deformation result is normal and the highway deformation trend is low risk, no alarm is issued. If the road deformation result is normal, and the road deformation trend is medium risk or high risk, a first-level alarm will be issued; If the road deformation result is slight deformation and the road deformation trend is low risk, a first-level alarm will be issued; If the road deformation result is slight deformation, and the road deformation trend is medium risk or high risk, a second-level alarm will be issued; If the road deformation result is severe deformation, and the road deformation trend is any of low risk, medium risk or high risk, a third-level alarm will be issued.

2. The deformation monitoring method for high-altitude frozen soil roads according to claim 1 is characterized in that: The sensor data include road surface strain data, road surface displacement data, roadbed strain data, roadbed displacement data, highway deep displacement data, and highway deep temperature and humidity data.

3. A deformation monitoring method for high-altitude frozen soil roads according to claim 2, characterized in that: The environmental sensors include a temperature and humidity sensor, an air pressure sensor and a wind speed sensor; The environmental data includes temperature and humidity data, air pressure data and wind speed data.

4. A deformation monitoring system for a high-altitude frozen soil highway, implemented in a deformation monitoring method for a high-altitude frozen soil highway as described in any one of claims 1 to 3, characterized in that: The system comprises: Sensor systems and central data processing systems; The sensor system comprises: a data acquisition module configured to acquire sensor data, road deformation data, and environmental data; a data transmission module configured to transmit sensor data, road deformation data, and environmental data to a central data processing system; The central data processing system comprises: A preprocessing module is configured to preprocess the sensor data, the road surface deformation data and the environmental data to obtain first sensor data, first road surface deformation data and first environmental data; A storage module configured to store first sensor data, first road deformation data, first environment data, and historical deformation data; A data fusion module is configured to fuse the first sensor data, the first road deformation data, and the first environment data to obtain monitoring data; An analysis module configured to use a machine learning model on the monitoring data to obtain a highway deformation result and a highway deformation risk trend; The alarm module is configured to determine whether to issue an alarm based on the highway deformation result and the highway deformation risk trend.

Images