An Internet of Things-based railway tunnel environment monitoring system and method

Abstract

The present invention discloses an Internet of Things-based railway tunnel environment monitoring system and method, which relates to the technical field of data monitoring and management. The railway tunnel environment monitoring method analyzes historical monitoring data to evaluate the abnormal risk of the tunnel's dynamic environment, and can identify potential risks in advance; by training an abnormal prediction model, it can obtain risk prediction results in real time and issue early warnings; when new monitoring data is input, it captures abnormal risk parameters through automatic comparison and outputs warning information; by marking abnormal risks and evaluating risk levels, the severity of risks can be clearly identified, and based on the risk levels, the scope of influence and weights of abnormal risks are set, and multiple processing strategies are set to effectively respond to different degrees of abnormalities and improve the response efficiency. It realizes the automatic monitoring and intelligent management of railway tunnels, thereby effectively improving the monitoring accuracy and risk early warning ability of the tunnel's dynamic environment, ensuring the safe operation of the tunnel and reducing accidents.

Description

Technical Field

[0001] The present invention relates to the technical field of data monitoring and management, and particularly to a railway tunnel environment monitoring system and method based on the Internet of Things. Background Art

[0002] With the continuous advancement of current transportation construction, the number of long tunnels and extra-long tunnels is increasing. As the traffic volume in the tunnel increases, the concentration of tunnel pollutants also increases year by year, and the traditional tunnel operation control methods are difficult to meet the requirements. In order to save the operation cost of railway tunnels, it is necessary to collect the environmental parameters in the tunnels to achieve intelligent operation control of the tunnels. In addition, the collection and processing of railway tunnel environmental parameters are crucial for the safe operation of the tunnels and play an important role in realizing intelligent railway transportation and ensuring train operation safety.

[0003] During the operation of the public mobile network along the railway, as an important infrastructure in the railway tunnel environment monitoring, the stability of the power environment in the public network equipment chamber is directly related to the operation efficiency of the entire public mobile network along the line. Traditional monitoring methods mostly rely on manual inspections, which are not only inefficient but also difficult to achieve real-time monitoring of tunnel environmental parameters and rapid response to abnormal situations. With the development of technology, automated and intelligent monitoring systems have gradually become the mainstream of railway tunnel monitoring. The railway tunnel environment monitoring system based on the Internet of Things proposed by the present invention improves the monitoring efficiency and accuracy through centralized monitoring and analysis processing capabilities, and ensures the stable operation and safety of railway tunnels.

[0004] In summary, how to more accurately monitor the abnormal risk parameters of the railway tunnel power environment and respond to abnormal situations in a timely manner is an urgent problem to be solved and optimized for the railway tunnel environment monitoring system based on the Internet of Things. Summary of the Invention

[0005] The present invention provides a railway tunnel environment monitoring system and method based on the Internet of Things, which solves the technical problems of how to improve the monitoring efficiency and accuracy of the railway tunnel power environment and respond to and process abnormal situations in a timely manner.

[0006] To solve the above technical problems, the present invention provides a railway tunnel environment monitoring system and method based on the Internet of Things, and the specific technical solutions are as follows:

[0007] In a first aspect, a railway tunnel environment monitoring method based on the Internet of Things includes:

[0008] Obtaining historical monitoring data of a railway tunnel to obtain abnormal risk parameters of the power environment;

[0009] Wherein, based on the historical monitoring data, a variety of power environment parameters are obtained;

[0010] Set multiple abnormal risk indicators;

[0011] Compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters; the comparison difference parameters are the potential abnormal risk parameters;

[0012] Evaluate the potential abnormal risk parameters to obtain the abnormal risk parameters of the power environment; among them, when there are errors in the monitored potential abnormal risk parameter data, obtain the real abnormal risk parameters through the comparison difference parameters, and evaluate the abnormal risk degree according to the set threshold range;

[0013] Based on the abnormal risk parameters, mark abnormal risk marks respectively; evaluate the risk levels of the abnormal risk parameters with different abnormal risk marks; based on the evaluation results, obtain the abnormal level data of the abnormal risk parameters;

[0014] Among them, obtain the electromagnetic vector field through sensors and tunnel boundary conditions, construct a wave mode packet layer to obtain the charge density and current density on the dielectric interface, and obtain the electromagnetic wave propagation constant; when the propagation constant is close to 0 and the direction is perpendicular or parallel to the ground, determine that the tunnel propagation is abnormal and evaluate the power environment risk;

[0015] Construct a data set with historical monitoring data and multiple abnormal risk parameters to train and obtain an abnormal prediction model; obtain the abnormal risk prediction result of the tunnel power environment through the abnormal prediction model; according to the prediction result, respond to the preset warning threshold and output a warning message;

[0016] According to the prediction result and the warning message, determine the affected range of the abnormal risk parameters and obtain the abnormal risk weight; according to the abnormal risk weight, formulate multiple abnormal handling strategies.

[0017] As a further optimization scheme of the present invention, obtain the tunnel electromagnetic vector field through sensors combined with the boundary conditions of the tunnel; according to the obtained tunnel electromagnetic vector field, construct a wave mode packet layer for the railway tunnel; in the wave mode packet layer, the charge density p and the current density j are equal, that is ; where k represents the spatial frequency of the radio wave, represents the partial derivative symbol, and t represents the current time node;

[0018] According to the equality of the charge density p and the current density j, obtain the unit normal vectors of the two dielectric interfaces of the charge density and the current density; according to the unit normal vectors of the two dielectrics, obtain the free charge surface density and the conduction current surface density on the dielectric interface;

[0019] According to the free charge surface density P and the conduction current surface density J, through ; to obtain the propagation constant of electromagnetic waves in free space; where, k0 represents the propagation constant of electromagnetic waves in free space, represents the wavelength of electromagnetic waves, k P represents the spatial frequency of the free charge surface density, k J represents the spatial frequency of the conduction current surface density;

[0020] When electromagnetic waves are transmitted in free space, the propagation constant of the electromagnetic waves approaches 0, and it propagates approximately perpendicular to or parallel to the ground, then the electromagnetic waves propagate normally in the railway tunnel environment; otherwise, the normal propagation of electromagnetic waves in the railway tunnel environment is doubtful, which will cause the wireless channel state to be in a risk state; furthermore, obtain the abnormal risk parameters of the power environment.

[0021] As a further optimization scheme of the present invention, compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters, including:

[0022] The power environment parameters represent the parameter data of the operating state of the environment inside the railway tunnel; the abnormal risk indicators represent the risk standard values for the abnormality of the parameter data of the operating state of the environment inside the railway tunnel;

[0023] Preprocess multiple power environment parameters to obtain a standardized data set;

[0024] For the operation data of each time node of the standardized data set of the railway tunnel, through to obtain comparison difference parameters, and multiple comparison difference parameters constitute a difference degree index;

[0025] Where, Di represents the i-th comparison difference parameter, n represents the number of power environment parameter items, x i represents the i-th power environment parameter, y i represents the abnormal risk indicator corresponding to the i-th power environment parameter.

[0026] As a further optimization scheme of the present invention, evaluate the potential abnormal risk parameters to obtain the abnormal risk parameters of the power environment, including:

[0027] When comparison difference parameters are generated due to errors in the monitoring data, pass the obtained comparison difference parameters through to obtain the true abnormal risk parameters, and the true abnormal risk parameters are the abnormal risk parameters of the power environment; R t represents the true abnormal risk parameter, w i represents the true abnormal weight, n represents the number of comparison difference parameters; the comparison difference parameters are the potential abnormal risk parameters;

[0028] By setting the abnormal risk threshold range [T i , Tj , through to obtain the abnormal risk degree of the abnormal risk parameter of the power environment; where R t (T) represents the abnormal risk degree of the true abnormal risk parameter, K represents the abnormal risk degree evaluation weight, and t represents t of the true abnormal risk parameters;

[0029] When R t (T) < T i , then the abnormal risk degree of the abnormal risk parameter of the power environment is at low risk; when T i < R t (T) < T j , then the abnormal risk degree of the abnormal risk parameter of the power environment is at medium risk; when R t (T) > T j , then the abnormal risk degree of the abnormal risk parameter of the power environment is at high risk.

[0030] As a further optimized solution of the present invention, the abnormal prediction model includes:

[0031] Construct a data set with the historical acquired monitoring data and various abnormal risk parameters to generate structural data, and encode the structural data into sequence data to train the abnormal prediction model;

[0032] Input the sequence data into the abnormal prediction model; the abnormal prediction model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, and transmit the intermediate representation data of multiple hidden layers to the output layer, and the output layer outputs the prediction result representing the abnormal risk of the power environment of the railway tunnel;

[0033] Input at least one data item in the newly acquired railway tunnel monitoring data into the abnormal prediction model, and the output layer outputs the prediction result representing the abnormal risk of the power environment parameters in the newly acquired railway tunnel monitoring data.

[0034] As a further optimized solution of the present invention, in response to a preset warning threshold and output a warning message, including:

[0035] Based on the prediction result of the abnormal prediction model, and set a warning threshold corresponding to different abnormal risk degrees of the abnormal risk parameter of the power environment; design a corresponding alarm response mechanism based on the preset warning threshold to quickly respond to abnormal situations and issue a response alarm;

[0036] When in a low-risk abnormal situation, the alarm response mechanism flashes the alarm light once every preset time interval, and continuously monitors the abnormal risk state of the power environment parameters at the alarm light flashing node. When there is a risk rising trend in the low-risk abnormal situation at the t-th time node, the preset time interval of the alarm light flashing is shortened;

[0037] When in a medium-risk abnormal situation, the alarm response mechanism makes the alarm light flash without time intervals, accompanied by a siren sound at a preset time interval, records the abnormal risk parameter information at the current time node, and sets a preset time period to regularly evaluate the risk trend of the obtained abnormal risk parameter information to determine whether there is a risk of deterioration at the (t + 1)-th time node;

[0038] When in a high-risk abnormal situation, the alarm response mechanism issues a warning alarm through the alarm light and the siren sound without time intervals, sets the railway tunnel section in the high-risk abnormal situation, closes the corresponding railway tunnel section at the current time node, and conducts a secondary survey to avoid frequent occurrence of abnormal risks in multiple railway sections.

[0039] As a further optimized solution of the present invention, based on the prediction result and warning information of the abnormal prediction model, to obtain the abnormal risk affected range of the abnormal risk parameters, including:

[0040] Pass the real abnormal risk parameters through to obtain the abnormal risk affected metric value corresponding to the real abnormal risk parameters; In the formula, represents the abnormal risk affected metric value, represents the normal power environment parameter of the prediction result, represents the power environment parameter;

[0041] Establish a three-dimensional space coordinate in the railway tunnel, obtain any number of space coordinate points in each quadrant of the coordinate system, and pass the multiple space coordinate points through respectively to obtain the three-dimensional space abnormal risk affected range of the abnormal risk affected metric value;

[0042] In the formula, represents the abnormal risk affected range, K represents the spatial kernel function, N represents the number of space coordinate points, x r 、y r and z r respectively represent the three-dimensional coordinates of the r-th coordinate point.

[0043] As a further optimized solution of the present invention, the abnormal risk affected range of the abnormal risk parameters further includes:

[0044] When is abnormal, if and only if ; that is to obtain the abnormal risk affected time node range of the abnormal risk affected metric value; In the formula, represents the abnormal risk affected range at the t-th time node, represents the normal risk parameter threshold of the power environment; t i and tj respectively represent the i-th and j-th abnormal risk time nodes;

[0045] Based on the abnormal risk spreading range in the three-dimensional space and the abnormal risk spreading time node range, according to , to obtain the upper and lower limits of the abnormal risk spreading range, according to adjustment factor to dynamically adjust the upper and lower limits of the abnormal risk spreading range.

[0046] As a further optimization scheme of the present invention, according to the abnormal risk spreading range, to judge the abnormal risk weight of the abnormal risk parameter, and according to the abnormal risk weight corresponding to the dynamic environment abnormal risk parameter, set a variety of abnormal handling strategies, including:

[0047] Based on the spreading range and the severity of each risk factor, through assign abnormal risk weights to the severities of different risk factors; in the formula, represents the abnormal risk weight of the abnormal risk parameter pi, represents the abnormal empowerment function;

[0048] According to the abnormal risk weight, abnormal risk spreading range and abnormal risk parameter level, to plan abnormal handling strategies;

[0049] The abnormal handling strategies include that for low risk levels, measures such as monitoring, recording and warning can be taken without immediate intervention; for medium risk levels, increase manual monitoring, adjust system parameters or perform local repairs; for high risk levels, start an emergency response mechanism for global intervention, shutdown or isolation of the fault area.

[0050] In a second aspect, the system is provided with an electronic device including a memory, a processor, and a program of a railway tunnel environment monitoring method based on the Internet of Things stored on the memory and executable on the processor. When the program of the railway tunnel environment monitoring method based on the Internet of Things is executed by the processor, the steps of a railway tunnel environment monitoring method based on the Internet of Things are implemented. The system includes:

[0051] Data acquisition module: It is used to obtain historical monitoring data of the railway tunnel to obtain dynamic environment abnormal risk parameters;

[0052] Data recognition module: It is used to obtain various power environment parameters based on historical monitoring data; set various abnormal risk indicators; compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters; the comparison difference parameters are potential abnormal risk parameters; evaluate the potential abnormal risk parameters to obtain power environment abnormal risk parameters; based on the obtained power environment abnormal risk parameters, mark abnormal risk marks respectively; evaluate the risk levels of abnormal risk parameters with different abnormal risk marks; based on the evaluation results, obtain the abnormal level data of the abnormal risk parameters.

[0053] Early warning response module: It is used to construct a data set with historical monitoring data and various abnormal risk parameters to train an abnormal prediction model; obtain new monitoring data, input the new monitoring data into the abnormal prediction model to obtain the prediction result of the tunnel power environment abnormal risk; according to the prediction result of the tunnel power environment abnormal risk, respond to a preset early warning threshold and output early warning information.

[0054] Abnormal handling module: It is used to obtain the scope of influence of the abnormal risk of the abnormal risk parameters based on the prediction result and early warning information of the abnormal prediction model; judge the abnormal risk weight of the abnormal risk parameters according to the scope of influence of the abnormal risk; set various abnormal handling strategies according to the abnormal risk weight corresponding to the power environment abnormal risk parameters.

[0055] The present invention has at least the following beneficial effects: By analyzing historical monitoring data, the present invention can accurately extract potential abnormal risk parameters and evaluate the abnormal risk of the tunnel power environment based on these parameters, which helps to identify possible risks in advance.

[0056] By training the abnormal prediction model, the prediction result of the tunnel power environment abnormal risk can be obtained in real time. When new monitoring data is input, the system can automatically compare and output early warning information, so as to achieve early warning and dynamic monitoring and reduce the possibility of sudden events.

[0057] Marking and evaluating the risk level of abnormal risks can help to more clearly identify the severity of different risks. Determining different scopes of influence and risk weights of abnormal risks according to the risk level provides a quantitative basis for subsequent risk management and handling.

[0058] According to the weight corresponding to the abnormal risk parameters, various targeted abnormal handling strategies can be set to ensure that preventive measures are taken for different degrees of abnormalities and improve the response efficiency.

[0059] By establishing an anomaly prediction system based on historical data and models, the automated monitoring and intelligent management of railway tunnels are realized, the decision-making process is optimized, and the management efficiency and safety are improved. Using this method can effectively improve the monitoring accuracy of the tunnel dynamic environment, enhance the risk early warning ability, provide a scientific basis for tunnel management, reduce the possibility of accidents, and ensure the safe operation of railway tunnels. Description of the Drawings

[0060] Figure 1 It is a schematic flow chart of a method for monitoring the environment of a railway tunnel based on the Internet of Things provided by an embodiment of the present invention;

[0061] Figure 2 It is a schematic diagram of a system for monitoring the environment of a railway tunnel based on the Internet of Things provided by an embodiment of the present invention;

[0062] Figure 3 It is a model diagram of a monitoring point of a railway tunnel provided by an embodiment of the present invention;

[0063] Figure 4 It is a schematic connection diagram of a monitoring device and a monitoring host provided by an embodiment of the present invention;

[0064] Figure 5 It is a processing flow when an abnormal situation is detected by a monitoring platform provided by an embodiment of the present invention. Detailed Embodiments

[0065] The following further describes the present application in detail with reference to the drawings. It is necessary to point out here that the following detailed embodiments are only used to further illustrate the present application and should not be construed as limiting the protection scope of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.

[0066] A system and method for monitoring the environment of a railway tunnel based on the Internet of Things provided by this embodiment are specifically implemented as follows:

[0067] As Figure 1 shown, a method for monitoring the environment of a railway tunnel based on the Internet of Things includes the following steps:

[0068] Step 11, obtaining historical monitoring data of a railway tunnel to obtain dynamic environment anomaly risk parameters;

[0069] Step 12, based on the historical monitoring data, obtaining various dynamic environment parameters;

[0070] Setting various anomaly risk indicators;

[0071] Comparing the dynamic environment parameters with the anomaly risk indicators to obtain comparison difference parameters; the comparison difference parameters are the potential anomaly risk parameters;

[0072] Evaluate the potential abnormal risk parameters to obtain power environment abnormal risk parameters;

[0073] Based on the obtained power environment abnormal risk parameters, mark abnormal risk marks respectively;

[0074] Evaluate the risk levels of the abnormal risk parameters with different abnormal risk marks;

[0075] Based on the evaluation results, obtain the abnormal level data of the abnormal risk parameters;

[0076] Step 13: Construct a data set with the historically obtained monitoring data and various abnormal risk parameters to train an abnormal prediction model; obtain new monitoring data, input the new monitoring data into the abnormal prediction model to obtain the prediction result of the tunnel power environment abnormal risk; according to the prediction result of the tunnel power environment abnormal risk, respond to a preset warning threshold and output a warning message;

[0077] Step 14: Based on the prediction result and warning message of the abnormal prediction model, obtain the abnormal risk spreading range of the abnormal risk parameters; according to the abnormal risk spreading range, judge the abnormal risk weight of the abnormal risk parameters; according to the abnormal risk weight corresponding to the power environment abnormal risk parameters, set various abnormal handling strategies.

[0078] As Figure 4 shown, in step 11, install multiple environmental sensors, power consumption detection devices, perimeter intrusion alarms and image acquisition devices in the public network equipment chamber and equipment monitoring ports of the railway tunnel, so that the multiple monitoring devices can effectively cover all effective areas in the entire chamber;

[0079] The environmental sensors are used to monitor environmental parameters such as temperature, humidity, air pressure, and wind speed in the chamber; the power consumption detection device is used to monitor the power consumption of the public network equipment in the chamber; the perimeter intrusion alarm device is used to detect whether there is any illegal intrusion behavior around the chamber; the image acquisition device is used to monitor the condition in the chamber in real time and record video materials;

[0080] The environmental sensors include temperature and humidity sensors, water immersion sensors, intelligent air conditioner sensors, meteorological sensors, liquid level sensors, and smoke sensors. By monitoring various environmental parameters in the tunnel in real time, the environmental conditions in the chamber can be comprehensively grasped. The intelligent air conditioner sensor can automatically adjust the operating state of the air conditioner according to changes in temperature and humidity to maintain a constant temperature and humidity environment in the chamber cabinet. The meteorological sensor is used to monitor meteorological changes outside the chamber, such as wind speed, air pressure, etc., so as to adjust the environment in the chamber in a timely manner. The liquid level sensor is used to monitor the water accumulation situation in the chamber. Once an abnormality is detected, the drainage system can be immediately activated to prevent equipment damage. The smoke sensor is used to monitor in real time whether there is a fire risk in the chamber. Once smoke is detected, the alarm system will be immediately activated to ensure the safe evacuation of personnel.

[0081] The power consumption detection device includes power on / off monitoring, intelligent air switch monitoring, regulated power supply monitoring, power meter, intelligent electricity meter, and battery monitoring. The power on / off monitoring can monitor the current and voltage changes of the power supply system equipment in the chamber in real time to ensure stability and safety. The intelligent air switch monitoring device can monitor the on / off state of the circuit in real time. Once an abnormality is found, the power supply can be immediately cut off to prevent the expansion of circuit faults. The regulated power supply monitoring device ensures the stability of the supply voltage and avoids damage to equipment caused by voltage fluctuations. The power meter and intelligent electricity meter are used to accurately measure the power consumption, which is convenient for energy management and cost control. The battery monitoring device monitors the charge and discharge state of the battery in real time to ensure that the battery can provide a stable backup power supply in case of emergency. Through the comprehensive application of these power consumption detection devices, the centralized monitoring system for the power environment of the public network equipment chamber in the railway tunnel can effectively prevent and reduce power failures.

[0082] The perimeter intrusion alarm device can cover the entire perimeter area of the chamber to ensure the timely detection and response to any illegal intrusion behavior. The device usually includes door magnetic switches, audible and visual alarms, infrared microwave detection, infrared pair shooting, vibration detection, electronic fences, etc., which can detect different types of intrusion behaviors and send alarm information to the monitoring platform. In addition, the system may also be equipped with an access control system to further strengthen security protection. The access control system can record the information of the personnel entering and leaving to ensure that only authorized personnel can enter the chamber, thereby improving the overall security management level.

[0083] The monitoring range of the image acquisition device is the public network equipment chamber in the railway tunnel and the key areas where the equipment installed therein is located.

[0084] Connect all monitoring devices to the monitoring host through ports. The monitoring host, which is the IoT data collection gateway terminal, is responsible for collecting and processing the data sent up by on-site devices. This host has functions such as video and data collection, data processing, protocol conversion, linkage control, and alarm management. It supports wired networks and can still operate independently, store data, and perform linkage control when the network is interrupted. The data collection terminal uploads the collected data, processed results, alarm information, etc. to the monitoring platform device.

[0085] The monitoring host has powerful data processing capabilities and can perform real-time analysis on the collected environmental parameters, power consumption data, and image data to promptly detect abnormal situations. The monitoring host also has a remote control function and can automatically adjust the device operation status according to the monitoring data, such as adjusting the air conditioner temperature, starting the drainage system, etc., to adapt to environmental changes. In addition, the monitoring host can also transmit data and alarm information to the remote monitoring center in real time through the network to achieve remote monitoring and management.

[0086] Send the monitoring data obtained through multiple data interfaces to the monitoring platform via the front-end monitoring host. The monitoring platform is used to summarize the data sent up by the monitoring hosts in all chambers within the region to achieve data sharing and system integration. At the same time, it has an intelligent analysis function and can predict potential risks and issue early warnings based on historical data and real-time data.

[0087] The monitoring platform has a friendly user interface and can intuitively display the environmental conditions and device operation status in the chamber, facilitating operators to quickly understand the on-site situation. The monitoring platform also has data storage and historical data analysis functions and can record and analyze long-term monitoring data to provide a scientific basis for the maintenance and management of railway tunnels. Through data mining and pattern recognition technologies, the monitoring platform can identify abnormal change trends in environmental parameters to provide support for preventive maintenance.

[0088] In the implementation of the present invention, based on the power environment monitoring data obtained in step 11, preprocess a variety of power environment monitoring data to obtain standardized monitoring data;

[0089] Define abnormal risk indicators according to the preprocessed historical monitoring data and empirical values. For example:

[0090] Set the threshold range: based on statistical methods (such as the 3σ rule) or domain knowledge; define specific dangerous areas or fluctuation criteria; abnormal risk indicators can be divided into univariate and multivariate:

[0091] Univariate: independently evaluate based on each power parameter;

[0092] Multivariate: consider the interaction relationship of multiple power parameters.

[0093] Set a single-variable abnormal risk index T, through ; where μ represents the mean vector of the parameters, σ represents the standard deviation, and k represents a constant (usually 2 or 3);

[0094] The multi-variable index is obtained through ; where X represents the vector of power environment parameters,

[0095] Σ represents the covariance matrix; D 2 represents the multi-variable abnormal risk index;

[0096] Compare the power environment parameter vector X with the risk index T one by one; if X > T or X < T, record the abnormal difference; the comparison difference parameter is expressed as ΔX = X - T; if X exceeds the upper and lower threshold ranges, it is regarded as abnormal, and record the numerical value of the ΔX comparison difference parameter; use the comparison difference parameter ΔX of the previous step as the potential abnormal risk parameter; conduct preliminary screening and classification on these parameters to determine their sources.

[0097] Further analyze the potential abnormal risk parameters to determine their severity and scope of influence; use statistical analysis, machine learning, or model prediction methods to evaluate the possible consequences they may cause;

[0098] Set risk markers (such as low risk, medium risk, high risk) according to the power environment abnormal risk parameters; implement through rules or classification algorithms;

[0099] Based on the risk markers, conduct a specific evaluation of their levels; the evaluation can be based on multiple dimensions, such as: occurrence frequency, impact degree, controllability; according to the aforementioned evaluation, output the evaluation results for subsequent decision support or abnormal handling.

[0100] In step 13, obtain the preprocessed historical monitoring data of the power environment and various abnormal risk parameters through step 12, and establish a training data set containing input features and corresponding abnormal results.

[0101] Input features: including tunnel power environment parameters (such as vibration, acceleration, temperature, pressure, displacement, etc.); target variable (label): abnormal risk parameters (such as fluctuations exceeding the threshold, abnormal levels, potential risk markers); after the data set undergoes feature engineering (such as feature selection, dimensionality reduction, normalization), it is used to train a machine learning model.

[0102] Using historical data to reflect the actual situation, the model can predict future risks more accurately; combining multi-dimensional parameters and historical samples, the model can adapt to complex environments and predict more potential abnormalities; using RNN (Recurrent Neural Network) or LSTM (Long Short-Term Memory Network), which is suitable for time series data prediction;

[0103] Real-time monitoring data (such as the latest data obtained by sensors) is used as input and substituted into the trained anomaly prediction model for inference; the model outputs the anomaly risk prediction results, including: the likelihood of anomaly (such as probability value or classification result); the anomaly level (low risk, medium risk, high risk); through rapid prediction of new data, anomaly identification can be completed in a short time; the trained model is automatically updated according to dynamic monitoring data and continuously optimized; according to the prediction results, it is judged whether to trigger an early warning. Set multiple early warning thresholds: low-risk early warning: prompt potential risks, medium-risk early warning: suggest taking preventive measures, high-risk early warning: enforce emergency handling; if the prediction result exceeds the early warning threshold, output early warning information: including anomaly parameters, anomaly degree, and possible affected range; risk avoidance: take effective measures before the problem escalates; flexible response: different levels of early warnings allow hierarchical processing to improve the efficiency of resource allocation. According to the prediction results, combined with the distribution characteristics of anomaly parameters in space or time, estimate the affected range of the anomaly; judge the affected range through the model or rules:

[0104] Spatial range: the geographical area or equipment range affected by the anomaly parameter (such as the specific location of the tunnel);

[0105] Temporal range: the time window during which the anomaly may persist.

[0106] To be clearer about the affected range of the anomaly, avoid false alarms or missed alarms; help quickly locate the problem area and reduce the response time; combine information such as the affected range of the anomaly, anomaly level, and occurrence probability to calculate the anomaly risk weight: ; where W represents the anomaly risk weight, P represents the probability of anomaly occurrence (the probability value predicted by the model), S represents the affected range of the anomaly (such as the area of the spatial region or the number of affected devices), and I represents the degree of anomaly impact (such as possible economic losses or equipment damage degree); the higher the risk weight, the higher-priority processing strategy is required, and resources are allocated according to the weight to address the most critical risks; by introducing multiple factors through various risk weight calculations, the decision-making is more reasonable.

[0107] According to the risk weight, formulate a processing strategy, including: low risk: observation or lightweight processing (such as adjusting parameters), medium risk: preventive intervention (such as increasing the monitoring frequency, strengthening inspections), high risk: emergency handling (such as immediately stopping construction, activating the emergency plan); allocate limited resources to the most needed places; dynamically adjust the processing plan based on real-time prediction and weight.

[0108] The above steps cooperate with each other. By analyzing historical monitoring data, potential abnormal risk parameters can be accurately extracted, and the abnormal risk of the tunnel's dynamic environment can be evaluated, enabling the early identification of possible risks. By training an abnormal prediction model, risk prediction results can be obtained in real time and early warnings can be issued to reduce the occurrence of emergencies. When new monitoring data is input, the system can automatically compare and output warning information to achieve early warning and dynamic monitoring. By marking abnormal risks and evaluating risk levels, the severity of risks can be clearly identified, and the scope of influence and weight of abnormal risks can be set based on the risk level, providing a quantitative basis for subsequent management. In addition, according to the weight of abnormal risks, multiple processing strategies can be set to effectively respond to different degrees of abnormalities and improve response efficiency. This method realizes the automated monitoring and intelligent management of railway tunnels by establishing an abnormal prediction system for historical data and models, optimizes the decision-making process, improves management efficiency and safety, thereby effectively improving the monitoring accuracy and risk early warning ability of the tunnel's dynamic environment, ensuring the safe operation of the tunnel, and reducing the occurrence of accidents.

[0109] In another preferred embodiment of the present invention, it includes: combining sensors with the boundary conditions of the tunnel to obtain the electromagnetic vector field of the tunnel; constructing a wave mode packet layer for the railway tunnel according to the obtained electromagnetic vector field of the tunnel; in the wave mode packet layer, the charge density p and the current density j are equal, that is ; where k represents the spatial frequency of the electromagnetic wave, represents the symbol of partial derivative, and t represents the current time node;

[0110] According to the equality of the charge density p and the current density j, the unit normal vector of the two dielectric interfaces of the charge density and the current density is obtained; according to the unit normal vectors of the two dielectrics, the free charge surface density and the conduction current surface density on the dielectric interface are obtained;

[0111] According to the free charge surface density P and the conduction current surface density J, through ; to obtain the propagation constant of the electromagnetic wave in free space; where k0 represents the propagation constant of the electromagnetic wave in free space, represents the wavelength of the electromagnetic wave, k P represents the spatial frequency of the free charge surface density, k J represents the spatial frequency of the conduction current surface density;

[0112] When the electromagnetic wave is transmitted in free space, if the propagation constant of the electromagnetic wave approaches 0 and is approximately perpendicular to or parallel to the ground, the electromagnetic wave propagates normally in the railway tunnel environment; otherwise, the normal propagation of the electromagnetic wave in the railway tunnel environment is in doubt, which will cause the wireless channel state to be in a risk state; and then the abnormal risk parameters of the dynamic environment are obtained.

[0113] In the implementation of the present invention, sensors are deployed in the tunnel to monitor the electromagnetic field data in the tunnel in real time. These sensors can measure electromagnetic vector field components in the tunnel, including but not limited to the electric field and magnetic field. These data will be transmitted to the computing system for analysis. The electromagnetic field in the tunnel is affected by the geometric shape, material, and surrounding physical environment factors of the tunnel, so it can reflect the electromagnetic wave propagation characteristics of the tunnel.

[0114] Construct a wave mode envelope: Based on the obtained tunnel electromagnetic vector field, establish a wave mode envelope model. In this model, there is a balance relationship between the charge density and the current density, that is, the charge and the current are equal in the wave mode envelope model; especially during the propagation of electromagnetic waves, the distribution of charge and current will affect the propagation characteristics of the wave. This is based on the wave theory of electromagnetic waves, that is, the coupling relationship between the charge density and the current density reflects the dynamic characteristics of the electromagnetic field in the tunnel; according to the wave equation of the electromagnetic field, deal with the mutual relationship between the charge density and the current density in the model. This involves solving the propagation mode of electromagnetic waves in different media, and the equality relationship between the charge density and the current density provides the basis for the following derivation. Ensure the stability and regularity of the propagation of electromagnetic waves.

[0115] Utilize the relationship between the charge density and the current density to further determine the interfaces between two different media. The unit normal vector between these interfaces is very important for subsequent calculations. It can help to further describe the behavior of electromagnetic waves at the interfaces between different media, especially at the interfaces between the tunnel wall and the surrounding air and other different media; further through the unit normal vector, obtain the free charge surface density and the conduction current surface density on the interface. The free charge surface density is usually the number of free charges per unit area, indicating the distribution of charges on the surface of the medium; the conduction current surface density is the intensity of the flowing current per unit area, indicating the distribution of current on the surface. Reflects the distribution of electromagnetic waves inside and outside the tunnel and helps to describe the intensity and interaction of the electric field and magnetic field at the interface.

[0116] Utilize the spatial frequencies of the free charge surface density and the conduction current surface density. The spatial frequency usually refers to the frequency of periodic changes in space, indicating the rate of change of the wave mode in space. The spatial frequencies of the free charge surface density and the current surface density are related to the electromagnetic field distribution under the wave mode. Usually, they can be obtained by solving Maxwell's equations and performing Fourier transforms on the electric field and magnetic field; or the spatial distribution characteristics of the charge density and the current density can be derived using boundary conditions, the characteristics of the medium (such as dielectric constant and conductivity), and the wave equation, and then their spatial frequencies can be obtained;

[0117] By Obtain the propagation constant k of electromagnetic waves in free space; the propagation constant is a key parameter of the propagation characteristics of electromagnetic waves, which affects signal attenuation, reflection, and transmission efficiency; based on the propagation constant of electromagnetic waves, the propagation situation of electromagnetic waves in the tunnel can be judged. In the formula, f represents the frequency of electromagnetic waves, that is, the spatial frequency, and c represents the propagation rate of electromagnetic waves, represents the pi; when the propagation constant of electromagnetic waves in free space approaches zero, and the propagation direction of electromagnetic waves is close to perpendicular or parallel to the ground, this indicates that electromagnetic waves can propagate normally in the tunnel. If the propagation constant deviates from this characteristic, it means that there is an abnormality in the propagation of electromagnetic waves, which may lead to a decrease in the signal quality or an increase in interference of wireless communication; by analyzing the change of the propagation constant of electromagnetic waves, the propagation state of wireless signals in the tunnel can be further obtained. If the propagation constant of electromagnetic waves is abnormal, it means that there is an abnormality in the propagation of electromagnetic waves in the tunnel environment, thus entering the "risk state", which may lead to unstable wireless signals or communication interruption. At this time, the system will output "dynamic environment abnormal risk parameters" to prompt that preventive measures need to be taken, such as adjusting communication equipment, improving the tunnel structure, or strengthening signal enhancement, etc.

[0118] This method can timely identify the abnormal propagation of electromagnetic waves in the tunnel environment through real-time monitoring of the electromagnetic field in the tunnel by sensors, providing data support for the maintenance and optimization of communication equipment in the tunnel; by obtaining the propagation constant of electromagnetic waves, the state of the wireless channel can be evaluated. If the signal propagation is abnormal, it can be discovered and adjusted in advance, thereby improving the reliability of wireless communication in railway tunnels and ensuring the safety and smoothness of train operation; by monitoring the propagation situation of electromagnetic waves, problems such as signal attenuation, reflection, or interference can be predicted, and risk warnings can be issued in time to reduce the occurrence of wireless communication failures. Especially in scenarios with extremely high communication requirements such as high-speed railways, it can effectively reduce communication failures caused by electromagnetic interference; this technology can provide a scientific basis for the design, construction, and maintenance of tunnels. For example, factors such as the tunnel wall material, shape, and equipment layout can be optimized through this technology to improve the propagation characteristics of electromagnetic waves and enhance the communication quality in the tunnel; risk assessment of the electromagnetic environment in railway tunnels helps to improve the predictability of potential communication failures, strengthen the safety management of tunnel operation, and avoid safety accidents caused by communication interruption.

[0119] In a preferred embodiment of the present invention, in step 12, comparing the dynamic environment parameters with the abnormal risk indicators to obtain comparison difference parameters further includes:

[0120] Step 121, the dynamic environment parameters represent various parameter data of the operating state of the environment in the railway tunnel; the abnormal risk indicators represent the risk standard values of the occurrence of abnormalities in various parameter data of the operating state of the environment in the railway tunnel;

[0121] Step 122: Preprocess the multiple power environment parameters to obtain a standardized data set;

[0122] Step 123: For the operation data of each time node in the standardized data set of the railway tunnel, through , to obtain comparison difference parameters, and multiple said comparison difference parameters constitute a difference degree index;

[0123] In the formula, Di represents the i-th comparison difference parameter, n represents the number of power environment parameter items, and x i represents the i-th power environment parameter, and y i represents the abnormal risk index corresponding to the i-th power environment parameter.

[0124] In the implementation of the present invention, in Step 121, multiple key data indicators of the operating state of the tunnel environment are mainly defined, such as temperature, humidity, pressure, air flow velocity, air composition, etc. These parameters reflect the basic conditions of the tunnel environment and are the basis for evaluating the safe operation and operation efficiency of the tunnel.

[0125] The abnormal risk index sets some standardized risk values to indicate the possible abnormal risks that may be triggered when the power environment parameters deviate from the normal range. For example, when the temperature, humidity, etc. exceed the set safety thresholds, it may indicate problems with the ventilation, drainage, etc. facilities of the tunnel or potential safety hazards.

[0126] By clearly defining each power environment parameter and abnormal risk index, a unified standard can be provided for subsequent monitoring, analysis, and early warning. This enables tunnel management personnel to pay attention to the changes in environmental parameters in real time during operation and take corresponding measures in a timely manner to ensure the safe operation of the railway tunnel.

[0127] In Step 122, in order to ensure that the multiple power environment parameter data collected at different time nodes can be effectively compared and analyzed, it is necessary to standardize the data. The preprocessing may include steps such as noise removal, filling in missing data, and smoothing the data to remove interference factors and make the data more reliable. Standardization then converts the parameter data from different sources into the same scale so that they can be compared and calculated on the same platform. Common standardization methods include converting the data into a standard normal distribution with a mean of 0 and a variance of 1, or normalizing the data to a specific range (such as the [0,1] interval).

[0128] The beneficial effects of preprocessing and standardization are that they can eliminate the influence of data differences under different devices, different times, and different measurement conditions, and ensure the accuracy of subsequent analysis and model training. At the same time, the standardized data can improve the processing efficiency, facilitate automated analysis and comparison, and thus effectively improve the real-time performance and accuracy of risk assessment.

[0129] In the dataset standardized in Step 122, for the tunnel environment data at each time node, compare it with the previous historical data or standard operation data, and calculate the differences. These difference parameters are defined by comparing the deviation between the standardized data and the reference data; based on the calculated comparison difference parameters, construct a difference degree index. The difference degree index is a metric value used to comprehensively measure the change in the environmental state at different time nodes. If the value of the difference degree index is large, it means that the environment inside the tunnel has changed to a large extent, and there may be safety risks. On the contrary, if the value of the difference degree index is small, it indicates that the environmental change is small and the risk is low.

[0130] By comparing the difference parameters and constructing the difference degree index, it is possible to achieve real-time monitoring of the operating state of the tunnel environment and quickly identify possible abnormal changes. The difference degree index can be used as an early warning signal to reflect potential risks or hidden dangers in advance, and help management personnel take appropriate intervention measures. Based on the analysis of the difference degree index, it can also provide a scientific basis for decision-making and improve the safety and reliability of the tunnel.

[0131] The above steps cooperate with each other to achieve efficient monitoring, real-time analysis and risk early warning of the railway tunnel environment. Specifically: Step 121 ensures clear goals for data collection and monitoring by defining key environmental parameters and abnormal risk indicators; Step 122 ensures the quality and consistency of data through data preprocessing and standardization, making subsequent analysis more accurate; Step 123 provides an effective means for timely detecting abnormal changes and potential risks through comparison and calculation of the difference degree index; thus, it can improve the safety management level of the tunnel, prevent potential safety accidents, and improve the operating efficiency and reliability of the railway tunnel.

[0132] In a preferred embodiment of the present invention, in Step 12, the potential abnormal risk parameters are evaluated to obtain power environment abnormal risk parameters, and it further includes:

[0133] Step 124, when there are errors in the monitoring data resulting in comparison difference parameters, pass the obtained comparison difference parameters through , to obtain the true abnormal risk parameters, and the true abnormal risk parameters are the power environment abnormal risk parameters; R t represents the true abnormal risk parameter, w i represents the true abnormal weight, and n represents the number of comparison difference parameters; the comparison difference parameters are the potential abnormal risk parameters;

[0134] Step 125, by setting the abnormal risk threshold range [T i , T j , through to obtain the abnormal risk degree of the power environment abnormal risk parameters; in the formula, R t(T) represents the abnormal risk degree of the true abnormal risk parameter, K represents the evaluation weight of the abnormal risk degree, and t represents t of the true abnormal risk parameters;

[0135] Step 126, when R t (T) < T i , then the abnormal risk degree of the power environment abnormal risk parameter is at a low risk; when T i < R t (T) < T j , then the abnormal risk degree of the power environment abnormal risk parameter is at a medium risk; when R t (T) > T j , then the abnormal risk degree of the power environment abnormal risk parameter is at a high risk.

[0136] In the implementation of the present invention, in step 124, during the actual monitoring process, the collected monitoring data may have certain errors due to various factors (such as equipment errors, environmental interference, etc.), which will affect subsequent analysis and judgment. To ensure the accuracy of risk assessment, these data need to be corrected.

[0137] First, by comparing the tunnel environment parameters with historical data or standard operation data, the "comparison difference parameter" is calculated. These differences reflect the deviation between the monitoring data and the standard data, and may indicate potential abnormal risks; after error correction, the comparison difference parameter is further processed to obtain the true abnormal risk parameter. The true abnormal risk parameter represents the corrected risk value, which can more accurately reflect the abnormal risks in the tunnel environment; through weighted averaging or other methods of the comparison difference parameter, the true abnormal risk parameter is obtained; the data after error correction is more accurate: by correcting the errors of the monitoring data, the calculated true abnormal risk parameter is more credible, avoiding false alarms and misjudgments caused by errors, and ensuring the reliability of the risk assessment result; through the correction of the comparison difference parameter, the assessment of abnormal risks can be made more precise, reducing possible missed alarms and false alarms.

[0138] Abnormal risk threshold range [Ti, Tj] in Step 125: In this step, two thresholds are set to divide different levels of abnormal risk degrees. These thresholds are set based on factors such as historical data, safety standards of the tunnel environment, and risk tolerance. The division of the threshold range is usually based on experience, statistical analysis, or expert knowledge. By comparing the abnormal parameter values with the threshold range [Ti, Tj], the abnormal risk degree of the environment can be calculated. Through this calculation, the current risk level of the tunnel environment can be judged; by setting a clear risk threshold range, different risk levels can be accurately distinguished, enabling managers to better understand the current safety status of the tunnel and take appropriate countermeasures; the threshold setting can help quickly determine whether the tunnel environment requires special attention, ensuring timely response when potential risks are discovered and preventing accidents from occurring.

[0139] Division of low risk, medium risk, and high risk in Step 126, intuitive division of risk levels: By dividing the real abnormal risk parameters, the risk status of the tunnel can be intuitively obtained, facilitating managers to make quick decisions; this step helps identify potential dangerous areas and time periods, and through hierarchical management, it avoids being caught off guard when the risk is too high. The high-risk stage can trigger timely emergency handling to prevent the situation from deteriorating further; according to different risk levels, managers can allocate resources targeted, avoiding over-response or under-response. For example, for the low-risk state, only routine inspections may be required; for the high-risk state, emergency rescue forces need to be deployed.

[0140] In a preferred embodiment of the present invention, the abnormal prediction model in Step 13 further includes:

[0141] Step 131, constructing a data set from the historically obtained monitoring data and multiple abnormal risk parameters to generate structured data, and encoding the structured data into sequence data to train the abnormal prediction model;

[0142] Step 132, inputting the sequence data into the abnormal prediction model; the abnormal prediction model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer, and an output layer. Transmitting the intermediate representation data of multiple hidden layers to the output layer, and the output layer outputs the prediction result representing the abnormal risk of the dynamic environment of the railway tunnel;

[0143] Step 133, inputting at least one data item in the newly obtained railway tunnel monitoring data into the abnormal prediction model, and the output layer outputs the prediction result representing the abnormal risk of the dynamic environment parameters in the newly obtained railway tunnel monitoring data.

[0144] In the implementation of the present invention, step 131 extracts various types of data of the railway tunnel from the historical monitoring data, including dynamic environment parameters such as temperature, humidity, pressure, vibration, displacement, etc.; combines the obtained historical monitoring data with abnormal risk parameters (such as specific abnormal thresholds, environmental pressure indicators, etc.) to construct a comprehensive data set. This data set is the basis for training the abnormal prediction model.

[0145] The constructed data set is converted into structured data by a certain method (such as time series, feature engineering, etc.) and further encoded into sequence data. This step converts the original monitoring data into a format suitable for model processing, usually time series data, so that the model can capture patterns and changes in the time dimension.

[0146] Through data structuring and encoding, it is ensured that the data has been standardized and preprocessed before entering the model, avoiding interference from raw data noise; by combining historical monitoring data with abnormal risk parameters, the model can comprehensively capture potential abnormal risks in the tunnel dynamic environment, enhancing the accuracy of prediction.

[0147] Step 132 inputs data into the abnormal prediction model: The data encoded as sequence data above is input into the abnormal prediction model. This model is generally a deep neural network (such as a multi-layer perceptron, MLP), which includes multiple layers (such as an input layer, multiple hidden layers, and an output layer).

[0148] Model structure: Input layer: Receives the processed data (such as sequence data) as the input of the model.

[0149] Hidden layers (the first, second, and third layers): Multiple hidden layers are used to gradually extract high-level features from the data. Each layer learns more complex patterns than the previous layer, so as to be able to capture the temporal characteristics and complex non-linear relationships of the data.

[0150] Output layer: Finally, all intermediate representations (hidden layer outputs) are passed to the output layer, where the output layer will calculate an abnormal risk prediction result.

[0151] Through the design of the multi-layer neural network, the model can extract deep features from the data, learn complex and implicit rules, and improve the accuracy of abnormal risk prediction; Non-linear relationship processing: Multiple hidden layers can process non-linear relationships in the data, and these non-linear relationships are usually crucial for accurate risk prediction.

[0152] Step 133 Input new data: When new monitoring data (such as real-time data in the tunnel) is acquired, at least one data item (such as temperature, pressure, etc.) will be input into the already trained anomaly prediction model; Model output: The input new data will be processed through each layer of the model, and the final output layer will generate a new prediction result, representing the predicted value of the anomaly risk of the railway tunnel dynamic environment corresponding to the new data.

[0153] In this way, the model can receive new monitoring data in real time and make immediate predictions, providing dynamic anomaly risk warnings; Respond promptly to potential anomalies: Once new data is input, the system can respond promptly according to the prediction results of the model, take necessary preventive or treatment measures, and reduce potential risks; As the data is continuously updated, the model can continuously adapt to new data characteristics, improving its prediction accuracy over time.

[0154] As Figure 5 shown, in a preferred embodiment of the present invention, in step 13, in response to a preset warning threshold and outputting a warning message, it further includes:

[0155] Step 134, based on the prediction results of the anomaly prediction model, and setting warning thresholds corresponding to different anomaly risk levels according to the dynamic environment anomaly risk parameters; Designing a corresponding alarm response mechanism based on the preset warning threshold to quickly respond to abnormal situations and issue response alarms;

[0156] Step 135, when in a low-risk abnormal situation, the alarm response mechanism makes the alarm light flash once every preset time interval, and continuously monitors the abnormal risk status of the dynamic environment parameters at the alarm light flashing node. When there is an upward trend in the risk at the t-th time node of the low-risk abnormal situation, the preset time interval for the alarm light to flash is shortened;

[0157] Step 136, when in a medium-risk abnormal situation, the alarm response mechanism makes the alarm light flash without time intervals, accompanied by a siren sound at a preset time interval, records the abnormal risk parameter information at the current time node, and sets a preset time period to regularly evaluate the risk trend of the acquired abnormal risk parameter information to determine whether there is a risk of deterioration at the (t + 1)-th time node;

[0158] Step 137, when in a high-risk abnormal situation, the alarm response mechanism issues a warning alarm without time intervals through the alarm light and the siren sound, sets the railway tunnel section in the high-risk abnormal situation, closes the corresponding railway tunnel section at the current time node, and conducts a secondary survey to avoid frequent occurrence of abnormal risks in multiple railway sections.

[0159] Specifically, when the monitoring platform detects an abnormal situation, the attached abnormal handling unit can classify different types of abnormalities and take preventive measures, such as sending alerts, automatically adjusting the device operation status, or notifying maintenance personnel. In addition, the abnormal handling unit also has a self-diagnosis function, which can detect its own operation status to ensure timely response in case of abnormalities. By combining this centralized monitoring with abnormal handling, the safety and reliability of the power environment in the tunnel chamber of the public network equipment for railway tunnels can be effectively improved.

[0160] The abnormal handling unit includes an intelligent alarm system and a remote maintenance system. The intelligent alarm system can automatically select an appropriate alarm method, such as text message, phone call, email, etc., according to the type and severity of the abnormal situation, to ensure that relevant personnel can receive the alarm information in a timely manner. The remote maintenance system allows maintenance personnel to remotely log in to the monitoring platform to perform diagnostic and maintenance operations on the equipment, improving the maintenance efficiency and reducing the number and cost of on-site maintenance.

[0161] In the implementation of the present invention, step 134 can identify abnormal situations with different risk levels based on the prediction results of the abnormal prediction model. Corresponding warning thresholds are set according to different abnormal risk parameters. This threshold is used to determine when to trigger an alarm response; design an alarm response mechanism to ensure effective response and alarm under different risk levels; through the prediction results of the abnormal prediction model, the warning threshold can be dynamically adjusted to ensure that the alarm is triggered only when there is a real risk, thus avoiding excessive alarms; design a dedicated response mechanism for each risk level to enable timely and effective responses in different risk states.

[0162] In step 135, when the system detects a low-risk abnormal situation, the alarm light flashes once every set time and continuously monitors the abnormal situation. If the abnormal risk shows an upward trend at a certain time point, the time interval of the alarm light flashing is shortened to enhance the warning effect; only the alarm light flashing is used to remind at low risk, without overly disturbing the operator and maintaining the normal operation of the working environment;: as the abnormal risk increases, the alarm frequency speeds up, which helps to timely remind the operator to pay attention to potential problems and avoid the further expansion of risks.

[0163] In step 136, in the case of medium risk, the alarm response mechanism not only flashes the alarm light but also is accompanied by regular alarm sounds. At the same time, the system records the current abnormal risk parameter information and periodically evaluates the risk trend. By comparing the risk changes at the current and future time nodes, it is judged whether there is a risk of deterioration.; Through the cooperation of the alarm light and the alarm sound, ensure that the operator can timely discover the abnormal situation and take actions; by regularly evaluating the abnormal risk parameters, pre-judge the potential risk of deterioration, which helps to formulate countermeasures to prevent the spread of risks.

[0164] In step 137, in the case of a high-risk anomaly, the alarm light and siren will work continuously without interruption, emitting a strong warning signal. In addition, the system will immediately close the tunnel section according to the high-risk railway tunnel section information and conduct a secondary survey to prevent possible larger-scale risks; the continuously uninterrupted alarm light and siren can quickly attract the high attention of the operators, ensuring that no potential threats are missed in high-risk situations; closing the high-risk section and conducting a secondary survey helps to avoid catastrophic events, reduce personnel and equipment losses, and ensure railway safety.

[0165] In a preferred embodiment of the present invention, in step 14, based on the prediction result and warning information of the anomaly prediction model, to obtain the abnormal risk spread range of the abnormal risk parameter, further includes:

[0166] Step 141, passing the real abnormal risk parameter through to obtain the abnormal risk spread metric value corresponding to the real abnormal risk parameter; in the formula, represents the abnormal risk spread metric value, represents the normal power environment parameter of the prediction result, represents the power environment parameter;

[0167] Step 142, establish a three-dimensional space coordinate in the railway tunnel, obtain any number of space coordinate points in each quadrant of the coordinate system, and pass the multiple space coordinate points through respectively to obtain the three-dimensional space abnormal risk spread range of the abnormal risk spread metric value;

[0168] In the formula, represents the abnormal risk spread range, K represents the spatial kernel function, N represents the number of space coordinate points, x r , y r and z r respectively represent the three-dimensional coordinates of the r-th coordinate point.

[0169] In the implementation of the present invention, in step 141, the current actual abnormal risk parameters are first obtained. These parameters are obtained through the prediction model and are usually closely related to the state of the power environment (such as temperature, humidity, pressure, speed, etc.) and other relevant dynamic factors (such as vehicle state, operation of equipment in the tunnel, etc.); calculate the corresponding abnormal risk spread metric value using the real abnormal risk parameter. This metric value reflects the spread and influence degree of the abnormal risk in space, indicating the influence range of the abnormal event on the surrounding environment; estimate the severity of the abnormal environment parameter by obtaining the difference between the actual parameter and the predicted parameter; this metric value helps to quantify the risk and predict the influence range of the abnormal event.

[0170] By obtaining the abnormal risk spread measurement value, the propagation intensity and scope of abnormal events can be accurately evaluated, providing a scientific basis for subsequent emergency responses. As time goes by, the system can monitor the changes in risk parameters in real time and update the risk spread measurement value, thus providing dynamic risk assessment. By quantifying the abnormal risk spread measurement value, the possible consequences of abnormal situations can be understood more clearly, enabling reasonable decisions to be made.

[0171] In step 142, inside the railway tunnel, a three-dimensional space coordinate system is first established. This coordinate system is used to represent the positions of all spatial points inside the tunnel. Each spatial point can be identified by the coordinate point (x, y, z). Then, the three-dimensional space is divided into several quadrants (or regions) through the three-dimensional coordinate system. There can be multiple spatial coordinate points within each quadrant. These coordinate points represent points at different positions inside the tunnel, and each point has a potential abnormal risk spread measurement value. Based on the known abnormal risk spread measurement value, calculate the propagation scope of this measurement value in the three-dimensional space. The spatial expansion of abnormal risks can be predicted through a certain mathematical model or simulation method (such as diffusion model, thermodynamics model, etc.). This calculation will generate a three-dimensional space range, indicating the area that the risk may affect inside the tunnel.

[0172] By establishing a three-dimensional coordinate system, the abnormal risk status of each area inside the tunnel can be clearly known, which helps to locate the risk source and potential high-risk areas. Marking the abnormal risk spread range in the three-dimensional space can help operators and managers more intuitively understand the spatial distribution of risks, enabling timely decision-making for prevention. It does not rely solely on a single risk point but spatially models the risks throughout the tunnel, enhancing the comprehensive perception ability of risks in complex environments. Based on the calculation results of these spatial coordinate points and spread ranges, managers can scientifically evaluate the risk levels of each area and conduct targeted resource allocation and emergency responses according to the spatial distribution of risks. For example, if a certain area is greatly affected, it may need to be immediately closed for personnel evacuation or equipment inspection.

[0173] In a preferred embodiment of the present invention, the abnormal risk spread range of the abnormal risk parameters in step 14 further includes:

[0174] Step 143, when is abnormal, if and only if ; that is , to obtain the abnormal risk spread time node range of the abnormal risk spread measurement value. In the formula, represents the abnormal risk spread range at the t-th time node, represents the normal risk parameter threshold of the power environment; t i and t jrepresent the i-th and j-th abnormal risk time nodes respectively;

[0175] Step 144, based on the three-dimensional space abnormal risk spread range and the abnormal risk spread time node range, according to to obtain the upper and lower limits of the abnormal risk spread range, and according to adjustment factor to dynamically adjust the upper and lower limits of the abnormal risk spread range.

[0176] In the implementation of the present invention, the "abnormal" risk situation is defined in step 143. Generally, an abnormal risk event refers to a risk exceeding a certain preset threshold. This threshold is usually the normal risk parameter obtained through the analysis of historical data and the risk assessment model; the normal power environment risk parameter in the absence of abnormal events, and the risk parameter exceeding a certain set threshold is considered "abnormal".

[0177] In step 143, our goal is to determine the change of the "abnormal risk spread measurement value" at different time nodes, and the propagation range of the abnormal risk in the time dimension. When obtaining the normal risk parameter threshold of the abnormal risk spread range. That is, only when the abnormal risk spread measurement value exceeds this threshold will it be regarded as a real abnormal risk parameter.

[0178] Time node: By tracking the abnormal risk spread measurement value, the start and end time points of the abnormality are identified. On the time axis, that is, the i-th and j-th time nodes, which mark the time boundaries of the abnormal risk spread range.

[0179] The change in the time dimension reflects the start and end moments of the risk spread, thereby determining the time range of the abnormal risk spread. By analyzing the time node range of the abnormal risk, the specific moment when the abnormal event occurs can be accurately captured. This can more effectively conduct early warning and intervention; as time goes by, the system can dynamically adjust the risk assessment according to new data, thereby ensuring the timeliness and accuracy of the risk management strategy; by accurately judging the time node range of the risk spread, relevant management personnel can initiate emergency responses at critical time points, thereby reducing potential losses.

[0180] In step 144, the system no longer conducts risk assessment solely based on static spatial dimensions. Instead, it combines the interaction of time and spatial dimensions to dynamically adjust the upper and lower limits of the risk spread range. This means not only considering the range of abnormal events expanding in three-dimensional space but also the trend of the risk changing over time. In step 142, the abnormal risk spread range in three-dimensional space has been obtained, and now these ranges need to be adjusted according to the new time node and risk spread measurement values. For example, over time, the risk may expand or contract to other areas of the tunnel. Adjustment factors (such as environmental changes, changes in risk sources, etc.) will affect the upper and lower limits of the risk spread range.

[0181] Through the influence of the time dimension, the duration of abnormal risks can be identified. Combining this information, the spatial range can be dynamically adjusted. For example, if the abnormal risk increases within a certain time period, it may be necessary to expand the spatial range of risk prevention and control; if the abnormal risk weakens within another time period, the spread range can be reduced. The adjustment factor is a coefficient dynamically calculated based on variables such as the time, environment, and tunnel state of risk spread, and it is used to make real-time adjustments to the abnormal risk spread range. The adjustment factor is usually calculated based on the following factors: environmental changes (such as temperature, humidity, etc.); equipment status (such as the health status of monitoring equipment in the tunnel); changes in traffic flow and train speed; changes in risk sources (such as train failures, fires, etc.). By integrating these factors, the adjustment factor can reflect environmental changes and accordingly adjust the risk spread range in space.

[0182] By dynamically adjusting the upper and lower limits of the risk spread range through comprehensive analysis of time and space, the risk of a certain area in the tunnel can be more accurately evaluated to ensure appropriate countermeasures are taken. When an abnormality occurs in the tunnel, the system can dynamically adjust the coverage and response speed of emergency measures according to the latest risk spread range. For example, if the risk spread range of a certain area expands, it may be necessary to dispatch resources in advance for evacuation or strengthen safety monitoring. In the complex environment of the tunnel, the spread of risk may be affected by various factors. By dynamically adjusting the risk spread range, the system can respond more flexibly to emergencies and avoid response lags caused by overly static risk prediction. Through precise adjustment of the risk spread range, relevant departments can reasonably allocate resources and avoid waste of resources. For example, in time periods and areas with a smaller risk spread range, the input of emergency resources can be reduced, while in high-risk time periods, efforts can be concentrated on emergency response. In short, this process not only enhances the spatial and temporal understanding of abnormal risk events but also provides a flexible adjustment mechanism for actual operations, making the risk management of railway tunnels more intelligent and real-time.

[0183] In a preferred embodiment of the present invention, in step 14, according to the scope of influence of the abnormal risk, the abnormal risk weight of the abnormal risk parameter is judged, and according to the abnormal risk weight corresponding to the abnormal risk parameter of the power environment, a variety of abnormal handling strategies are set, and it further includes:

[0184] Step 145, based on the scope of influence and the severity of each risk factor, through assign abnormal risk weights to the severities of different risk factors; in the formula, represents the abnormal risk weight of the abnormal risk parameter pi, represents the abnormal empowerment function;

[0185] Step 146, according to the abnormal risk weight, the scope of abnormal risk influence and the abnormal risk parameter level, plan the abnormal handling strategy;

[0186] The abnormal handling strategy includes that for a low risk level, measures such as monitoring, recording and early warning can be taken, and immediate intervention is not required; for a medium risk level, increase manual monitoring, adjust system parameters or perform local repair; for a high risk level, start an emergency response mechanism, perform global intervention, shut down or isolate the faulty area;

[0187] In the implementation of the present invention, the scope of influence in step 145 refers to the breadth or degree to which the risk may affect the system. For example, a certain fault may only affect a single module, or may affect the entire system.

[0188] Evaluate the consequences when the risk occurs. For example, data loss, service interruption, system crash, etc.

[0189] The abnormal risk weight represents the weight or priority of each risk factor in the abnormal state. This weight will reflect the urgency of the risk factor and the need for priority handling. The calculation of the weight can depend on various factors, such as the scope of influence, severity, occurrence probability, etc.

[0190] By assigning weights to risk factors according to the scope of influence and severity, the risk management process can be made more detailed and scientific; this step helps to determine which abnormalities are the most urgent and which can be handled later, thus providing a clear basis for subsequent decision-making.

[0191] Step 146 plans the exception handling strategy based on the exception risk weight, the affected scope, and the exception risk parameter level assigned in Step 145; the exception risk parameter level refers to the risk level calculated based on the weight, such as low, medium, and high risk levels; different handling measures are formulated for different risk levels. Lower risks may not require immediate response, while higher risks may require urgent handling; by formulating different response strategies for different risk levels, resources can be rationally allocated to avoid overreaction or resource waste; according to different risk scenarios, appropriate response measures are selected to enable the system to flexibly respond to various emergencies.

[0192] Step 147 takes different exception handling measures for different risk levels according to the plan in Step 146: for low-risk exception events, measures such as monitoring, recording, and early warning are usually taken. Through real-time monitoring and recording, the evolution trend of the risk can be detected in a timely manner, but no immediate intervention measures are required.

[0193] Medium risk level: For the medium risk level, manual monitoring, system parameter adjustment, or partial repair are usually increased. At this time, the risk is relatively high, and some intervention measures are required, but a global response is not necessary; High risk level: For the high risk level, the system needs to activate the emergency response mechanism for global intervention, shutdown, or isolation of the faulty area to ensure the stability and security of the system.

[0194] Taking different response measures according to different risk levels can avoid overreaction and resource waste, and at the same time ensure that necessary emergency measures are taken in high-risk situations; through targeted handling of different risk levels, the normal operation and security of the system can be better guaranteed; monitoring and early warning can be carried out at low risk to detect potential problems in advance, and emergency measures can be started in a timely manner at high risk to prevent catastrophic consequences.

[0195] Such as Figure 2 shown, an Internet of Things-based railway tunnel environment monitoring system includes:

[0196] Data acquisition module: It is used to obtain the historical monitoring data of the railway tunnel to obtain the abnormal risk parameters of the power environment;

[0197] Data recognition module: It is used to obtain various power environment parameters based on historical monitoring data; set various abnormal risk indicators; compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters; the comparison difference parameters are potential abnormal risk parameters; evaluate the potential abnormal risk parameters to obtain power environment abnormal risk parameters; based on the obtained power environment abnormal risk parameters, mark abnormal risk marks respectively; evaluate the abnormal risk parameters with different abnormal risk marks to obtain the abnormal level data of the abnormal risk parameters.

[0198] Early warning response module: It is used to construct a data set with historical monitoring data and various abnormal risk parameters to train and obtain an abnormal prediction model; obtain new monitoring data, input the new monitoring data into the abnormal prediction model to obtain the abnormal risk prediction result of the tunnel power environment; according to the abnormal risk prediction result of the tunnel power environment, respond to a preset early warning threshold and output an early warning message.

[0199] Abnormal handling module: It is used to obtain the abnormal risk spread range of the abnormal risk parameters based on the prediction result and early warning information of the abnormal prediction model; judge the abnormal risk weight of the abnormal risk parameters according to the abnormal risk spread range; set various abnormal handling strategies according to the abnormal risk weight corresponding to the power environment abnormal risk parameters.

[0200] Such as Figure 3As shown, the railway tracks are displayed inside the tunnel, and multiple monitoring devices (such as sensors) are arranged along the tunnel, which may be used to monitor the structural health status of the tunnel, environmental parameters (such as humidity, temperature, smoke), or the train operation status; the monitoring devices are connected to the local monitoring site through wireless communication to form a regional monitoring network; the local monitoring site is located near the tunnel and collects data of the devices inside the tunnel through wireless signals; the local monitoring site processes or stores the monitoring data and communicates with a higher-level control center through the dedicated railway communication network; the functions of the local site may include: monitoring the tunnel operation status, initially analyzing the data, detecting anomalies, and uploading the data to the remote center; the dedicated railway communication network, the cloud-shaped area in the figure represents the dedicated railway communication network (such as railway optical fiber network, wireless communication network), which acts as a transmission bridge; through the dedicated railway communication network, the monitoring data is transmitted from the local site to a higher-level monitoring and control center to ensure the reliability and security of data transmission; the railway monitoring and dispatching center is the core node of the entire system, which is used to centrally process the tunnel monitoring data and perform high-level analysis and decision-making. It may have the following functions: remotely and real-time monitoring the tunnel and train operation conditions, cross-regional abnormal event analysis (such as geological disasters, equipment failures, fires, etc.), and issuing control or dispatching commands to the local monitoring sites on-site; data flow and communication, the yellow lightning-shaped arrows in the figure represent the transmission paths of data and commands. The data is collected from the monitoring devices inside the tunnel, transmitted to the local monitoring site, and then sent to the railway monitoring and dispatching center through the dedicated railway communication network; the commands and instructions of the monitoring and dispatching center can also be sent to the local monitoring site and on-site devices through the communication network.

[0201] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a well-known general intelligent device. Therefore, the object of the present invention can also be achieved only by providing a program product containing program code for implementing the method or system. It should also be noted that in the device and method of the present invention, obviously, each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present invention. And, the steps of performing the above series of processes can naturally be executed in chronological order according to the described order, but it is not necessary to be executed in chronological order. Some steps can be executed in parallel or independently of each other.

Claims

1. A method for monitoring the railway tunnel environment based on the Internet of Things, characterized in that, Including the following methods: Obtain the historical monitoring data of the railway tunnel to obtain the abnormal risk parameters of the power environment; Among them, based on the historical monitoring data, obtain various power environment parameters; Set various abnormal risk indicators; Compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters; The comparison difference parameter is the potential abnormal risk parameter; Evaluate the potential abnormal risk parameter to obtain the abnormal risk parameter of the power environment; Among them, when there are errors in the monitored potential abnormal risk parameter data, obtain the true abnormal risk parameter through the comparison difference parameter, and evaluate the abnormal risk degree according to the set threshold range; Based on the abnormal risk parameter, mark the abnormal risk mark respectively; evaluate the risk level of the abnormal risk parameter with different abnormal risk marks; based on the evaluation result, obtain the abnormal level data of the abnormal risk parameter; Among them, obtain the electromagnetic vector field through the sensor and the tunnel boundary conditions, construct the wave mode packet layer to obtain the charge density and current density on the medium interface, and obtain the electromagnetic wave propagation constant; when the propagation constant is 0 and the direction is perpendicular or parallel to the ground, determine that the tunnel propagation is abnormal and evaluate the power environment risk; Construct a data set with the historical monitoring data and various abnormal risk parameters to train and obtain an abnormal prediction model; obtain the abnormal risk prediction result of the tunnel power environment through the abnormal prediction model; according to the prediction result, respond to the preset warning threshold and output a warning message; According to the prediction result and the warning message, determine the affected range of the abnormal risk parameter and obtain the abnormal risk weight; according to the abnormal risk weight, formulate various abnormal handling strategies.

2. The method for monitoring the railway tunnel environment based on the Internet of Things according to claim 1, wherein Obtain the tunnel electromagnetic vector field through the sensor combined with the boundary conditions of the tunnel; according to the obtained tunnel electromagnetic vector field, construct a wave mode packet layer for the railway tunnel; in the wave mode packet layer, the charge density p and the current density j are equal to obtain the unit normal vectors of the charge density and the current density on the two medium interfaces; According to the unit normal vectors of the two media, obtain the free charge surface density and the conduction current surface density on the medium interface; according to the free charge surface density P and the conduction current surface density J, obtain the propagation constant of the electromagnetic wave in free space; When the electromagnetic wave is transmitted in free space, if the electromagnetic wave propagation constant is 0 and it propagates perpendicular to or parallel to the ground, the electromagnetic wave propagates normally in the railway tunnel environment; otherwise, the normal propagation of the electromagnetic wave in the railway tunnel environment is in doubt, resulting in the wireless channel state being in a risk state; furthermore, obtain the abnormal risk parameter of the power environment.

3. The method for monitoring the environment of a railway tunnel based on the Internet of Things according to claim 1, wherein, Compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters, including: The power environment parameters represent the parameter data of the operating state of the railway tunnel environment; the abnormal risk indicators represent the risk standard values for the abnormal occurrence of the parameter data of the operating state of the railway tunnel environment; Preprocess multiple power environment parameters to obtain a standardized data set; The operation data of each time node of the standardized data set of the railway tunnel is used to obtain comparison difference parameters, and multiple said comparison difference parameters constitute a difference degree index; Where Di represents the i-th comparison difference parameter, n represents the number of dynamic environment parameter items, x i represents the i-th dynamic environment parameter, and y i represents the abnormal risk index corresponding to the i-th dynamic environment parameter.

4. The method for monitoring the environment of a railway tunnel based on the Internet of Things according to claim 3, characterized in that, Evaluate the potential abnormal risk parameter to obtain the abnormal risk parameter of the power environment, including: When there are errors in the monitoring data and comparison difference parameters are generated, the obtained comparison difference parameters are passed through to obtain the true anomaly risk parameters, and the true anomaly risk parameters are the power environment anomaly risk parameters; R t represents the true anomaly risk parameter, w i represents the true anomaly weight, and n represents the number of comparison difference parameters; the comparison difference parameters are the potential anomaly risk parameters; By setting the abnormal risk threshold range [T i , T j , by to obtain the abnormal risk degree of the power environment abnormal risk parameter; where R t (T) represents the abnormal risk degree of the true abnormal risk parameter, K represents the abnormal risk degree evaluation weight, and t represents t of the true abnormal risk parameters; When R t (T) < T i , the abnormal risk degree of the abnormal risk parameter of the power environment is at low risk; when T i < R t (T) < T j , the abnormal risk degree of the abnormal risk parameter of the power environment is at medium risk; when R t (T) > T j , the abnormal risk degree of the abnormal risk parameter of the power environment is at high risk.

5. A method for monitoring the railway tunnel environment based on the Internet of Things according to claim 4, characterized in that, The abnormal prediction model includes: Generate structured data by constructing a dataset from historical monitored data and multiple anomaly risk parameters, encode the structured data into sequence data to train the anomaly prediction model; input the sequence data into the anomaly prediction model; the anomaly prediction model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer, and an output layer, transmit the intermediate representation data of multiple hidden layers to the output layer, and the output layer outputs the prediction result of the dynamic environment anomaly risk of the railway tunnel. Input at least one data item in the newly obtained railway tunnel monitored data into the anomaly prediction model, and the output layer outputs the prediction result of the anomaly risk of the dynamic environment parameters in the newly obtained railway tunnel monitored data. Based on the prediction result of the anomaly prediction model and the warning information, obtain the anomaly risk spread range of the anomaly risk parameters; according to the anomaly risk spread range, judge the anomaly risk weight of the anomaly risk parameters; according to the anomaly risk weight corresponding to the dynamic environment anomaly risk parameters, set multiple anomaly handling strategies.

6. The method for monitoring the railway tunnel environment based on the Internet of Things according to claim 5, wherein, Respond to a preset warning threshold and output warning information, including: Based on the prediction result of the anomaly prediction model, set a warning threshold corresponding to different anomaly risk levels of the dynamic environment anomaly risk parameters; design a corresponding alarm response mechanism based on the preset warning threshold to quickly respond to abnormal situations and issue corresponding alarms. When in a low-risk abnormal situation, the alarm response mechanism flashes the alarm light once every preset time interval, and continuously monitors the abnormal risk status of the dynamic environment parameters at the alarm light flashing node. When there is an upward risk trend in the low-risk abnormal situation at the t-th time node, shorten the preset time interval of the alarm light flashing. When in a medium-risk abnormal situation, the alarm response mechanism flashes the alarm light without time interval, accompanied by a siren sound at a preset time interval, records the abnormal risk parameter information at the current time node, and sets a preset time period to regularly evaluate the risk trend of the obtained abnormal risk parameter information to judge whether there is a deteriorating risk situation at the (t + 1)-th time node. When in a high-risk abnormal situation, the alarm response mechanism issues a warning alarm without time interval through the alarm light and the siren sound, set the railway tunnel section in the high-risk abnormal situation, close the corresponding railway tunnel section at the current time node, and conduct a second survey to avoid frequent occurrence of abnormal risks in multiple railway sections.

7. The method for monitoring the railway tunnel environment based on the Internet of Things according to claim 6, wherein Based on the prediction result of the anomaly prediction model and the warning information, obtain the anomaly risk spread range of the anomaly risk parameters, including: Using A(x i ) = ||x i - f θ (x i )|| to obtain the anomaly risk spread metric value corresponding to the true anomaly risk parameter; in the formula, A(x i ) represents the anomaly risk spread metric value, f θ (x i ) represents the normal power environment parameter of the prediction result, and x i represents the power environment parameter; Establish a three-dimensional space coordinate in a railway tunnel, obtain any number of spatial coordinate points in each quadrant of the coordinate system, and respectively pass the multiple spatial coordinate points through to obtain the three-dimensional space abnormal risk spread range of the abnormal risk spread measurement value; Wherein, R(x, y, z) represents the abnormal risk affected range, K represents the spatial kernel function, N represents the number of spatial coordinate points, x r , y r and z r respectively represent the three-dimensional coordinates of the r-th coordinate point.

8. The method for monitoring the railway tunnel environment based on the Internet of Things according to claim 7, characterized in that, The anomaly risk spread range of the anomaly risk parameters also includes: When x i is abnormal if and only if A(x i ) > τ; that is, W(t) = {t|A(x t ) > τ, t ∈ [t i , t j}, to obtain the abnormal risk propagation time node range of the abnormal risk propagation metric value; where W(t) represents the abnormal risk propagation range at the t-th time node, τ represents the normal risk parameter threshold of the power environment; t i and t j respectively represent the i-th and j-th abnormal risk time nodes; Based on the scope of the three-dimensional space abnormal risk impact and the range of the abnormal risk impact time nodes, according to [R(x, y, z), W(t)] = {x i | Max(α), Min(α)}, to obtain the upper and lower limits of the scope of the abnormal risk impact, and according to the α adjustment factor, to dynamically adjust the upper and lower limits of the scope of the abnormal risk impact.

9. The method for monitoring the railway tunnel environment based on the Internet of Things according to claim 8, characterized in that, According to the anomaly risk spread range, judge the anomaly risk weight of the anomaly risk parameters. According to the anomaly risk weight corresponding to the dynamic environment anomaly risk parameters, set multiple anomaly handling strategies, including: Based on the scope of influence and the severity of each risk factor, through W pi = f(A(x i ), R(x, y, z), W(t)) to assign corresponding abnormal risk weights to the severity of different risk factors; in the formula, W pi represents the abnormal risk weight of the abnormal risk parameter pi, and f represents the abnormal empowerment function; Plan the anomaly handling strategy according to the anomaly risk weight, anomaly risk spread range, and anomaly risk parameter level. The abnormal handling strategy includes, for low risk levels, taking monitoring, recording, and warning measures without the need for immediate intervention; for medium risk levels, increasing manual monitoring, adjusting system parameters, or performing partial repairs; for high risk levels, activating an emergency response mechanism for global intervention, shutdown, or isolation of the fault area.

10. A railway tunnel environment monitoring system based on the Internet of Things, characterized in that, The system is configured with an electronic device including a memory, a processor, and a program for an Internet of Things-based railway tunnel environment monitoring method stored on the memory and executable on the processor. When the program for the Internet of Things-based railway tunnel environment monitoring method is executed by the processor, it implements the steps of an Internet of Things-based railway tunnel environment monitoring method as described in any one of claims 1-9. The system includes: A data collection module: It is used to obtain historical monitoring data of railway tunnels to obtain power environment abnormal risk parameters. A data identification module: It is used to obtain various power environment parameters based on historical monitoring data; set abnormal risk indicators; compare the power environment parameters with the abnormal risk indicators to obtain comparison difference parameters; the comparison difference parameters are potential abnormal risk parameters; evaluate the potential abnormal risk parameters to obtain power environment abnormal risk parameters; based on the obtained power environment abnormal risk parameters, mark abnormal risk marks respectively; evaluate the abnormal risk parameters with different abnormal risk marks to obtain the abnormal level data of the abnormal risk parameters; based on the evaluation results, obtain the abnormal level data of the abnormal risk parameters. An early warning response module: It is used to construct a data set with the historical monitoring data and various abnormal risk parameters to train and obtain an abnormal prediction model; obtain new monitoring data, input the new monitoring data into the abnormal prediction model to obtain the prediction result of the abnormal risk of the tunnel power environment; according to the prediction result of the abnormal risk of the tunnel power environment, respond to a preset early warning threshold and output an early warning message. An abnormal handling module: It is used to obtain the scope of influence of the abnormal risk of the abnormal risk parameters based on the prediction result and early warning message of the abnormal prediction model; judge the abnormal risk weight of the abnormal risk parameters according to the scope of influence of the abnormal risk; set various abnormal handling strategies according to the abnormal risk weight corresponding to the power environment abnormal risk parameters.

Images