Railway tunnel structure state remote intelligent monitoring system and method

Abstract

The invention discloses a railway tunnel structure state remote intelligent monitoring system and method, and relates to the technical field of intelligent monitoring, and the system comprises a sensing layer which is used for collecting the related information of a tunnel structure; the transmission layer is used for transmitting the relevant information of the tunnel structure acquired by the sensing layer to the processing layer; the processing layer is used for inputting the related information of the tunnel structure into a tunnel simulation model to generate simulation data, inputting the simulation data and the related information of the tunnel structure into a risk prediction model, outputting a risk category and a risk level, and calling a guidance strategy according to the risk category and the risk level; and the application layer is used for displaying the related information, the risk category, the risk level and the guidance strategy of the tunnel structure, and generating an execution command according to the input instruction. Through introduction of the tunnel simulation model and the deep neural network, early warning can be carried out on possible structural abnormity or safety risks of the tunnel, corresponding countermeasures are matched, and the influence caused by emergencies is reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and more particularly to a remote intelligent monitoring system and method for the structural state of a railway tunnel. Background Art

[0002] With the continuous expansion of the railway network and the increase in the operating time of existing lines, railway tunnels, as one of the key infrastructures, are of vital importance in ensuring the safe and efficient operation of trains. During long-term use, tunnels will face dual challenges from the external environment and internal operating conditions, including but not limited to changes in geological conditions, the impact of extreme weather, and dynamic loads caused by frequent train traffic. The combined effect of these factors may cause stress concentration, structural deformation, cracks, and even more serious damage to the tunnel, thus posing a safety hazard.

[0003] Traditional railway tunnel monitoring and maintenance methods mainly include regular manual inspections and the use of basic monitoring equipment. However, this method has significant limitations:

[0004] Long inspection cycle: Due to human resource limitations, manual inspections often cannot achieve high frequency and full coverage, resulting in potential problems remaining undetected for a long time.

[0005] Single-modal data: Existing monitoring equipment can usually only provide a specific type of monitoring data (such as stress or vibration), and lacks the ability to effectively integrate and deeply analyze data from different sources, which can easily lead to low accuracy in risk prediction.

[0006] Therefore, how to improve the working efficiency and accuracy of the monitoring system is an urgent problem to be solved by those skilled in the art. Summary of the invention

[0007] In view of this, the present invention provides a remote intelligent monitoring system and method for the structural status of a railway tunnel, which overcomes the above-mentioned defects.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] A remote intelligent monitoring system for railway tunnel structure status, comprising: a perception layer, a transmission layer, a processing layer and an application layer;

[0010] The perception layer is used to collect relevant information about the tunnel structure; the relevant information about the tunnel structure includes tunnel image data, point cloud data, tunnel structure status data and environmental data;

[0011] The transmission layer is used to transmit the relevant information of the tunnel structure collected by the perception layer to the processing layer;

[0012] The processing layer is used to input the relevant information of the tunnel structure into a tunnel simulation model to generate simulation data, input the simulation data and the relevant information of the tunnel structure into a risk estimation model, output a risk category and a risk level, and call a guidance strategy according to the risk category and the risk level;

[0013] The application layer is used to display the relevant information of the tunnel structure, the risk category, the risk level and the guidance strategy, and generate an execution command according to the input instruction.

[0014] Optionally, the perception layer includes an image data acquisition module, a point cloud data acquisition module and a tunnel structure status data acquisition module;

[0015] The image data acquisition module is used to acquire image data of the tunnel;

[0016] The point cloud data acquisition module is used to collect point cloud data of the tunnel;

[0017] The tunnel structure status data acquisition module is used to collect stress data, deformation data, vibration data and environmental data of the tunnel.

[0018] Optionally, the processing layer includes a data pre-processing module, a data storage module, a risk estimation module and a guidance strategy generation module;

[0019] The data pre-processing module is used to pre-process the relevant information of the tunnel structure and obtain pre-processed data;

[0020] The data storage module is used to classify and store the pre-processed data;

[0021] The risk estimation module is used to generate the simulation data based on the pre-processed data input into the tunnel simulation model, and generate the risk category and risk level using the risk estimation model according to the simulation data and the pre-processed data;

[0022] The guidance strategy generation module is used to call a pre-stored guidance strategy according to the risk category and the risk level.

[0023] Optionally, the data pre-processing module includes a data cleaning submodule, a data noise reduction submodule, a data alignment submodule, and a data compression submodule;

[0024] The data cleaning submodule is used to remove abnormal values ​​and supplement missing values ​​from the relevant information of the tunnel structure to generate cleaned data;

[0025] The data denoising submodule is used to perform smoothing and / or multi-scale decomposition on the cleaned data, filter out noise, and generate denoised data;

[0026] The data alignment submodule is used to perform spatiotemporal alignment and time synchronization on the noise reduction data to generate aligned data;

[0027] The data compression submodule is used to analyze and extract features from the alignment data, compress the extracted features, and obtain preprocessed data.

[0028] Optionally, the processing layer also includes a monitoring device self-correction module for performing life prediction and monitoring mode management on the monitoring device.

[0029] Optionally, the application layer includes a command input module, a positioning module, a data display module, and a command generation module;

[0030] The command input module is used to receive external control commands;

[0031] The positioning module is used to provide location data according to the risk category and risk level;

[0032] The data display module is used to display the relevant information of the tunnel structure, the risk category, the risk level, the location data and the guidance strategy;

[0033] The command generation module is used to split the control instruction or the guidance strategy into execution commands.

[0034] A method for remote intelligent monitoring of railway tunnel structure status, comprising the following specific steps:

[0035] Collecting relevant information of the tunnel structure, wherein the relevant information of the tunnel structure includes image data, point cloud data, tunnel structure status data and environmental data of the tunnel;

[0036] Preprocessing the relevant information of the tunnel structure to obtain preprocessing data;

[0037] Inputting the preprocessed data into a tunnel simulation model to generate simulation data, inputting the simulation data and the preprocessed data into a risk prediction model to output a risk category and a risk level; and calling a guidance strategy according to the risk category and the risk level;

[0038] The guidance strategy is split into execution commands, and the execution commands are transmitted to the execution device.

[0039] Optional preprocessing steps are:

[0040] Identify data types in the relevant information of the tunnel structure, the data types including image type, point cloud type and sensor type;

[0041] The data of different data types are preprocessed according to the data type and a preset preprocessing method.

[0042] Optionally, the steps for obtaining the risk category and risk level are:

[0043] Inputting the preprocessed data into the tunnel simulation model to simulate different operation scenarios to obtain multiple simulation data;

[0044] Performing local spatial features on the plurality of simulation data and the pre-processed data to generate a spatial feature map;

[0045] Performing time feature extraction on the spatial feature map to obtain spatiotemporal features;

[0046] The spatiotemporal features are classified and predicted to obtain the risk level and the risk type.

[0047] Optionally, the steps for obtaining the guidance strategy are:

[0048] Build a plan library based on historical data;

[0049] A corresponding plan is matched from a plan library according to the risk level and the risk type, and a guiding strategy is generated according to the plan.

[0050] It can be seen from the above technical solutions that the present invention discloses a remote intelligent monitoring system and method for the structural state of a railway tunnel, which has the following beneficial effects compared with the prior art:

[0051] 1. The present invention integrates multi-modal monitoring equipment to collect various data such as stress, deformation, vibration, image, point cloud, etc. of the tunnel structure in real time, providing all-round and dynamic monitoring of the health status of the tunnel structure, ensuring timely discovery of potential hazards and ensuring safe operation of the tunnel;

[0052] 2. The present invention performs spatiotemporal alignment processing on the collected multimodal data to ensure that the data from different collection devices are effectively analyzed and compared under the same spatiotemporal reference, thereby improving the accuracy of the analysis and the utilization rate of the data;

[0053] 3. The present invention can provide early warning of possible structural anomalies or safety risks in the tunnel through the introduction of tunnel simulation models and deep neural networks, and match corresponding response measures to reduce the impact of emergencies;

[0054] 4. The present invention optimizes the monitoring scheme through a dynamic monitoring strategy adjustment scheme, improves monitoring efficiency and accuracy, and saves resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0056] Figure 1 Provide a schematic diagram of the system structure for the present invention;

[0057] Figure 2 A schematic diagram of the method provided by the present invention;

[0058] Figure 3 A schematic diagram of a tunnel simulation model provided by the present invention;

[0059] Figure 4 A cross-sectional schematic diagram of a tunnel simulation model is provided for the present invention. DETAILED DESCRIPTION

[0060] 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.

[0061] This embodiment discloses a remote intelligent monitoring system for railway tunnel structure status. Figure 1 As shown, it includes: perception layer, transmission layer, processing layer and application layer;

[0062] The perception layer is used to collect relevant information about the tunnel structure; the relevant information about the tunnel structure includes tunnel image data, point cloud data, tunnel structure status data and environmental data;

[0063] The transmission layer is used to transmit the relevant information of the tunnel structure collected by the perception layer to the processing layer;

[0064] The processing layer is used to input relevant information of the tunnel structure into the tunnel simulation model to generate simulation data, input the simulation data and relevant information of the tunnel structure into the risk estimation model, output the risk category and risk level, and call the guidance strategy according to the risk category and risk level;

[0065] The application layer is used to display relevant information, risk categories, risk levels and guidance strategies of the tunnel structure, and generate execution commands based on input instructions.

[0066] In one embodiment, the perception layer includes an image data acquisition module, a point cloud data acquisition module, and a tunnel structure status data acquisition module;

[0067] An image data acquisition module, used for acquiring image data of the tunnel;

[0068] Point cloud data acquisition module, used to collect point cloud data of the tunnel;

[0069] The tunnel structure status data acquisition module is used to collect tunnel stress data, deformation data, vibration data and environmental data.

[0070] Furthermore, in this embodiment, the image data acquisition module adopts a camera; the point cloud data acquisition module adopts a three-dimensional laser scanner; the tunnel structure status data acquisition module adopts a variety of sensors or optical fiber sensors; the various sensors include stress sensors (strain gauges), which are installed on the tunnel lining to detect stress changes caused by changes in geological conditions or external loads; displacement meters (deformation sensors), which are used to monitor the deformation of the tunnel wall or supporting structure; accelerometers (vibration sensors), which are used to record dynamic loads on the tunnel structure, such as vibrations caused by natural phenomena such as traffic flow and earthquakes; temperature sensors, which are used to monitor temperature changes in the tunnel; humidity sensors, which are used to monitor humidity changes in the tunnel; crack monitors, which are used to accurately measure the width and development trend of cracks; wind speed and direction sensors, which are used to monitor changes in wind speed and direction in the tunnel; water level sensors, which are used to monitor changes in water level in the tunnel; and rain gauges, which are used to monitor rainfall in the area where the tunnel is located.

[0071] In one embodiment, the processing layer includes a data pre-processing module, a data storage module, a risk estimation module, and a guidance strategy generation module;

[0072] A data pre-processing module, used to pre-process relevant information of the tunnel structure and obtain pre-processed data;

[0073] A data storage module, used for classifying and storing preprocessed data;

[0074] A risk estimation module is used to generate simulation data based on preprocessed data input into the tunnel simulation model, and generate risk categories and risk levels using the risk estimation model according to the simulation data and the preprocessed data;

[0075] The guidance strategy generation module is used to call pre-stored guidance strategies according to risk categories and risk levels.

[0076] In one embodiment, the data pre-processing module includes a data cleaning submodule, a data noise reduction submodule, a data alignment submodule, and a data compression submodule;

[0077] The data cleaning submodule is used to remove abnormal values ​​and supplement missing values ​​from the relevant information of the tunnel structure to generate cleaned data;

[0078] A data denoising submodule is used to perform smoothing and / or multi-scale decomposition on the cleaned data, filter out noise, and generate denoised data;

[0079] A data alignment submodule is used to perform spatiotemporal alignment and time synchronization on the noise reduction data to generate aligned data;

[0080] The data compression submodule is used to analyze and extract features from the aligned data, compress the extracted features, and obtain preprocessed data.

[0081] Furthermore, in the data cleaning submodule, an isolation forest or clustering algorithm is used to detect abnormal pixel values ​​or areas in the image data. For the detected abnormal values, they are selected for removal; for the missing pixel values, the interpolation of adjacent pixels (such as bilinear interpolation) or the image content-based repair algorithm (such as the Criminisi algorithm) can be used to fill them; for the point cloud data, the spatial distance and density of each point in the point cloud are calculated to identify and remove isolated points or abnormally dense areas, and the missing points are filled based on linear interpolation, nearest neighbor interpolation or surface reconstruction-based algorithms (such as Poisson surface reconstruction); for the tunnel structure status data, the threshold is set using the mean and standard deviation, and the abnormal data points exceeding the threshold are detected and removed, and the time series model (such as ARIMA) or regression model (such as linear regression, random forest regression) based on historical data is used to predict and fill the missing values.

[0082] In the data denoising submodule, for image data, mean filtering, median filtering or Gaussian filtering is used to remove noise in the image, retaining edges and details; then wavelet transform is used for multi-scale decomposition to extract detail information and low-frequency background in the image and filter out noise components; for point cloud data, statistical filtering or radius filtering is used to remove noise points, retain the main geometric structure, and then the point cloud data is converted into a three-dimensional image format, and then wavelet transform is applied for denoising, and a low-pass filter is used to remove high-frequency noise and retain useful signal components; for tunnel structure status data, wavelet transform is used to identify and filter out noise components in the signal, retain useful structural information, and fill in the gaps in the data using linear interpolation, polynomial interpolation or predictive interpolation based on time series models.

[0083] In one embodiment, the data alignment submodule includes a time alignment unit and a space alignment unit.

[0084] In one embodiment, the time alignment unit includes:

[0085] Calibrate the subunit to ensure that the timestamps of all devices (including those collecting image data, point cloud data, tunnel structure status data, and environmental data) are based on the same time reference through GPS or NTP services;

[0086] The time window matching subunit sets a reasonable time window (such as every 1 minute or every 10 seconds) according to the difference in data collection frequency; within each time window, the collected data is aggregated to obtain representative statistics such as average, maximum, and minimum values; data points from different sources are aligned according to the time window to ensure that they are compared and analyzed within the same time window.

[0087] In one embodiment, the specific structure of the space alignment unit includes:

[0088] The standard coordinate construction subunit is used to set a unified spatial coordinate system for all equipment and sensors monitoring the tunnel, such as the tunnel centerline as the x-axis, the vertical direction as the y-axis, and the longitudinal direction as the z-axis;

[0089] The position calibration subunit is used to define the precise position of each acquisition device inside the tunnel, including the distance from the tunnel entrance, height, and longitudinal position, using the measurement data during tunnel construction;

[0090] The spatial coordinate transformation subunit is used to uniformly map the data in different coordinate systems into standard coordinates using methods such as affine transformation or homogeneous coordinate transformation;

[0091] A spatial resolution unification subunit is used to unify the spatial resolution of data from different acquisition devices according to upsampling or downsampling;

[0092] The data filling submodule is used to select appropriate interpolation methods, such as inverse distance weighted interpolation and Kriging interpolation, according to the spatial distribution characteristics and missing conditions of the data, to ensure the integrity of the data for the missing data points in space.

[0093] In one embodiment, the spatial alignment unit further includes data verification, which is used to verify the data after the spatial and temporal alignment to ensure the accuracy and consistency of the data.

[0094] In one embodiment, the spatial coordinate transformation subunit calculates transformation parameters, such as rotation matrix, translation vector, etc., based on the installation position of the acquisition device and its coordinates in different coordinate systems, and then transforms the data in different coordinate systems into standard coordinates according to the transformation parameters.

[0095] This embodiment can achieve spatiotemporal alignment of four types of data, namely image data, point cloud data, tunnel structure status data and environmental data, through the data alignment submodule, ensuring the consistency of multi-source data within the same time and space range, so that data from different sensors can be analyzed and compared under the same framework, which helps to improve the data preprocessing accuracy of the railway tunnel monitoring system and the reliability of monitoring analysis.

[0096] In one embodiment, the specific structure of the risk estimation module is:

[0097] The feature prediction submodule is used to input the preprocessed data into the tunnel simulation model to perform risk simulation under different operation scenarios and obtain simulation data;

[0098] The risk prediction submodule is used to input the simulation data and preprocessed data into the risk estimation model to obtain the risk level and risk type.

[0099] In one embodiment, the processing layer further includes a monitoring device self-correction module for performing life prediction and monitoring mode management on the monitoring device.

[0100] Furthermore, monitoring mode management is to calculate the safety factor of the tunnel structure according to the risk level and risk type, supplement the monitoring frequency and the density of monitoring equipment for sections with lower safety factors, increase the monitoring space density and time density for areas with increased risk, and stop monitoring or reduce the monitoring frequency for sections with reduced risk.

[0101] In one embodiment, the guidance strategy generation module includes:

[0102] Automatic matching unit, used to automatically match relevant plans from the plan library according to the results of risk prediction;

[0103] An automatic sorting unit, used to arrange the matching guidance strategies in order of relevance;

[0104] Guide strategy generation, used to match relevant data according to the plan, such as risk level data.

[0105] In one embodiment, the application layer includes a command input module, a positioning module, a data display module, and a command generation module;

[0106] A command input module, used for receiving external control commands;

[0107] Positioning module, used to provide location data according to risk category and risk level;

[0108] The data display module is used to display relevant information of the tunnel structure, risk category, risk level, location data and guidance strategies;

[0109] The command generation module is used to split the control instructions or guidance strategies into execution commands.

[0110] Another aspect of the present embodiment includes a method for remote intelligent monitoring of the structural state of a railway tunnel, which is applied to the above system, such as Figure 2 As shown, the specific steps are:

[0111] Step 1: Collect relevant information of the tunnel structure, which includes tunnel image data, point cloud data, tunnel structure status data and environmental data;

[0112] Step 2: preprocess relevant information of the tunnel structure to obtain preprocessed data;

[0113] Step 3: Input the preprocessed data into the tunnel simulation model to generate simulation data, input the simulation data and the preprocessed data into the risk estimation model, output the risk category and risk level; call the guidance strategy according to the risk category and risk level;

[0114] Step 4: Split the guidance strategy into execution commands, and transmit the execution commands to the execution device.

[0115] In one embodiment, the pre-processing step in step 2 is:

[0116] Identify the data types in the relevant information of the tunnel structure, including image type, point cloud type and sensor type (including tunnel structure status data and environmental data);

[0117] According to the data type, data of different data types are cleaned, denoised, aligned and compressed according to the preset preprocessing method.

[0118] In one embodiment, the steps for obtaining the risk category and risk level are:

[0119] Perform feature extraction based on preprocessed data;

[0120] The preprocessed data is input into the tunnel simulation model to simulate different operation scenarios and obtain multiple simulation data;

[0121] Multiple simulation data and preprocessed data are input into the risk estimation model to obtain risk level and risk type.

[0122] In one embodiment, the steps of constructing the tunnel simulation model are:

[0123] Using image data and point cloud data as reference, import them into the selected modeling software, such as CAD or BIM software, to preliminarily construct the geometric outline of the tunnel; fine-tune the geometric outline of the tunnel based on the tunnel construction drawings, and integrate the relevant data of the tunnel into the tunnel geometric model, such as lining material, thickness, strength, etc., add them to the model in the form of attributes, and annotate and classify the various parts of the tunnel; then integrate the collected tunnel structure status data and environmental data, such as stress, temperature, humidity and wind speed, into the model in the form of data tables or sensors.

[0124] In one embodiment, after the tunnel simulation model is built, a professional simulation verification tool is used to perform static and dynamic verification analysis on the model and make necessary adjustments and optimizations to the model based on the verification results. The optimization content includes the model's geometric shape, structural parameters, environmental data, etc., to improve the accuracy and reliability of the simulation. The tunnel simulation model constructed in this embodiment, such as Figure 3 and Figure 4 As shown in the figure, the model can truly reflect the spatial structure and morphological characteristics of the tunnel.

[0125] In one embodiment, the operation scenarios are defined according to the actual needs and potential risks of tunnel operation, including train operation scenarios (the parameters of the train in the operation scenarios are adjustable), emergency braking scenarios, disaster scenarios (fire, flood, etc.) and extreme weather (heavy rain, heavy snow, etc.); and the corresponding simulation parameters are configured for the defined scenarios. For example, in the train operation scenario, the speed, load, and direction of operation of the train are set; in the emergency braking scenario, the starting time and deceleration of braking are set; in the fire scenario, the location, size, and duration of the fire source are set.

[0126] In one embodiment, the steps of acquiring simulation data are:

[0127] After initializing the tunnel simulation model, the model simulates various state changes inside the tunnel according to the set scenarios and parameters; and uses the data generated during the simulation process as simulation data, including temperature distribution, pressure changes, structural deformation and other characteristics inside the tunnel, as well as other relevant information such as train operation status and environmental parameters.

[0128] In one embodiment, the steps of constructing the risk estimation model are:

[0129] Data collection: Obtain historical data on tunnel risk events over the past period of time from the tunnel management department, including the time, location, type (such as structural cracks, groundwater leakage, landslide, etc.), level (minor, moderate, severe) of the risk events, as well as related images, structural state parameters (such as stress, etc.) and environmental data (such as rainfall, seismic activity).

[0130] Data preprocessing: different methods are used to preprocess data according to their types, and the preprocessed data is input into the tunnel simulation model to obtain simulation data. After aligning the simulation data with the mathematical data, the format is converted into a four-dimensional tensor;

[0131] Model construction: Using ConvLSTM as the model architecture, build an initial risk estimation model;

[0132] Model training: Based on known risk events in historical data, labels are created for each piece of data. The labels include the type and level of the risk event.

[0133] The converted data and corresponding labels are input into the initial risk estimation model, and the model parameters are optimized through the back propagation algorithm so that the model can accurately predict tunnel risks;

[0134] Cross-validation: Use cross-validation to evaluate the performance of the model and ensure the stability and generalization ability of the model;

[0135] Hyperparameter tuning: Adjust the model architecture and hyperparameters (such as learning rate, batch size, number of iterations, etc.) based on the validation results to improve the accuracy and efficiency of the model.

[0136] In one embodiment, the steps for obtaining the risk level and risk type are:

[0137] After the pre-processed data and simulation data are input into the risk estimation model, local spatial features are performed to generate a spatial feature map;

[0138] Perform temporal feature extraction on the spatial feature map to obtain spatiotemporal features;

[0139] Classify and predict the spatiotemporal features to output risk level or risk type labels.

[0140] The steps to obtain the guidance strategy are:

[0141] Construct a plan library based on historical data. The plan library is to establish a database containing various emergency plans. The plans should cover different types of risks and different levels of emergency response measures.

[0142] According to the risk level and risk type, the corresponding plan is matched from the plan library, and the guidance strategy is generated according to the plan. The guidance strategy includes the risk type and level of the risk point, recommended emergency measures, etc.; managers can formulate specific emergency response plans based on this information.

[0143] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0144] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A remote intelligent monitoring system for railway tunnel structure status, characterized in that: include: Perception layer, transport layer, processing layer and application layer; The perception layer is used to collect relevant information about the tunnel structure; The relevant information of the tunnel structure includes tunnel image data, point cloud data, tunnel structure status data and environment data; The transmission layer is used to transmit the relevant information of the tunnel structure collected by the perception layer to the processing layer; The processing layer is used to input the relevant information of the tunnel structure into a tunnel simulation model to generate simulation data, input the simulation data and the relevant information of the tunnel structure into a risk estimation model, output a risk category and a risk level, and call a guidance strategy according to the risk category and the risk level; The application layer is used to display the relevant information of the tunnel structure, the risk category, the risk level and the guidance strategy, and generate an execution command according to the input instruction.

2. A railway tunnel structure status remote intelligent monitoring system according to claim 1, characterized in that: The perception layer includes an image data acquisition module, a point cloud data acquisition module and a tunnel structure status data acquisition module; The image data acquisition module is used to acquire image data of the tunnel; The point cloud data acquisition module is used to collect point cloud data of the tunnel; The tunnel structure status data acquisition module is used to collect stress data, deformation data, vibration data and environmental data of the tunnel.

3. A railway tunnel structure status remote intelligent monitoring system according to claim 1, characterized in that: The processing layer includes a data pre-processing module, a data storage module, a risk estimation module and a guidance strategy generation module; The data pre-processing module is used to pre-process the relevant information of the tunnel structure and obtain pre-processed data; The data storage module is used to classify and store the pre-processed data; The risk estimation module is used to generate the simulation data based on the pre-processed data input into the tunnel simulation model, and generate the risk category and risk level using the risk estimation model according to the simulation data and the pre-processed data; The guidance strategy generation module is used to call a pre-stored guidance strategy according to the risk category and the risk level.

4. A railway tunnel structure status remote intelligent monitoring system according to claim 3, characterized in that: The data pre-processing module includes a data cleaning submodule, a data noise reduction submodule, a data alignment submodule, and a data compression submodule; The data cleaning submodule is used to remove abnormal values ​​and supplement missing values ​​from the relevant information of the tunnel structure to generate cleaned data; The data denoising submodule is used to perform smoothing and / or multi-scale decomposition on the cleaned data, filter out noise, and generate denoised data; The data alignment submodule is used to perform spatiotemporal alignment and time synchronization on the noise reduction data to generate aligned data; The data compression submodule is used to analyze and extract features from the alignment data, compress the extracted features, and obtain preprocessed data.

5. A railway tunnel structure status remote intelligent monitoring system according to claim 1, characterized in that: The processing layer also includes a monitoring device self-correction module for performing life prediction and monitoring mode management on the monitoring device.

6. A railway tunnel structure status remote intelligent monitoring system according to claim 1, characterized in that: The application layer includes a command input module, a positioning module, a data display module, and a command generation module; The command input module is used to receive external control commands; The positioning module is used to provide location data according to the risk category and risk level; The data display module is used to display the relevant information of the tunnel structure, the risk category, the risk level, the location data and the guidance strategy; The command generation module is used to split the control instruction or the guidance strategy into execution commands.

7. A remote intelligent monitoring method for railway tunnel structural status, characterized in that: The specific steps are: Collecting relevant information of the tunnel structure, wherein the relevant information of the tunnel structure includes image data, point cloud data, tunnel structure status data and environmental data of the tunnel; Preprocessing the relevant information of the tunnel structure to obtain preprocessing data; Inputting the preprocessed data into a tunnel simulation model to generate simulation data, inputting the simulation data and the preprocessed data into a risk prediction model, and outputting a risk category and a risk level; Invoking a guidance strategy according to the risk category and the risk level; The guidance strategy is split into execution commands, and the execution commands are transmitted to the execution device.

8. A method for remote intelligent monitoring of railway tunnel structure status according to claim 7, characterized in that: The preprocessing steps are: Identify data types in the relevant information of the tunnel structure, the data types including image type, point cloud type and sensor type; The data of different data types are preprocessed according to the data type and a preset preprocessing method.

9. A method for remote intelligent monitoring of railway tunnel structure status according to claim 7, characterized in that: The steps to obtain risk categories and risk levels are: Inputting the preprocessed data into the tunnel simulation model to simulate different operation scenarios to obtain multiple simulation data; Performing local spatial features on the plurality of simulation data and the pre-processed data to generate a spatial feature map; Performing time feature extraction on the spatial feature map to obtain spatiotemporal features; The spatiotemporal features are classified and predicted to obtain the risk level and the risk type.

10. A method for remote intelligent monitoring of railway tunnel structure status according to claim 7, characterized in that: The steps to obtain the guidance strategy are: Build a plan library based on historical data; A corresponding plan is matched from a plan library according to the risk level and the risk type, and a guiding strategy is generated according to the plan.

Images