Method and system for predicting gradual change information of frozen soil foundation of highway in long-term operation

Abstract

The present invention discloses a method and system for predicting the gradual change information of frozen soil foundation of highway in long-term operation, and relates to the technical field of frozen soil treatment of highway. In step S1 of the method, the foundation meltwater condition coefficient Rsx is generated by integrating multi-source data, thereby realizing the assessment of the risk of foundation meltwater. Step S2 forms the frozen soil state coefficient ZTx through multi-dimensional collection of freezing and thawing data sets, and clarifies the seasonal frozen soil area and the non-seasonal frozen soil area. Step S3 corrects the frozen soil state coefficient by using the vegetation coverage rate ZbF and the terrain elevation value Gc obtained by remote sensing technology. Step S4 realizes the real-time monitoring and differential treatment of the foundation state by establishing a frozen soil feature library and a frozen soil foundation gradual change treatment plan library; Step S5 collects the annual average sunlight exposure time Gz of the corresponding frozen soil area that has been allocated; the frozen soil prediction and treatment model predicts and analyzes the matching gradual change treatment plan, and optimizes the treatment plan to improve the long-term operation and maintenance of the frozen soil foundation.

Description

Technical Field

[0001] The present invention relates to the technical field of highway frozen soil treatment, and in particular to a method and system for predicting gradual change information of frozen soil foundation of a highway during long-term operation. Background Art

[0002] With the continuous development and expansion of transportation networks, the highway system has become one of the key infrastructures connecting cities and rural areas. However, in cold climates, the long-term operation of frozen soil foundations of highways faces a series of challenges, one of which is the gradual changes in the foundation during operation. These changes may include fluctuations in groundwater levels, seasonal changes in temperature, and soil expansion and contraction caused by freeze-thaw cycles, all of which pose a potential threat to the stability and safety of frozen soil foundations of highways.

[0003] At present, the monitoring and management of frozen foundations of highways mainly rely on traditional inspections and periodic soil exploration. The disadvantage of this method is that it lacks the real-time and accuracy of the gradual change information of the foundation. Traditional methods are difficult to provide accurate predictions of the gradual change information of frozen foundations, which may lead to misjudgment of the foundation status, thus affecting the long-term operation and maintenance of frozen foundations.

[0004] In traditional methods, the detailed division of permafrost areas is often neglected, and there is a lack of comprehensive consideration of factors such as vegetation coverage and sunlight exposure time. This makes it easy for traditional methods to make overly macroscopic assessments of the entire area, and fail to deeply explore the foundation gradient information in different regions and seasons, resulting in the localization and individual characteristics of the gradient changes not being fully considered. In addition, existing technologies often lack differentiated methods in governance. Usually, once a problem with the foundation is detected, the conventional treatment method is often to adopt relatively uniform maintenance and improvement measures, while ignoring the differentiated needs in different regions and different gradient states. Such a treatment approach may lead to a waste of resources and unsatisfactory governance results. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the deficiencies of the prior art, the present invention provides a method and system for predicting the gradual change information of frozen soil foundation of highway in long-term operation, so as to solve the problems mentioned in the background technology.

[0007] (II) Technical solution

[0008] To achieve the above purpose, the present invention is implemented by the following technical scheme: a method for predicting the gradual change information of frozen soil foundation of highway in long-term operation, comprising the following steps:

[0009] S1. Pre-acquire a GPS electronic map of the highway frozen soil foundation area to be predicted, obtain the groundwater depth of the highway frozen soil foundation area in the form of random sampling, and obtain the annual average precipitation and snowfall in the frozen soil foundation area, and comprehensively generate a foundation meltwater condition coefficient Rsx. When the foundation meltwater condition coefficient Rsx is higher than a preset condition threshold, it indicates that there is a frozen soil meltwater risk, and the first processing information is sent to the system;

[0010] S2. When obtaining the first processing information, the frozen soil foundation area of ​​the highway is divided into multiple detection areas, detection points are set in the multiple detection areas, and the sensor group is assembled to detect the frozen soil state characteristics in two seasons, namely winter and summer, to establish a freezing data set and a melting data set, which are summarized as a frozen soil state characteristic data set, and a frozen soil state coefficient ZTx is generated according to the frozen soil state characteristic data set, and the frozen soil is determined to be a seasonal frozen soil area or a non-seasonal frozen soil area according to the frozen soil state coefficient ZTx;

[0011] S3, obtaining the vegetation coverage rate ZbF and the terrain elevation value Gc in the seasonal frozen soil area and the non-seasonal frozen soil area, correcting the frozen soil state coefficient ZTx, and using the correction result as the frozen soil melting coefficient Rh, re-dividing the detection area according to the value of the frozen soil melting coefficient Rh to obtain the division result;

[0012] S4. Based on the division results, establish a frozen soil feature database; store the geographical information of the frozen soil area, including topography, landform, soil depth, temperature distribution and deformation data; store the frozen soil state data of the frozen soil area, including soil temperature, humidity and water content, record soil type, frozen soil depth and seasonal change characteristics; obtain the gradual management plan of frozen soil foundation from public channels, summarize and obtain the information of frozen soil management plan, and establish a frozen soil foundation gradual management plan database;

[0013] Match the characteristics of the frozen soil area with the gradual change management schemes in the frozen soil foundation gradual change management scheme library, and output the corresponding frozen soil management scheme;

[0014] S5. Collect soil structure data and annual average sunlight exposure time Gz of the corresponding assigned permafrost areas, obtain the corresponding gradual governance plan, establish and train a convolutional neural network CNN, and generate a frozen land prediction and governance model. The frozen land prediction and governance model predicts and analyzes the matching gradual governance plan, obtains the prediction and analysis results, and selects the optimized governance plan based on the prediction and analysis results.

[0015] Preferably, the S1 specifically includes:

[0016] S11, obtaining a GPS electronic map of the highway frozen ground area to be predicted, where the GPS electronic map includes spatial resolution and detailed geographic information;

[0017] S12. In the frozen soil foundation area of ​​the highway, random sampling is performed to obtain groundwater depth data, and the groundwater depth Dxs is obtained by using groundwater level monitoring wells or water level sensor equipment according to the wellbore measurement method;

[0018] S13, obtaining the annual average precipitation JS and annual average snowfall JX in the frozen soil foundation area of ​​the highway from the meteorological station to evaluate the surface moisture state and the melting of frozen soil, and calculate the foundation meltwater condition coefficient Rsx;

[0019] S14. Compare the calculated foundation meltwater condition coefficient Rsx with the preset condition threshold. If the foundation meltwater condition coefficient Rsx is higher than the condition threshold, it indicates the risk of frozen soil meltwater. The first label is set in the current highway frozen soil foundation area, and the first processing information is sent to the system.

[0020] Preferably, the collected groundwater depth Dxs, annual average precipitation JS and annual average snowfall JX are used, after dimensionless processing, to generate the foundation meltwater condition coefficient Rsx according to the following formula:

[0021]

[0022] In the formula, W1>W2, W1 and W2 are the proportional coefficients of the annual average precipitation JS and the annual average snowfall JX, and W1+W2=1.0.

[0023] Preferably, S2 specifically includes:

[0024] S21. Use geographic information system (GIS) tools to divide the frozen soil foundation area of ​​the highway into multiple detection areas, and set detection points in each area;

[0025] S22. In winter, the soil temperature and humidity values ​​of the surface layer, frozen soil surface and active layer are collected through the sensor group, and the groundwater level depth is measured at the same time, forming a frozen data set;

[0026] S23. In summer, the soil temperature and moisture values ​​at the same location are collected by the sensor group, and the groundwater level depth is measured at the same time, forming a melting data set;

[0027] S24. The freezing data set and the melting data set are merged to form a comprehensive frozen soil state characteristic data set; including the soil temperature and moisture values ​​of the surface layer, frozen soil surface and active layer at different detection points in winter and summer, as well as the information of groundwater level depth.

[0028] Preferably, S2 also includes:

[0029] S26. For each detection point, calculate the average values ​​of soil temperature and moisture of the surface layer, frozen soil surface and active layer in winter and summer, as well as the average value of groundwater level depth;

[0030] S27, based on the average value calculated in S26, dimensionless processing, the frozen soil state coefficient ZTx is calculated by the following formula:

[0031]

[0032] In the formula, DdbcT1 represents the mean surface temperature in winter, XdbcS1 represents the mean surface humidity in summer; XdbcT2 represents the mean surface temperature in summer, DdbcS2 represents the mean surface humidity in winter; DdtcT1 represents the mean surface temperature of the frozen soil layer in winter, XdtcS1 represents the mean surface humidity of the frozen soil layer in summer; XdtcT2 represents the mean surface temperature of the frozen soil layer in summer, DdtcS2 represents the mean surface humidity of the frozen soil layer in winter; DhdcT1 represents the mean temperature of the active layer in winter, XhdcS1 represents the mean humidity of the active layer in summer; XhdcT2 represents the mean temperature of the active layer in summer, DhdcS2 represents the mean humidity of the active layer in winter; SdxsJz represents the mean depth of the groundwater level in summer, DdxsJz represents the mean depth of the groundwater level in winter;

[0033] S28. Preset a first threshold value, which serves as the dividing line between seasonal frozen soil and non-seasonal frozen soil. For each detection point, compare the calculated frozen soil state coefficient ZTx with the first threshold value. If the frozen soil state coefficient ZTx is greater than the first threshold value, it is determined to be a seasonal frozen soil area; if the frozen soil state coefficient ZTx is less than or equal to the first threshold value, it is determined to be a non-seasonal frozen soil area.

[0034] Preferably, S3 specifically includes:

[0035] S31. Use remote sensing technology to obtain image data of seasonal frozen soil and non-seasonal frozen soil areas, extract vegetation coverage area from image data, and obtain vegetation coverage rate ZbF; use radar interferometry InSAR technology to obtain digital elevation model DEM, use synthetic aperture radar SAR satellite data to generate radar elevation image, SAR technology penetrates clouds and vegetation to obtain surface elevation information, use stereo image matching technology to compare optical images of uniform objects at different viewing angles, calculate and obtain initial terrain elevation value Gc1, use GPS for ground measurement, obtain second terrain elevation value Gc2, correct the initial terrain elevation value Gc1, and obtain terrain elevation value Gc;

[0036] S32, the frozen soil state coefficient ZTx is corrected by the collected vegetation coverage rate ZbF and terrain elevation value Gc, and a correction model is established, and the frozen soil melting coefficient Rh is obtained by the following formula;

[0037] Rh=α*ZTx+β*ZbF+γ*Gc+∈

[0038] In the formula, α represents the first correction coefficient, β represents the second correction coefficient, represents the influence of vegetation coverage on the frozen soil state; γ represents the influence of terrain elevation on the frozen soil state, and ∈ represents the error term;

[0039] Preferably, S3 also includes:

[0040] S33. Set a stable threshold, a partial melting threshold and a full melting threshold, and compare the frozen soil melting coefficient Rh with the stable threshold, the partial melting threshold and the full melting threshold respectively. If it matches within the stable threshold interval, it is represented as a stable state area; if it matches within the partial melting threshold interval, it is represented as a partial melting state area; if it matches within the full melting threshold interval, it is represented as a full melting state area.

[0041] Preferably, the S4 specifically includes:

[0042] S41, storing the acquired geographic information data in a database; including topography, landform, soil depth, temperature distribution and deformation data; each data item is identified, organized and classified;

[0043] S42. Frozen soil status data is stored in the same database; including soil temperature, humidity and water content data; each data point can be associated with a specific geographic location and has a timestamp to reflect the temporal characteristics of the data;

[0044] S43, establish a frozen soil characteristic database, integrate geographic information and frozen soil status data into the frozen soil characteristic database; record soil type and frozen soil depth information in the frozen soil characteristic database; and for seasonal change characteristics, record the changes in soil temperature, humidity and water content in different seasons;

[0045] S44. Obtain information on gradual management plans for frozen ground from public channels, including monitoring and maintenance measures and environmental control plans;

[0046] S45. Summarize and organize the acquired gradual change control scheme information and establish a database of gradual change control schemes for frozen soil foundations, including detailed descriptions of the schemes, application conditions and effects;

[0047] S46. Using a database query language or a programming language, matching the features in the frozen soil feature library with the solutions in the frozen soil foundation gradual change management solution library based on soil type, frozen soil depth, and seasonal changes, and outputting the corresponding frozen soil management solution according to the matching results; including:

[0048] Stable state area: regular inspection and monitoring, and establishment of a maintenance protection layer on the foundation surface;

[0049] Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and strengthen drainage control;

[0050] In the area of ​​complete melting: improve the foundation, strengthen pile foundation, reinforce grouting and strengthen drainage control; and regulate the environment to control and reduce soil moisture.

[0051] Preferably, the S5 specifically includes:

[0052] S52, obtaining a treatment plan for the corresponding frozen soil area from an existing frozen soil foundation gradual change treatment plan library; the obtained plan is matched with the soil structure data and the annual average sunlight exposure time Gz;

[0053] S53, organizing the collected soil structure data and sunshine time data into a format suitable for the convolutional neural network CNN, using the existing soil structure data and annual average sunshine time Gz data, and the corresponding frozen land management plan as labels, to train the convolutional neural network CNN; using the trained frozen land prediction and management model to predict and analyze the new soil structure data and annual average sunshine time Gz data; obtaining the optimization prediction results of the frozen land prediction and management model for the frozen land management plan, including:

[0054] Stable state area: Regular inspection and monitoring, and building a maintenance protection layer on the foundation surface, using waterproof material protection layer, and increasing vegetation coverage when the annual average sunshine exposure time Gz is greater than the sunshine exposure threshold, increasing shading time in warm seasons, and increasing lighting in cold seasons to control the surface temperature;

[0055] Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and increase shading time in warm seasons and strengthen drainage control when the annual average sunshine duration Gz is greater than the sunshine threshold;

[0056] In the area of ​​full melting state: improve the foundation, add bored piles and vibrating piles, reinforce with grouting, increase the viscosity of the soil and strengthen drainage control; and when the annual average sunshine exposure time Gz is greater than the sunshine exposure threshold, increase the vegetation coverage, increase the shade time in the warm season, regulate the environment, add geothermal system and control and reduce soil moisture.

[0057] The system for predicting the gradual change of frozen soil foundation of highway in long-term operation includes data acquisition module, monitoring point setting and data set establishment module, frozen soil state classification module, vegetation terrain acquisition module, correction module, frozen soil feature library establishment module, matching module, soil structure and sunlight exposure time acquisition module, model training module and optimization management module;

[0058] The data acquisition module obtains the GPS electronic map of the highway frozen soil foundation area to be predicted; randomly samples the groundwater depth, and obtains the precipitation and snowfall to generate the foundation meltwater condition coefficient Rsx;

[0059] The monitoring point setting and data set establishment module is used to divide the highway frozen soil foundation area into multiple detection areas, and set detection points to collect soil temperature, humidity, and groundwater depth data in winter and summer through sensor groups to form frozen data sets and melted data sets;

[0060] The frozen soil state classification module calculates the frozen soil state coefficient ZTx based on the collected data and divides the seasonal frozen soil area into the non-seasonal frozen soil area;

[0061] Vegetation terrain acquisition module, using remote sensing technology to obtain vegetation coverage ZbF and terrain elevation Gc;

[0062] The correction module is used to correct the frozen soil state coefficient ZTx according to the vegetation coverage rate ZbF and the terrain elevation value Gc through the correction model to obtain the frozen soil thawing coefficient Rh;

[0063] The frozen soil characteristic database establishment module is used to store geographic information and frozen soil status data, including topography, landform, soil depth, temperature distribution and deformation data, as well as soil temperature, humidity and water content, and obtain gradual change management plan information from public channels to establish a frozen soil foundation gradual change management plan library;

[0064] The matching module is used to match the features in the frozen soil foundation feature library with the solutions in the frozen soil foundation gradual change treatment solution library according to the division results, and output the corresponding frozen soil treatment solution;

[0065] The soil structure and sunshine exposure time acquisition module is used to collect soil structure data and annual average sunshine exposure time Gz of the allocated area;

[0066] The model training module uses the convolutional neural network (CNN) for training to generate a frozen land prediction and management model;

[0067] The optimization management module uses the trained frozen land prediction and management model to perform predictive analysis on new soil structure and sunlight exposure time data to output an optimized frozen land management plan, including specific management measures for stable state areas, partially thawed state areas, and fully thawed state areas.

[0068] (III) Beneficial effects

[0069] The present invention provides a method and system for predicting the gradual change information of frozen soil foundation of highway in long-term operation. It has the following beneficial effects:

[0070] (1) The method for predicting the gradual change information of frozen soil foundation of highway in long-term operation, step S1 is used to integrate these data to generate the foundation meltwater condition coefficient Rsx, which is used to evaluate the meltwater risk of the foundation; step S2 determines whether the frozen soil is a seasonal frozen soil area or a non-seasonal frozen soil area based on the frozen soil state coefficient ZTx; step S3 obtains the vegetation coverage ZbF and terrain elevation value Gc in the seasonal frozen soil area and the non-seasonal frozen soil area. The frozen soil state coefficient ZTx is corrected using these data to form the frozen soil melting coefficient Rh. The detection area is re-divided based on the value of Rh to obtain the division result; step S4, based on the division result, a frozen soil feature library is established to store feature information such as geographic information, soil structure data, and temperature distribution. At the same time, a frozen soil foundation gradual change treatment plan library is constructed to summarize the public gradual change treatment plan information. It is used to match the characteristics of the frozen soil area with the gradual change treatment plan in the frozen soil foundation gradual change treatment plan library. Through this step, the corresponding frozen soil treatment plan is output. Step S5 collects the soil structure data of the corresponding frozen soil area that has been allocated and the annual average sunlight exposure time Gz. The corresponding gradual governance scheme is obtained, and a convolutional neural network (CNN) is established and trained to generate a frozen land prediction and governance model. The frozen land prediction and governance model performs prediction and analysis on the matching gradual governance scheme to obtain prediction and analysis results. Finally, the optimized governance scheme is selected based on the prediction and analysis results to improve the stability and safety of the frozen ground foundation.

[0071] (2) The method for predicting the gradual change information of highway frozen soil foundation in long-term operation, step S2 collects the soil temperature and humidity values ​​of the surface layer, frozen soil surface and active layer in winter and summer respectively, and measures the groundwater level depth at the same time, thus forming a frozen and thawed data set. This multi-dimensional data collection helps to fully understand the state of frozen soil foundation in different seasons and provides more detailed and comprehensive information. By merging the frozen data set and the thawed data set, the frozen soil state coefficient ZTx is calculated, and then divided according to the preset first threshold, the seasonal frozen soil area and the non-seasonal frozen soil area are clarified. It helps to eliminate seasonal and regional differences, making the model more universal and better adaptable to the foundation conditions under different conditions. This helps to better maintain and manage the frozen soil foundation of the highway and ensure its stability and safety in long-term operation.

[0072] (3) The method for predicting the gradual change information of the frozen soil foundation of highway in long-term operation, step S3 uses the vegetation coverage and terrain elevation values ​​obtained by remote sensing technology to correct the frozen soil state coefficient. This helps to more accurately reflect the impact of vegetation and terrain on the frozen soil state and improve the accuracy of the prediction. Combined with remote sensing technology and GPS measurement, multi-source data fusion is used to obtain terrain information. This method can make up for the shortcomings of a single data source and provide more comprehensive foundation characteristics. Using the corrected frozen soil thawing coefficient Rh, different thresholds are set for regional division, making the assessment of the frozen soil state more refined, which helps to propose targeted governance plans. Using data obtained by remote sensing technology, real-time monitoring and prediction of the foundation state can be achieved. This helps to timely discover potential problems, take corresponding governance and maintenance measures, and promote the safe and stable operation of frozen soil foundations of highways.

[0073] (4) This method of predicting the gradual change information of highway frozen soil foundation in long-term operation realizes differentiated management by proposing differentiated management plans for different frozen soil status areas, avoiding relatively uniform maintenance and improvement measures for the entire area, and improving the management effect. By establishing a database to store geographic information and frozen soil status data, real-time monitoring and timely maintenance of foundation conditions are achieved, which helps to discover potential problems in advance and take measures to ensure the safe operation of highway frozen soil foundations. By establishing a frozen soil foundation gradual change management plan library, sustainable maintenance plans for different situations are provided, providing a scientific basis for long-term operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 A schematic diagram of the method steps for predicting gradual change information of frozen soil foundation of highway in long-term operation according to the present invention;

[0075] Figure 2 The present invention is a schematic diagram of a flowchart for predicting gradual change information of frozen soil foundation of highway during long-term operation. DETAILED DESCRIPTION

[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0077] With the continuous development and expansion of transportation networks, the highway system has become one of the key infrastructures connecting cities and rural areas. However, in cold climates, the long-term operation of frozen soil foundations of highways faces a series of challenges, one of which is the gradual changes in the foundation during operation. These changes may include fluctuations in groundwater levels, seasonal changes in temperature, and soil expansion and contraction caused by freeze-thaw cycles, all of which pose a potential threat to the stability and safety of frozen soil foundations of highways.

[0078] At present, the monitoring and management of frozen foundations of highways mainly rely on traditional inspections and periodic soil exploration. The disadvantage of this method is that it lacks the real-time and accuracy of the gradual change information of the foundation. Traditional methods are difficult to provide accurate predictions of the gradual change information of frozen foundations, which may lead to misjudgment of the foundation status, thus affecting the long-term operation and maintenance of frozen foundations.

[0079] In traditional methods, the detailed division of permafrost areas is often neglected, and there is a lack of comprehensive consideration of factors such as vegetation coverage and sunlight exposure time. This makes it easy for traditional methods to make overly macroscopic assessments of the entire area, and fail to deeply explore the foundation gradient information in different regions and seasons, resulting in the localization and individual characteristics of the gradient changes not being fully considered. In addition, existing technologies often lack differentiated methods in governance. Usually, once a problem with the foundation is detected, the conventional treatment method is often to adopt relatively uniform maintenance and improvement measures, while ignoring the differentiated needs in different regions and different gradient states. Such a treatment approach may lead to a waste of resources and unsatisfactory governance results.

[0080] Example 1

[0081] The present invention provides a method for predicting the gradual change information of frozen soil foundation of highway in long-term operation, please refer to Figure 1 , including the following steps:

[0082] S1. Pre-acquire a GPS electronic map of the highway frozen soil foundation area to be predicted, obtain the groundwater depth of the highway frozen soil foundation area in the form of random sampling, and obtain the annual average precipitation and snowfall in the frozen soil foundation area, and comprehensively generate a foundation meltwater condition coefficient Rsx. When the foundation meltwater condition coefficient Rsx is higher than a preset condition threshold, it indicates that there is a frozen soil meltwater risk, and the first processing information is sent to the system;

[0083] S2. When obtaining the first processing information, the frozen soil foundation area of ​​the highway is divided into multiple detection areas, detection points are set in the multiple detection areas, and the sensor group is assembled to detect the frozen soil state characteristics in two seasons, namely winter and summer, to establish a freezing data set and a melting data set, which are summarized as a frozen soil state characteristic data set, and a frozen soil state coefficient ZTx is generated according to the frozen soil state characteristic data set, and the frozen soil is determined to be a seasonal frozen soil area or a non-seasonal frozen soil area according to the frozen soil state coefficient ZTx;

[0084] S3, obtaining the vegetation coverage rate ZbF and the terrain elevation value Gc in the seasonal frozen soil area and the non-seasonal frozen soil area, correcting the frozen soil state coefficient ZTx, and using the correction result as the frozen soil melting coefficient Rh, re-dividing the detection area according to the value of the frozen soil melting coefficient Rh to obtain the division result;

[0085] S4. Based on the division results, establish a frozen soil feature database; store the geographical information of the frozen soil area, including topography, landform, soil depth, temperature distribution and deformation data; store the frozen soil state data of the frozen soil area, including soil temperature, humidity and water content, record soil type, frozen soil depth and seasonal change characteristics; obtain the gradual management plan of frozen soil foundation from public channels, summarize and obtain the information of frozen soil management plan, and establish a frozen soil foundation gradual management plan database;

[0086] Match the characteristics of the frozen soil area with the gradual change management schemes in the frozen soil foundation gradual change management scheme library, and output the corresponding frozen soil management scheme;

[0087] S5. Collect soil structure data and annual average sunlight exposure time Gz of the corresponding assigned permafrost areas, obtain the corresponding gradual governance plan, establish and train a convolutional neural network CNN, and generate a frozen land prediction and governance model. The frozen land prediction and governance model predicts and analyzes the matching gradual governance plan, obtains the prediction and analysis results, and selects the optimized governance plan based on the prediction and analysis results.

[0088] In this embodiment, step S1 is used to synthesize these data to generate the foundation meltwater condition coefficient Rsx, which is used to evaluate the meltwater risk of the foundation; step S2 determines whether the frozen soil is a seasonal frozen soil area and a non-seasonal frozen soil area based on the frozen soil state coefficient ZTx; step S3 obtains the vegetation coverage rate ZbF and the terrain elevation value Gc in the seasonal frozen soil area and the non-seasonal frozen soil area. Use these data to correct the frozen soil state coefficient ZTx to form the frozen soil melting coefficient Rh. Re-divide the detection area based on the value of Rh to obtain the division result; step S4, based on the division result, establish a frozen soil feature library to store characteristic information such as geographic information, soil structure data, and temperature distribution. At the same time, a frozen soil foundation gradual change management solution library is constructed to summarize the public gradual change management solution information. It is also used to match the characteristics of the frozen soil area with the gradual change management solution in the frozen soil foundation gradual change management solution library. The corresponding frozen soil management solution is output through this step. Step S5 collects the soil structure data of the corresponding frozen soil area that has been allocated and the annual average sunlight exposure time Gz. The corresponding gradual governance scheme is obtained, and a convolutional neural network (CNN) is established and trained to generate a frozen land prediction and governance model. The frozen land prediction and governance model performs prediction and analysis on the matching gradual governance scheme to obtain prediction and analysis results. Finally, the optimized governance scheme is selected based on the prediction and analysis results to improve the stability and safety of the frozen ground foundation.

[0089] This method uses geographic information, meteorological data, sensor monitoring and other information in combination with advanced technical means to achieve all-round and multi-angle capture and analysis of the gradual change information of highway frozen soil foundation, thus providing more accurate, timely and effective management means for long-term operation.

[0090] Embodiment 2: This embodiment is an explanation of Embodiment 1. Specifically, S1 includes:

[0091] S11, obtaining a GPS electronic map of the highway frozen ground area to be predicted, where the GPS electronic map includes spatial resolution and detailed geographic information;

[0092] S12. In the frozen soil foundation area of ​​the highway, random sampling is performed to obtain groundwater depth data, and the groundwater depth Dxs is obtained by using groundwater level monitoring wells or water level sensor equipment according to the wellbore measurement method;

[0093] S13. Obtain the annual average precipitation JS and annual average snowfall JX in the frozen soil foundation area of ​​the highway from the meteorological station to evaluate the surface moisture status and the melting of frozen soil, and calculate the foundation meltwater condition coefficient Rsx; use the collected groundwater depth Dxs, annual average precipitation JS and annual average snowfall JX, after dimensionless processing, to generate the foundation meltwater condition coefficient Rsx according to the following formula:

[0094]

[0095] In the formula, W1>W2, W1 and W2 are the proportional coefficients of the annual average precipitation JS and the annual average snowfall JX, and W1+W2=1.0.

[0096] S14. Compare the calculated foundation meltwater condition coefficient Rsx with the preset condition threshold. If the foundation meltwater condition coefficient Rsx is higher than the condition threshold, it indicates the risk of frozen soil meltwater. The first label is set in the current highway frozen soil foundation area, and the first processing information is sent to the system.

[0097] In this embodiment, the first processing information is to respond to potential frozen soil thaw problems in a timely manner so as to take necessary management and preventive measures. S1 realizes real-time monitoring and evaluation of frozen soil foundation areas of highways, providing basic data and alarm information for subsequent management and prediction.

[0098] Embodiment 3: This embodiment is an explanation of Embodiment 1. Specifically, S2 includes:

[0099] S21. Use geographic information system (GIS) tools to divide the frozen soil foundation area of ​​the highway into multiple detection areas and set up detection points in each area. This step aims to divide the area into small pieces to more accurately monitor and understand the frozen soil status of each area.

[0100] S22. In winter, the soil temperature and humidity values ​​of the surface layer, frozen soil surface and active layer are collected through the sensor group, and the groundwater level depth is measured at the same time, forming a frozen data set;

[0101] S23. In summer, the soil temperature and moisture values ​​at the same location are collected by the sensor group, and the groundwater level depth is measured at the same time, forming a melting data set;

[0102] S24. The freezing data set and the melting data set are merged to form a comprehensive frozen soil state characteristic data set; including the soil temperature and moisture values ​​of the surface layer, frozen soil surface and active layer at different detection points in winter and summer, as well as the information of groundwater level depth.

[0103] S26. For each detection point, calculate the average values ​​of soil temperature and moisture of the surface layer, frozen soil surface and active layer in winter and summer, as well as the average value of groundwater level depth;

[0104] S27, based on the average value calculated in S26, dimensionless processing, the frozen soil state coefficient ZTx is calculated by the following formula:

[0105]

[0106] In the formula, DdbcT1 represents the mean surface temperature in winter, XdbcS1 represents the mean surface humidity in summer; XdbcT2 represents the mean surface temperature in summer, DdbcS2 represents the mean surface humidity in winter; DdtcT1 represents the mean surface temperature of the frozen soil layer in winter, XdtcS1 represents the mean surface humidity of the frozen soil layer in summer; XdtcT2 represents the mean surface temperature of the frozen soil layer in summer, DdtcS2 represents the mean surface humidity of the frozen soil layer in winter; DhdcT1 represents the mean temperature of the active layer in winter, XhdcS1 represents the mean humidity of the active layer in summer; XhdcT2 represents the mean temperature of the active layer in summer, DhdcS2 represents the mean humidity of the active layer in winter; SdxsJz represents the mean depth of the groundwater level in summer, DdxsJz represents the mean depth of the groundwater level in winter;

[0107] S28. Preset a first threshold value, which serves as the dividing line between seasonal frozen soil and non-seasonal frozen soil. For each detection point, compare the calculated frozen soil state coefficient ZTx with the first threshold value. If the frozen soil state coefficient ZTx is greater than the first threshold value, it is determined to be a seasonal frozen soil area; if the frozen soil state coefficient ZTx is less than or equal to the first threshold value, it is determined to be a non-seasonal frozen soil area.

[0108] In this embodiment, step S2 forms a frozen and thawed data set by collecting soil temperature and humidity values ​​of the surface layer, frozen soil surface and active layer in winter and summer respectively, and measuring the depth of the groundwater level at the same time. This multi-dimensional data collection helps to fully understand the state of frozen soil foundations in different seasons and provides more detailed and comprehensive information. By merging the frozen data set and the thawed data set, the frozen soil state coefficient ZTx is calculated, and then divided according to the preset first threshold, the seasonal frozen soil area and the non-seasonal frozen soil area are clarified. It helps to eliminate seasonal and regional differences, making the model more universal and better adaptable to foundation conditions under different conditions. This helps to better maintain and manage the frozen soil foundation of the highway and ensure its stability and safety in long-term operation.

[0109] Embodiment 4: This embodiment is an explanation of Embodiment 1. Specifically, S3 includes:

[0110] S31. Use remote sensing technology to obtain image data of seasonal frozen soil and non-seasonal frozen soil areas, extract vegetation coverage area from image data, and obtain vegetation coverage rate ZbF; use radar interferometry InSAR technology to obtain digital elevation model DEM, use synthetic aperture radar SAR satellite data to generate radar elevation image, SAR technology penetrates clouds and vegetation to obtain surface elevation information, use stereo image matching technology to compare optical images of uniform objects at different viewing angles, calculate and obtain initial terrain elevation value Gc1, use GPS for ground measurement, obtain second terrain elevation value Gc2, correct the initial terrain elevation value Gc1, and obtain terrain elevation value Gc;

[0111] S32, the frozen soil state coefficient ZTx is corrected by the collected vegetation coverage rate ZbF and terrain elevation value Gc, and a correction model is established, and the frozen soil melting coefficient Rh is obtained by the following formula;

[0112] Rh=α*ZTx+β*ZbF+γ*Gc+∈

[0113] In the formula, α represents the first correction coefficient, β represents the second correction coefficient, represents the influence of vegetation coverage on the frozen soil state; γ represents the influence of terrain elevation on the frozen soil state, and ∈ represents the error term;

[0114] S33. Set a stable threshold, a partial melting threshold and a full melting threshold, and compare the frozen soil melting coefficient Rh with the stable threshold, the partial melting threshold and the full melting threshold respectively. If it matches within the stable threshold interval, it is represented as a stable state area; if it matches within the partial melting threshold interval, it is represented as a partial melting state area; if it matches within the full melting threshold interval, it is represented as a full melting state area.

[0115] In this embodiment, the vegetation coverage rate and terrain elevation value obtained by remote sensing technology in step S3 are used to correct the permafrost state coefficient. This helps to more accurately reflect the impact of vegetation and terrain on the permafrost state and improve the accuracy of prediction. Combined with remote sensing technology and GPS measurement, terrain information is obtained by multi-source data fusion. Such a method can make up for the shortcomings of a single data source and provide more comprehensive foundation characteristics. Using the corrected permafrost thawing coefficient Rh, different thresholds are set for regional division, making the assessment of the permafrost state more refined, which helps to propose targeted governance plans. Using the data obtained by remote sensing technology, real-time monitoring and prediction of the foundation state are achieved. This helps to discover potential problems in a timely manner, take corresponding governance and maintenance measures, and promote the safe and stable operation of highway frozen foundations.

[0116] Example 5: This example is an explanation of Example 1. Specifically, S4 includes:

[0117] S41, storing the acquired geographic information data in a database; including topography, landform, soil depth, temperature distribution and deformation data; each data item is identified, organized and classified;

[0118] S42. Frozen soil status data is stored in the same database; including soil temperature, humidity and water content data; each data point can be associated with a specific geographic location and has a timestamp to reflect the temporal characteristics of the data;

[0119] S43, establish a frozen soil characteristic database, integrate geographic information and frozen soil status data into the frozen soil characteristic database; record soil type and frozen soil depth information in the frozen soil characteristic database; and for seasonal change characteristics, record the changes in soil temperature, humidity and water content in different seasons;

[0120] S44. Obtain information on gradual management plans for frozen ground from public channels, including monitoring and maintenance measures and environmental control plans;

[0121] S45. Summarize and organize the acquired gradual change control scheme information and establish a database of gradual change control schemes for frozen soil foundations, including detailed descriptions of the schemes, application conditions and effects;

[0122] S46. Using a database query language or a programming language, matching the features in the frozen soil feature library with the solutions in the frozen soil foundation gradual change management solution library based on soil type, frozen soil depth, and seasonal changes, and outputting the corresponding frozen soil management solution according to the matching results; including:

[0123] Stable state area: regular inspection and monitoring, and establishment of a maintenance protection layer on the foundation surface;

[0124] Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and strengthen drainage control;

[0125] In the area of ​​complete melting: improve the foundation, strengthen pile foundation, reinforce grouting and strengthen drainage control; and regulate the environment to control and reduce soil moisture.

[0126] In this embodiment, differentiated governance is achieved by proposing differentiated governance plans for areas with different frozen soil states, avoiding relatively uniform maintenance and improvement measures for the entire area, and improving the governance effect. By establishing a database to store geographic information and frozen soil state data, real-time monitoring and timely maintenance of foundation conditions are achieved, which helps to discover potential problems in advance and take measures to ensure the safe operation of frozen soil foundations on highways. By establishing a frozen soil foundation gradual change governance plan library, sustainable maintenance plans for different situations are provided, providing a scientific basis for long-term operation.

[0127] Embodiment 6: This embodiment is an explanation of Embodiment 1. Specifically, S5 includes:

[0128] S51, collecting soil structure data in the assigned corresponding frozen soil area, involving sampling and field measurement to obtain the physical and chemical properties of the soil; and obtaining the annual average sunshine exposure time Gz in the corresponding frozen soil area through meteorological information;

[0129] S52, obtaining a treatment plan for the corresponding frozen soil area from an existing frozen soil foundation gradual change treatment plan library; the obtained plan is matched with the soil structure data and the annual average sunlight exposure time Gz;

[0130] S53, organizing the collected soil structure data and sunshine time data into a format suitable for the convolutional neural network CNN, using the existing soil structure data and annual average sunshine time Gz data, and the corresponding frozen land management plan as labels, to train the convolutional neural network CNN; using the trained frozen land prediction and management model to predict and analyze the new soil structure data and annual average sunshine time Gz data; obtaining the optimization prediction results of the frozen land prediction and management model for the frozen land management plan, including:

[0131] Stable state area: Regular inspection and monitoring, and building a maintenance protection layer on the foundation surface, using waterproof material protection layer, and increasing vegetation coverage when the annual average sunshine exposure time Gz is greater than the sunshine exposure threshold, increasing shading time in warm seasons, and increasing lighting in cold seasons to control the surface temperature;

[0132] Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and increase shading time in warm seasons and strengthen drainage control when the annual average sunshine duration Gz is greater than the sunshine threshold;

[0133] In the area of ​​full melting state: improve the foundation, add bored piles and vibrating piles, reinforce with grouting, increase the viscosity of the soil and strengthen drainage control; and when the annual average sunshine exposure time Gz is greater than the sunshine exposure threshold, increase the vegetation coverage, increase the shade time in the warm season, regulate the environment, add geothermal system and control and reduce soil moisture.

[0134] In this embodiment, a frozen land prediction and management model is constructed through a convolutional neural network (CNN), and intelligent prediction and analysis is performed on new soil structure and sunlight exposure time data to obtain the model's optimized prediction results for different management plans, thereby improving the intelligence and accuracy of management. According to the prediction results of the model, differentiated management suggestions are provided for areas with different frozen soil states, so as to better adapt to the actual conditions of different regions and improve the feasibility and effectiveness of management. The factor of sunlight exposure time is taken into account in the management suggestions, making the management plan more environmentally adaptable and helping to maintain the stability of the frozen soil foundation of the highway under different seasons and climatic conditions.

[0135] Example 7, a system for predicting the gradual change of frozen soil foundation of highway in long-term operation, please refer to Figure 2 , including data acquisition module, monitoring point setting and data set establishment module, permafrost state classification module, vegetation terrain acquisition module, correction module, permafrost feature library establishment module, matching module, soil structure and sunlight exposure time acquisition module, model training module and optimization management module;

[0136] The data acquisition module obtains the GPS electronic map of the highway frozen soil foundation area to be predicted; randomly samples the groundwater depth, and obtains the precipitation and snowfall to generate the foundation meltwater condition coefficient Rsx;

[0137] The monitoring point setting and data set establishment module is used to divide the highway frozen soil foundation area into multiple detection areas, and set detection points to collect soil temperature, humidity, and groundwater depth data in winter and summer through sensor groups to form frozen data sets and melted data sets;

[0138] The frozen soil state classification module calculates the frozen soil state coefficient ZTx based on the collected data and divides the seasonal frozen soil area into the non-seasonal frozen soil area;

[0139] Vegetation terrain acquisition module, using remote sensing technology to obtain vegetation coverage ZbF and terrain elevation Gc;

[0140] The correction module is used to correct the frozen soil state coefficient ZTx according to the vegetation coverage rate ZbF and the terrain elevation value Gc through the correction model to obtain the frozen soil thawing coefficient Rh;

[0141] The frozen soil characteristic database establishment module is used to store geographic information and frozen soil status data, including topography, landform, soil depth, temperature distribution and deformation data, as well as soil temperature, humidity and water content, and obtain gradual change management plan information from public channels to establish a frozen soil foundation gradual change management plan library;

[0142] The matching module is used to match the features in the frozen soil foundation feature library with the solutions in the frozen soil foundation gradual change treatment solution library according to the division results, and output the corresponding frozen soil treatment solution;

[0143] The soil structure and sunshine exposure time acquisition module is used to collect soil structure data and annual average sunshine exposure time Gz of the allocated area;

[0144] The model training module uses the convolutional neural network (CNN) for training to generate a frozen land prediction and management model;

[0145] The optimization management module uses the trained frozen land prediction and management model to perform predictive analysis on new soil structure and sunlight exposure time data to output an optimized frozen land management plan, including specific management measures for stable state areas, partially thawed state areas, and fully thawed state areas.

[0146] This system not only improves the prediction accuracy of gradual change information in the long-term operation of frozen soil foundations of highways, but also provides specific and personalized management solutions, promoting significant results in improving the stability and safety of frozen soil foundations of highways.

[0147] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting gradual changes in frozen soil foundation of highway during long-term operation, characterized by: The following steps are included: S1. Pre-acquire a GPS electronic map of the highway frozen soil foundation area to be predicted, obtain the groundwater depth of the highway frozen soil foundation area in the form of random sampling, and obtain the annual average precipitation and snowfall in the frozen soil foundation area, and comprehensively generate a foundation meltwater condition coefficient Rsx. When the foundation meltwater condition coefficient Rsx is higher than a preset condition threshold, it indicates that there is a frozen soil meltwater risk, and the first processing information is sent to the system; The collected groundwater depth Dxs, annual average precipitation JS and annual average snowfall JX are used to generate the foundation meltwater condition coefficient Rsx according to the following formula after dimensionless processing: Where W1>W2, W1 and W2 are the proportional coefficients of the annual average precipitation JS and the annual average snowfall JX, and W1+W2=1.0; S2. When obtaining the first processing information, the frozen soil foundation area of ​​the highway is divided into multiple detection areas, detection points are set in the multiple detection areas, and the sensor group is assembled to detect the frozen soil state characteristics in two seasons, namely winter and summer, to establish a freezing data set and a melting data set, which are summarized as a frozen soil state characteristic data set, and a frozen soil state coefficient ZTx is generated according to the frozen soil state characteristic data set, and the frozen soil is determined to be a seasonal frozen soil area or a non-seasonal frozen soil area according to the frozen soil state coefficient ZTx; The S2 specifically includes: S21. Use geographic information system (GIS) tools to divide the frozen soil foundation area of ​​the highway into multiple detection areas, and set detection points in each area; S22. In winter, the soil temperature and humidity values ​​of the surface layer, frozen soil surface and active layer are collected through the sensor group, and the groundwater level depth is measured at the same time, forming a frozen data set; S23. In summer, the soil temperature and moisture values ​​at the same location are collected by the sensor group, and the groundwater level depth is measured at the same time, forming a melting data set; S24, merging the frozen data set and the thawed data set to form a comprehensive frozen soil state characteristic data set; including soil temperature and moisture values ​​of the surface layer, frozen soil surface and active layer at different detection points in winter and summer, as well as information on the depth of the groundwater level; S26. For each detection point, calculate the average values ​​of soil temperature and humidity of the surface layer and frozen soil layer, active layer in winter and summer, and the average value of groundwater level depth; S27, based on the average value calculated in S26, dimensionless processing, the frozen soil state coefficient ZTx is calculated by the following formula: In the formula, DdbcT1 represents the mean surface temperature in winter, XdbcS1 represents the mean surface humidity in summer; XdbcT2 represents the mean surface temperature in summer, DdbcS2 represents the mean surface humidity in winter; DdtcT1 represents the mean surface temperature of the frozen soil layer in winter, XdtcS1 represents the mean surface humidity of the frozen soil layer in summer; XdtcT2 represents the mean surface temperature of the frozen soil layer in summer, DdtcS2 represents the mean surface humidity of the frozen soil layer in winter; DhdcT1 represents the mean temperature of the active layer in winter, XhdcS1 represents the mean humidity of the active layer in summer; XhdcT2 represents the mean temperature of the active layer in summer, DhdcS2 represents the mean humidity of the active layer in winter; SdxsJz represents the mean depth of the groundwater level in summer, DdxsJz represents the mean depth of the groundwater level in winter; S28, presetting a first threshold value, the first threshold value is used as the dividing line between seasonal frozen soil and non-seasonal frozen soil, for each detection point, comparing the calculated frozen soil state coefficient ZTx with the first threshold value, if the frozen soil state coefficient ZTx is greater than the first threshold value, it is determined to be a seasonal frozen soil area; if the frozen soil state coefficient ZTx is less than or equal to the first threshold value, it is determined to be a non-seasonal frozen soil area; S3, obtaining the vegetation coverage rate ZbF and the terrain elevation value Gc in the seasonal frozen soil area and the non-seasonal frozen soil area, correcting the frozen soil state coefficient ZTx, and using the correction result as the frozen soil melting coefficient Rh, re-dividing the detection area according to the value of the frozen soil melting coefficient Rh to obtain the division result; The S3 specifically includes: S31. Use remote sensing technology to obtain image data of seasonal frozen soil and non-seasonal frozen soil areas, extract vegetation coverage area from image data, and obtain vegetation coverage rate ZbF; use radar interferometry InSAR technology to obtain digital elevation model DEM, use synthetic aperture radar SAR satellite data to generate radar elevation image, SAR technology penetrates clouds and vegetation to obtain surface elevation information, use stereo image matching technology to compare optical images of uniform objects at different viewing angles, calculate and obtain initial terrain elevation value Gc1, use GPS for ground measurement, obtain second terrain elevation value Gc2, correct the initial terrain elevation value Gc1, and obtain terrain elevation value Gc; S32, the frozen soil state coefficient ZTx is corrected by the collected vegetation coverage rate ZbF and terrain elevation value Gc, and a correction model is established, and the frozen soil melting coefficient Rh is obtained by the following formula; Rh=α*ZTx+β*ZbF+γ*Gc+∈ In the formula, α represents the first correction coefficient, β represents the second correction coefficient, represents the influence of vegetation coverage on the frozen soil state; γ represents the influence of terrain elevation on the frozen soil state, and ∈ represents the error term; S33, setting a stable threshold, a partial melting threshold and a full melting threshold, and comparing the frozen soil melting coefficient Rh with the stable threshold, the partial melting threshold and the full melting threshold respectively, and matching within the stable threshold interval, indicating a stable state area; matching within the partial melting threshold interval, indicating a partial melting state area; matching within the full melting threshold interval, indicating a full melting state area; S4. Based on the division results, establish a frozen soil feature database; store the geographical information of the frozen soil area, including topography, landform, soil depth, temperature distribution and deformation data; store the frozen soil state data of the frozen soil area, including soil temperature, humidity and water content, record soil type, frozen soil depth and seasonal change characteristics; obtain the gradual management plan of frozen soil foundation from public channels, summarize and obtain the information of frozen soil management plan, and establish a frozen soil foundation gradual management plan database; Match the characteristics of the frozen soil area with the gradual change management schemes in the frozen soil foundation gradual change management scheme library, and output the corresponding frozen soil management scheme; S5. Collect soil structure data and annual average sunshine exposure time Gz of the corresponding frozen soil area that has been assigned, obtain the corresponding frozen soil management plan, establish and train a convolutional neural network CNN, and generate a frozen soil prediction and management model. The frozen soil prediction and management model predicts and analyzes the matching frozen soil management plan, obtains the prediction and analysis results, and selects the optimized management plan based on the prediction and analysis results; The S5 specifically includes: S51, collecting soil structure data in the assigned corresponding frozen soil area, involving sampling and field measurement to obtain the physical and chemical properties of the soil; and obtaining the annual average sunshine exposure time Gz in the corresponding frozen soil area through meteorological information; S52, obtaining a frozen soil treatment plan corresponding to the frozen soil area from an existing frozen soil foundation gradual change treatment plan library; the obtained plan is matched with the soil structure data and the annual average sunlight exposure time Gz; S53, organizing the collected soil structure data and sunshine time data into a format suitable for the convolutional neural network CNN, using the existing soil structure data and annual average sunshine time Gz data, and the corresponding frozen land management plan as labels, to train the convolutional neural network CNN; using the trained frozen land prediction and management model to predict and analyze the new soil structure data and annual average sunshine time Gz data; obtaining the optimization prediction results of the frozen land prediction and management model for the frozen land management plan, including: Stable state area: Regular inspection and monitoring, and building a maintenance protection layer on the foundation surface, using waterproof material protection layer, and increasing vegetation coverage when the annual average sunshine exposure time Gz is greater than the sunshine exposure threshold, increasing shading time in warm seasons, and increasing lighting in cold seasons to control the surface temperature; Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and increase shading time in warm seasons and strengthen drainage control when the annual average sunshine duration Gz is greater than the sunshine threshold; In the area of ​​full melting: improve the foundation, add bored piles and vibrating piles, reinforce with grouting, increase the viscosity of the soil, and strengthen drainage control; and when the annual average sunshine time Gz is greater than the sunshine threshold, increase the vegetation coverage, increase the shade time in the warm season, regulate the environment, add geothermal system, and control and reduce soil moisture.

2. The method for predicting gradual change information of frozen soil foundation of highway in long-term operation according to claim 1 is characterized by: The S1 specifically includes: S11, obtaining a GPS electronic map of the highway frozen ground area to be predicted, where the GPS electronic map includes spatial resolution and detailed geographic information; S12. In the frozen soil foundation area of ​​the highway, random sampling is performed to obtain groundwater depth data, and the groundwater depth Dxs is obtained by using groundwater level monitoring wells or water level sensor equipment according to the wellbore measurement method; S13, obtaining the annual average precipitation JS and annual average snowfall JX in the frozen soil foundation area of ​​the highway from the meteorological station to evaluate the surface moisture state and the melting of frozen soil, and calculate the foundation meltwater condition coefficient Rsx; S14. Compare the calculated foundation meltwater condition coefficient Rsx with the preset condition threshold. If the foundation meltwater condition coefficient Rsx is higher than the condition threshold, it indicates the risk of frozen soil meltwater. The first label is set in the current highway frozen soil foundation area, and the first processing information is sent to the system.

3. The method for predicting gradual change information of frozen soil foundation of highway in long-term operation according to claim 1 is characterized by: The S4 specifically includes: S41, storing the acquired geographic information data in a database; including topography, landform, soil depth, temperature distribution and deformation data; each data item is identified, organized and classified; S42. Frozen soil status data is stored in the same database; including soil temperature, humidity and water content data; each data point can be associated with a specific geographic location and has a timestamp to reflect the temporal characteristics of the data; S43, establish a frozen soil characteristic database, integrate geographic information and frozen soil status data into the frozen soil characteristic database; record soil type and frozen soil depth information in the frozen soil characteristic database; and for seasonal change characteristics, record the changes in soil temperature, humidity and water content in different seasons; S44. Obtain information on gradual management plans for frozen ground from public channels, including monitoring and maintenance measures and environmental control plans; S45. Summarize and organize the acquired gradual change control scheme information and establish a database of gradual change control schemes for frozen soil foundations, including detailed descriptions of the schemes, application conditions and effects; S46. Using a database query language or a programming language, matching the features in the frozen soil feature library with the solutions in the frozen soil foundation gradual change management solution library based on soil type, frozen soil depth and seasonal changes, and outputting the corresponding frozen soil management solution according to the matching results; including: Stable state area: regular inspection and monitoring, and establishment of a maintenance protection layer on the foundation surface; Partially melted areas: real-time reinforcement measures to strengthen foundation support, including reinforced concrete pile foundations and underground retaining walls, and strengthen drainage control; In the area of ​​complete melting: improve the foundation, strengthen pile foundation, reinforce grouting and strengthen drainage control; and regulate the environment to control and reduce soil moisture.

4. A system for predicting gradual changes in frozen soil foundation of a highway during long-term operation, used to implement the method for predicting gradual changes in frozen soil foundation of a highway during long-term operation as described in any one of claims 1 to 3, characterized in that: It includes data acquisition module, monitoring point setting and data set establishment module, permafrost state classification module, vegetation terrain acquisition module, correction module, permafrost feature library establishment module, matching module, soil structure and sunlight exposure time acquisition module, model training module and optimization management module; The data acquisition module obtains the GPS electronic map of the highway frozen soil foundation area to be predicted; randomly samples the groundwater depth, and obtains the precipitation and snowfall to generate the foundation meltwater condition coefficient Rsx; The monitoring point setting and data set establishment module is used to divide the highway frozen soil foundation area into multiple detection areas, and set detection points to collect soil temperature, humidity and groundwater depth data in winter and summer through sensor groups to form frozen data sets and melted data sets; The frozen soil state classification module calculates the frozen soil state coefficient ZTx based on the collected data and divides the seasonal frozen soil area into the non-seasonal frozen soil area; Vegetation terrain acquisition module, using remote sensing technology to obtain vegetation coverage ZbF and terrain elevation Gc; The correction module is used to correct the frozen soil state coefficient ZTx according to the vegetation coverage rate ZbF and the terrain elevation value Gc through the correction model to obtain the frozen soil thawing coefficient Rh; The frozen soil characteristic database establishment module is used to store geographic information and frozen soil status data, including topography, landform, soil depth, temperature distribution and deformation data, as well as soil temperature, humidity and water content, and obtain gradual change management plan information from public channels to establish a frozen soil foundation gradual change management plan library; The matching module is used to match the features in the frozen soil feature library with the solutions in the frozen soil foundation gradual change treatment solution library according to the division results, and output the corresponding frozen soil treatment solution; The soil structure and sunshine exposure time acquisition module is used to collect soil structure data and annual average sunshine exposure time Gz of the allocated area; The model training module uses the convolutional neural network (CNN) for training to generate a frozen land prediction and management model; The optimization management module uses the trained frozen land prediction and management model to perform predictive analysis on new soil structure and sunlight exposure time data to output an optimized frozen land management plan, including specific management measures for stable state areas, partially thawed state areas, and fully thawed state areas.

Images