A tunnel safety monitoring and early warning management method and system

Abstract

The application provides a tunnel safety monitoring and early warning management method and system, and relates to the technical field of tunnel safety, comprising: a plurality of intelligent monitoring terminals connected on an RS485 bus, which collect multi-modal data of a tunnel environment, and perform abnormal environment alarm triggering analysis; when any intelligent monitoring terminal detects an alarm event, the alarm event is broadcast to other terminals through the RS485 bus, other terminals receiving the alarm event autonomously analyze diffusion probability atlas, calculate emergency evacuation routes, and control an escape indication device to display an evacuation direction; communication permission differentiation scheduling based on the RS485 bus is performed, and alarm event situation awareness based on multi-modal data dynamic updating is completed. The application solves the technical problem that the tunnel safety monitoring and early warning management of the prior art often adopts a centralized control mode, resulting in lag in information transmission and processing, a relatively long alarm response time, and the inability to timely and effectively dispose of emergencies.

Description

Technical Field

[0001] This invention relates to the field of tunnel safety technology, specifically to a tunnel safety monitoring, early warning, and management method and system. Background Technology

[0002] Because tunnels are often in enclosed environments and subject to various factors, safety accidents therein can cause significant losses. Therefore, tunnel safety monitoring and early warning are particularly important. Current tunnel safety monitoring and early warning management technologies often employ a centralized control model, processing all alarm events through a central control system. This approach leads to delays in information transmission and processing, especially in long tunnels or when there are high demands for multi-point monitoring. The long alarm response time hinders timely and effective emergency response. Furthermore, existing evacuation systems are typically based on fixed evacuation routes, neglecting the impact of environmental changes or emergencies on these routes. In the event of sudden disasters, gas leaks, fires, or other special circumstances, existing evacuation routes often become ineffective, jeopardizing personnel safety. Summary of the Invention

[0003] This application provides a tunnel safety monitoring and early warning management method and system, which aims to solve the technical problems that the existing tunnel safety monitoring and early warning management often adopts a centralized control mode, resulting in lag in information transmission and processing, long alarm response time, and inability to carry out timely and effective emergency response.

[0004] The first aspect of this application discloses a tunnel safety monitoring and early warning management method, the method comprising: multiple intelligent monitoring terminals connected to an RS485 bus to collect multimodal data of the tunnel environment, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video; and to perform environmental anomaly alarm triggering analysis based on the multimodal data; when any intelligent monitoring terminal detects an alarm event, it broadcasts the alarm event to other terminals via the RS485 bus; the other terminals receiving the alarm event autonomously analyze the diffusion probability spectrum based on the relative relationship between their own location and the location of the alarm event, calculate emergency evacuation routes, and control an integrated escape indicator device to display the corresponding evacuation direction; and, based on the safety attributes of the alarm event and the diffusion probability spectrum, to perform differentiated communication permission scheduling based on the RS485 bus for the multiple intelligent monitoring terminals, thereby completing alarm event situational awareness based on dynamic updates of multimodal data.

[0005] The second aspect disclosed in this application provides a tunnel safety monitoring and early warning management system. The system is used for the above-mentioned tunnel safety monitoring and early warning management method. The system includes: a multi-modal data acquisition module, which is used to collect multi-modal data of the tunnel environment, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video, from multiple intelligent monitoring terminals connected to the RS485 bus, and trigger analysis of environmental abnormality alarms based on the multi-modal data; an evacuation direction display module, which is used to broadcast an alarm event to other terminals through the RS485 bus when any intelligent monitoring terminal detects an alarm event. Other terminals receiving the alarm event independently analyze the diffusion probability map based on the relative relationship between their own positions and the position of the alarm event, calculate the emergency evacuation route, and control the integrated escape indicator device to display the corresponding evacuation direction; an alarm event situation awareness module, which is used to perform differential scheduling of communication permissions based on the RS485 bus for the multiple intelligent monitoring terminals according to the safety attributes of the alarm event and the diffusion probability map, and complete the situation awareness of the alarm event dynamically updated based on multi-modal data.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] Through multiple intelligent monitoring terminals connected to the RS485 bus, comprehensive monitoring of the tunnel environment can be achieved, and various modal data can be collected. This multi-modal data collection method provides rich environmental information and can effectively comprehensively evaluate the safety status of the tunnel; when any intelligent monitoring terminal detects an alarm event, the alarm event is broadcast to other terminals through the RS485 bus, enabling full-network linkage to ensure that each monitoring point can obtain alarm information in real time and respond collaboratively. Other terminals receiving the alarm event independently analyze the diffusion probability map based on the relative relationship between their own positions and the position of the alarm event, and calculate the emergency evacuation route. This method can perform intelligent analysis and path planning based on the data of terminals in different positions, dynamically adjust the evacuation route, and ensure the safe evacuation of personnel. Each terminal independently analyzes and provides targeted evacuation instructions based on its own position and the position of the alarm event, reducing human intervention, improving the degree of automation, and the accuracy of the evacuation route; according to the safety attributes of the alarm event and the diffusion probability map, differential scheduling of communication permissions is implemented for different intelligent monitoring terminals. This mechanism optimizes the utilization of communication resources by reasonably scheduling the bus communication frequency and ensures that important information can be transmitted in a timely and accurate manner. Based on the dynamic update of multi-modal data and combined with the communication scheduling strategy, the situation awareness of the alarm event can be updated in real time, enabling the system to adjust the evacuation route or issue an alarm in a timely manner according to the spatio-temporal evolution, and reflecting the safety changes of the tunnel environment in real time.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of a tunnel safety monitoring and early warning management method provided in an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of a tunnel safety monitoring and early warning management system provided in an embodiment of this application.

[0011] Figure labeling: Multimodal data acquisition module 10, evacuation direction display module 20, alarm event situation awareness module 30. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown in the figure, this application provides a tunnel safety monitoring and early warning management method, the method including:

[0014] Multiple intelligent monitoring terminals connected to the RS485 bus collect multimodal data on the tunnel environment, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video. Based on the multimodal data, environmental anomaly alarm triggering analysis is performed.

[0015] Multiple intelligent monitoring terminals are installed inside the tunnel, distributed in different locations and connected to a unified monitoring system via an RS485 bus. Each intelligent monitoring terminal collects various types of data, including: gas: monitoring the concentration of gases in the tunnel, including H2S, CO, SO2, CO2, NO2, NH3, H2, CH4, N2, oxygen, etc.; dust: monitoring the concentration of fine particulate matter in the air to ensure air quality; wind speed: real-time monitoring of wind speed in the tunnel, especially under the operating conditions of the ventilation system; temperature and humidity: monitoring temperature and humidity data in the tunnel to promptly detect potential fire risks or mold growth; radioactive radiation: monitoring for the presence of abnormal radioactive materials in the tunnel; video surveillance: real-time monitoring of the tunnel's internal environment, equipment operation, and personnel activities via 360-degree panoramic cameras.

[0016] By performing real-time analysis of the collected multimodal data, abnormal situations can be identified. For example, trend analysis of environmental parameters based on historical data can identify abnormal fluctuations. Alarms can be triggered by setting thresholds, such as when the gas concentration exceeds safety standards or the temperature and humidity exceed the standards.

[0017] When any intelligent monitoring terminal detects an alarm event, it broadcasts the alarm event to other terminals via the RS485 bus. The other terminals that receive the alarm event autonomously analyze the diffusion probability spectrum based on the relative relationship between their own location and the location of the alarm event, calculate the emergency evacuation route, and control the integrated escape indicator device to display the corresponding evacuation direction.

[0018] When any intelligent monitoring terminal detects an anomaly and triggers an alarm, it broadcasts the alarm information to other intelligent monitoring terminals via the RS485 bus. Upon receiving the alarm information, all other intelligent monitoring terminals perform further analysis based on the received event content, including alarm type and location. Specifically, each intelligent monitoring terminal autonomously analyzes a diffusion probability map based on its location and the location relationship of the alarm event. The diffusion probability map reflects the environmental risk propagation at each location. For example, if a gas leak occurs at a tunnel entrance, the diffusion range of the gas leak is calculated based on wind speed and direction, providing a reference for emergency evacuation. Based on the diffusion probability map, the optimal emergency evacuation route is calculated and indicated by escape guidance devices, such as arrow indicator lights.

[0019] Based on the security attributes of the alarm event and the diffusion probability map, differentiated communication permission scheduling based on the RS485 bus is performed for the multiple intelligent monitoring terminals to complete the alarm event situational awareness based on dynamic updates of multimodal data.

[0020] The safety attributes of alarm events, such as fire and gas leak, are combined with the diffusion probability map to dynamically allocate different communication priorities to multiple intelligent monitoring terminals, including the core disaster group, diffusion monitoring group and safety background group. Based on the actual situation of the alarm event and the diffusion prediction, differentiated scheduling is performed through the RS485 bus to ensure that the most important environmental data can be obtained first in emergency situations, so as to carry out effective alarm event situational awareness.

[0021] Furthermore, in addition to the functions mentioned above, this method also integrates the following important functions: an emergency telephone, providing an instant communication channel between personnel inside the tunnel and the external rescue command center, ensuring rapid contact with rescue forces and timely transmission of on-site information in the event of an emergency; an emergency loudspeaker, which can guide the rapid and orderly evacuation of personnel inside the tunnel through sound broadcasts in emergency situations, playing a crucial guiding role, especially in environments with low visibility or poor audio-visual conditions; emergency lighting in case of power outages or power failures in the tunnel, ensuring that personnel can see clear escape routes during emergency evacuations and avoiding secondary disasters caused by power outages; and a remote power-off device / feeder circuit breaker, used to remotely control the power-off function of electrical equipment inside the tunnel, which can quickly cut off the power supply in the event of dangerous events such as fires or gas leaks, preventing further disasters caused by power failures. The integration of these functions comprehensively enhances the intelligence level of the tunnel safety monitoring and early warning system, providing more diversified guarantees for timely response to tunnel accidents and safe evacuation of personnel.

[0022] Furthermore, multiple intelligent monitoring terminals connected to the RS485 bus collect multimodal data on the tunnel environment, and perform environmental anomaly alarm triggering analysis based on the multimodal data, including:

[0023] Historical tunnel safety events are retrieved based on multimodal data types and categorized according to safety attributes to establish M safety event attributes and M associated data types. Based on the M associated data types, coupled alarm training is performed on the M safety event attributes to establish a multimodal coupled alarm analysis model. This model is then embedded into multiple intelligent monitoring terminals to perform alarm analysis on the multimodal data collected by the terminals. When an alarm is triggered, the safety attributes of the alarm event and the alarm trigger modal data are output.

[0024] Detailed data on historical tunnel safety incidents are collected through multimodal data retrieval, including various tunnel safety incidents such as fires, gas leaks, collapses, and explosions. In order to more effectively identify and handle different types of safety incidents, these historical tunnel safety incidents are first classified according to safety attributes, including event type, degree of danger, risk area, and triggering conditions. Each safety event attribute corresponds to a set of associated data types. For example, temperature and humidity data are associated with fire or equipment failure events. Through data mining and machine learning methods, M associated data types for M safety event attributes are established to ensure that each safety event attribute can be accurately associated with the corresponding multimodal data type.

[0025] By utilizing the relationships between M security event attributes and M related data types, historical data is used for training. The goal is to enable the model to identify which data patterns and combinations of changes can effectively predict different types of security events. The model can employ decision trees, among other methods. Based on the training results, a multi-modal coupled alarm analysis model is established. This model is used to analyze real-time collected data to identify potential security risks in advance.

[0026] The trained multimodal coupled alarm analysis model is embedded into multiple intelligent monitoring terminals. This allows each terminal to perform alarm analysis locally without relying on a central server for data processing, thus improving system response speed. The intelligent monitoring terminals input real-time collected multimodal data into the embedded multimodal coupled alarm analysis model. If certain data combinations exceed safety thresholds or abnormal patterns occur, an alarm is triggered. When an alarm is triggered, the safety attributes of the alarm event are output, including event type, hazard level, and risk area. In addition, the alarm event also includes alarm trigger modal data, which further indicates the root cause of the event. For example, in the case of a gas leak alarm, the trigger modal data includes gas concentration, wind speed, temperature, and humidity. This data allows for accurate determination of the alarm's source and potential impact range.

[0027] Furthermore, when collecting multimodal data about the tunnel environment, the multiple intelligent monitoring terminals have equal communication privileges on the RS485 bus.

[0028] Multiple intelligent monitoring terminals communicate via an RS485 bus. All terminals have equal communication rights, meaning each terminal can exchange data with other terminals and the control center, ensuring the system's real-time performance and reliability. Terminals do not rely on a central server to initiate data queries or alarm responses; each terminal has independent rights to exchange data and broadcast alarms. This decentralized design improves the system's fault tolerance. For example, when a terminal detects an alarm event, it can broadcast the event to other terminals via the RS485 bus, allowing them to analyze and respond more effectively based on local information. The RS485 bus communication protocol has high anti-interference capabilities and a long transmission distance, making it suitable for environments like tunnels. Equal communication rights also ensure that all terminals can exchange information smoothly within the network, avoiding the risk of single points of failure.

[0029] Furthermore, other terminals receiving the alarm event, based on the relative position of their own location and the location of the alarm event, autonomously analyze the diffusion probability map, calculate emergency evacuation routes, and control the integrated escape indicator device to display the corresponding evacuation direction, including:

[0030] The system reads the tunnel construction progress and analyzes the distribution locations of safety exits. Based on the relative positions of their own locations and alarm event locations, the multiple intelligent monitoring terminals first calculate the longitudinal evacuation direction along the tunnel axis pointing to the distribution locations of the safety exits. It then reads the safety attributes and alarm trigger mode data of the alarm events, combines them with real-time wind speed and direction data, identifies the risk diffusion probability, and establishes the diffusion probability map. Using the longitudinal evacuation direction as a constraint, it identifies progressively decreasing evacuation directions based on the diffusion probability map and establishes the emergency evacuation route. Finally, it generates corresponding control commands based on the emergency evacuation route, driving the arrow LEDs of the escape indicator device to light up in sequence, forming a light flow guide.

[0031] The tunnel construction progress information, including excavation progress and adit opening status, is read in real time through an integrated tunnel construction management system or other monitoring systems. During the tunnel construction phase, the layout of safety exits is planned in advance. Based on the construction progress, the location of completed or planned safety exits within the tunnel, including the main entrance and construction adits, is analyzed in real time. Safety exits are critical points for emergency evacuation, and their location information needs to be dynamically updated according to the tunnel structure. As construction progresses, emergency evacuation strategies need to be adjusted based on the different distributions of safety exits.

[0032] Each intelligent monitoring terminal has a clearly defined location within the tunnel, determined using a coordinate system and positioning module. When an alarm event occurs, it records the specific location of the alarm. Based on the relative position of its own location and the location of the alarm event, the terminal calculates and analyzes the longitudinal evacuation direction from the alarm location to the nearest safety exit. Specifically, since the tunnel is laid out along a certain axis, the evacuation direction generally follows the main axis of the tunnel. Each terminal calculates the distribution location of the nearest safety exit from the alarm event location along the longitudinal direction of the tunnel axis. During the calculation process, the geographical information and structure of the tunnel are taken into account to avoid calculating unreasonable evacuation routes.

[0033] Wind speed and direction are key factors affecting airflow and the spread of smoke or toxic gases within tunnels. Wind speed and direction data help predict the direction and extent of the spread of hazardous substances, such as toxic gases and smoke. Real-time acquisition of wind speed and direction data, combined with alarm event data, allows for the calculation of risk spread probability—the likelihood and direction of the spread of various hazardous substances—based on real-time wind speed, direction, the safety attributes of alarm events, and trigger modal data. A spread probability map is a visualization tool that shows the degree of risk spread in different areas; each area in the map displays the magnitude of the risk and potential danger zones.

[0034] When calculating evacuation routes, the longitudinal evacuation direction is used as a constraint, prioritizing evacuation paths along the tunnel's main axis to ensure safety and efficiency. When assessing risk based on the diffusion probability map, progressively decreasing evacuation directions are identified; that is, the probability of risk diffusion gradually decreases as evacuation distance increases. These evacuation directions are dynamically adjusted according to the risk level and diffusion probability to ensure that relatively safe paths are always prioritized. Based on the aforementioned longitudinal evacuation direction constraint and diffusion probability map, the optimal emergency evacuation route is calculated. This route covers the best path from the alarm location to the safe exit, ensuring that high-risk areas are avoided as much as possible during evacuation.

[0035] Based on the calculated emergency evacuation routes, corresponding control commands are generated. These commands activate escape guidance devices within the tunnel, such as arrow indicators and LED lights. Following the emergency evacuation routes, the control commands sequentially illuminate the arrow indicators at different locations, showing which direction people should evacuate. Typically, the lights illuminate in a specific order, gradually guiding the evacuation direction and ensuring a smooth and orderly retreat to the safe exit. The sequence of arrow indicator lights creates a light flow guide, providing an intuitive and visual evacuation route. People can quickly and clearly find the escape path by observing the lighting sequence of the arrow indicators. This light flow guide is not only practical but also effectively reduces evacuation chaos and improves evacuation efficiency in emergencies.

[0036] Furthermore, based on the security attributes of the alarm event and the diffusion probability map, differentiated communication permission scheduling based on the RS485 bus is performed for the multiple intelligent monitoring terminals, including:

[0037] Based on the security attributes of the alarm events and the diffusion probability map, the multiple intelligent monitoring terminals in the tunnel are dynamically divided into different communication priority groups, including a core disaster group, a diffusion monitoring group, and a security background group; differentiated scheduling of communication permissions is performed for the core disaster group, the diffusion monitoring group, and the security background group, and the polling scheduling strategy of the RS485 bus is adjusted; according to the polling scheduling strategy, the multiple intelligent monitoring terminals are controlled to complete alarm event situational awareness based on dynamic updates of multimodal data.

[0038] When an alarm event occurs, the severity, danger range, and potential risk areas of the event are assessed based on the event's safety attributes and diffusion probability map. This dynamically divides the multiple intelligent monitoring terminals within the tunnel into three different communication priority groups: the core disaster group, containing the most severely affected monitoring terminals located near the alarm event location, has the highest priority and requires rapid response and processing to obtain the most urgent real-time data; the diffusion monitoring group, with terminals located in the diffusion area of ​​the alarm event, is responsible for continuously monitoring the risk diffusion, such as gas concentration and temperature changes, and has a lower priority, requiring attention to risk changes and providing real-time data updates; the safety background group, with terminals located in relatively safe areas, primarily monitors the tunnel's routine environmental data and has the lowest priority, its task being to continuously monitor the environmental background and promptly identify any possible abnormal changes. This division is dynamic; that is, as the alarm event progresses and spreads, the terminal group division and priorities are adjusted in real time.

[0039] In the RS485 bus, communication between devices follows a scheduling strategy, employing a polling method. In this mode, multiple terminals take turns sending data to the bus, but different groups of terminals have different scheduling priorities. Terminals in the core disaster recovery group have the highest priority, receiving shorter polling intervals to ensure frequent reporting of emergency data and receipt of the latest instructions and information. Terminals in the diffusion monitoring group have the next highest priority; these terminals monitor the diffusion of risk areas, so their polling intervals are slightly longer, but still relatively frequent, to promptly capture and report diffusion conditions. Terminals in the security background group have the lowest priority, with even longer polling intervals, and can even switch to a mode that only actively reports when data changes exceed a threshold, reducing unnecessary communication burden. This differentiated scheduling ensures that the most critical monitoring data is reported first in emergencies, thus avoiding network congestion or information delays and improving the efficiency and response speed of the tunnel monitoring system.

[0040] According to the polling scheduling strategy, each terminal is directed to perform different tasks. Each terminal collects and uploads its corresponding multimodal data according to the assigned communication priority. Based on the latest data uploaded by the terminal, the situation of the alarm event is assessed in real time, such as the scope of risk spread, potential affected areas, whether further evacuation is needed, etc., and a corresponding emergency response plan is formed.

[0041] Furthermore, the polling scheduling strategy specifically includes: actively polling terminals within the core disaster recovery group at a first interval; periodically querying terminals within the diffusion monitoring group at a second interval, wherein the first interval is shorter than the second interval; and switching terminals within the security background group to a mode that actively reports only when data changes exceed a threshold.

[0042] For terminals within the core disaster recovery group, short-interval active polling is used to ensure these terminals can upload data more frequently. These terminals are located in the central area of ​​the incident and require rapid response and processing; therefore, their polling uses a very short first interval to ensure timely reporting of critical data and receipt of emergency instructions. For the diffusion monitoring group, a longer timed query interval, i.e., a second interval, is used. These terminals are located in the risk diffusion area, and their task is to continuously monitor environmental changes. Because the risk diffusion speed is slower, these terminals do not need to upload data as frequently as the core disaster recovery group; therefore, their polling interval is relatively longer, with the second interval being longer than the first interval. For the safety background group, these terminals are located in relatively safe areas of the tunnel and are only responsible for monitoring routine environmental parameters. To avoid unnecessary communication burden, the mode of these terminals is switched to actively report only when data changes exceed a threshold. That is, these terminals only actively report data when environmental data changes significantly; otherwise, they will be in silent mode to reduce network load.

[0043] Furthermore, according to the polling scheduling strategy, controlling the multiple intelligent monitoring terminals to complete alarm event situational awareness based on dynamic updates of multimodal data includes:

[0044] The system continuously aggregates multimodal data streams uploaded from intelligent monitoring terminals of the core disaster response group, the diffusion monitoring group, and the safety background group; performs spatiotemporal evolution analysis of alarm events based on the multimodal data streams to generate spatiotemporal safety distribution data; generates a dynamic isosurface model of alarm event diffusion based on the spatiotemporal safety distribution data through an interpolation algorithm; and overlays the dynamic isosurface model with a preset tunnel 3D geographic information model to visualize the risk prediction map.

[0045] Data is continuously collected from intelligent monitoring terminals in different groups. The terminals in the core disaster response group primarily focus on critical environmental data surrounding the incident site, providing the most urgent real-time monitoring data. The terminals in the spread monitoring group monitor environmental parameters in the event spread area, and the collected multimodal data helps predict the potential direction of event expansion. The terminals in the safety background group monitor routine environmental data from the tunnel, which helps maintain normal environmental monitoring and serves as a benchmark for changes in other areas. Data from each terminal is uploaded in streaming form, and the data update frequency depends on the communication priority of each terminal.

[0046] Spatiotemporal evolution analysis based on multimodal data streams means analyzing not only environmental data at each moment but also the trends of these data over time and their spatial distribution. For example, the occurrence of a gas leak event involves not only real-time data at the leak point but also the analysis of the gas diffusion speed, range, and its evolutionary pattern over time. Based on multimodal data streams, spatiotemporal safety distribution data is synthesized, including: the risk status of each area at each moment; the temporal progression of the event's spread; and the spatial distribution of various risk factors, such as the relationship between temperature, humidity, wind speed, and other data and the event location. This data enables dynamic tracking of events, real-time updates to risk assessments, and the generation of safety distribution maps reflecting the spatiotemporal dynamics of the event.

[0047] Interpolation algorithms can extrapolate the expected values ​​of collected discrete data points in unmonitored areas. For example, using inverse distance-weighted interpolation, points closer to the sensor are assigned higher weights to extrapolate values ​​at other locations. Through interpolation, isosurfaces are obtained, displaying different data levels during the event's spread, such as gas leaks of different concentrations. These isosurfaces can help analyze the event's spread trend and identify different risk areas. Because alarm events are dynamic, and the spread process changes over time, the generated isosurface model needs to be continuously updated to reflect the real-time changes in risk areas.

[0048] A 3D geographic information model (GIS) of a tunnel is a digital representation of its structure, including its spatial layout, exit locations, walls, supporting structures, and passageways. This 3D model provides the geographic context for risk prediction mapping. By overlaying a dynamic isosurface model with the tunnel's 3D GIS model, an integrated risk prediction map is formed. This map displays the risk level of each area within the tunnel, including real-time risk distribution and trends. The risk prediction map is displayed in a 3D visualization, using color, shading, and transparency to represent different risk levels; for example, red areas represent high-risk areas, yellow areas represent medium-risk areas, and green areas represent low-risk areas. Users can rotate, zoom, and pan the view in 3D space to examine the risk situation at different locations, helping emergency responders to promptly understand the safety situation within the tunnel.

[0049] Furthermore, after visualizing the risk prediction map, it also includes:

[0050] Based on the dynamic isosurface model and the risk prediction map, the effectiveness of the emergency evacuation route is automatically evaluated. If the emergency evacuation route is identified as invalid, an alternative optimized evacuation route is automatically calculated and generated. Through the RS485 bus, an instruction is sent to the intelligent monitoring terminal located on the emergency evacuation route to update the guidance direction of the escape indicator device in order to switch to the alternative optimized evacuation route.

[0051] Based on dynamic isosurface models and risk prediction maps, the effectiveness of currently established emergency evacuation routes is assessed. The dynamic isosurface model displays diffusion areas at different concentrations, helping to determine which areas' evacuation routes might be blocked or affected. The risk prediction map provides a visual representation of risk areas, helping to analyze whether current evacuation routes pass through high-risk areas or are potentially cut off. The effectiveness assessment includes: checking whether evacuation routes are within risk areas, especially high-risk areas such as toxic gas leaks or fires; determining whether existing evacuation routes can smoothly reach safe exits and whether they are affected by diffused harmful gases, fires, smoke, etc.; and automatically assessing whether there are unsuitable evacuation routes by comparing the risk data in the dynamic isosurface with the location of the evacuation routes.

[0052] If existing evacuation routes are deemed ineffective, for example, due to being covered by hazardous materials or cut off by fire, the system automatically calculates and generates alternative optimized evacuation routes. This means calculating the safest, shortest, and most effective alternative routes to avoid high-risk areas and ensure that people can evacuate safely and quickly.

[0053] New alternative optimized evacuation route guidance information is sent to each terminal via RS485 bus. Upon receiving the instruction, each intelligent monitoring terminal adjusts its escape indication device and updates the new evacuation direction, such as updating the direction of indicator lights or arrows, to help people quickly find the correct evacuation route. Through these updates, people in the tunnel can be accurately guided to evacuate along the most efficient evacuation route. The real-time changes in indicator lights or directional arrows not only improve evacuation efficiency but also ensure that people avoid dangerous areas in case of emergencies.

[0054] Furthermore, the multiple intelligent monitoring terminals broadcast their own status information on the RS485 bus according to a preset period, and listen to the status broadcasts of adjacent terminals. When the first terminal goes offline abnormally, the adjacent second terminal detects the abnormality and reports the fault information of the first terminal to the gateway through the RS485 bus.

[0055] Each intelligent monitoring terminal broadcasts its own status information on the RS485 bus at preset intervals, such as every 5 or 10 seconds. This includes: the sensor's operating status (e.g., whether it is working properly); currently monitored environmental data (e.g., gas concentration, temperature, humidity); and whether the equipment is malfunctioning or offline. This broadcast status information is monitored by neighboring terminals. This mechanism helps ensure cooperation between terminals and timely detection of any changes or faults in the tunnel environment. If an intelligent monitoring terminal malfunctions, such as going offline or failing, neighboring terminals will detect this anomaly and report the fault information to the gateway via the RS485 bus. This fault detection mechanism can quickly identify problems with the monitoring terminals, ensuring the efficient operation of the monitoring system.

[0056] Example 2, based on the same inventive concept as the tunnel safety monitoring and early warning management method in the foregoing examples, such as... Figure 2 As shown in the figure, this application provides a tunnel safety monitoring and early warning management system, the system comprising:

[0057] The multimodal data acquisition module 10 is used to acquire multimodal data of the tunnel environment from multiple intelligent monitoring terminals connected to the RS485 bus, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video. Based on the multimodal data, it performs environmental anomaly alarm triggering analysis. The evacuation direction display module 20 is used to broadcast the alarm event to other terminals via the RS485 bus when any intelligent monitoring terminal detects an alarm event. Other terminals receiving the alarm event autonomously analyze the diffusion probability map based on the relative relationship between their own position and the alarm event position, calculate emergency evacuation routes, and control the integrated escape indicator device to display the corresponding evacuation direction. The alarm event situation awareness module 30 is used to perform differentiated communication permission scheduling based on the RS485 bus for the multiple intelligent monitoring terminals according to the safety attributes of the alarm event and the diffusion probability map, and complete alarm event situation awareness based on dynamic updates of multimodal data.

[0058] Furthermore, the multimodal data acquisition module 10 is used to perform the following operation steps:

[0059] Historical tunnel safety events are retrieved based on multimodal data types and categorized according to safety attributes to establish M safety event attributes and M associated data types. Based on the M associated data types, coupled alarm training is performed on the M safety event attributes to establish a multimodal coupled alarm analysis model. This model is then embedded into multiple intelligent monitoring terminals to perform alarm analysis on the multimodal data collected by the terminals. When an alarm is triggered, the safety attributes of the alarm event and the alarm trigger modal data are output.

[0060] Furthermore, when collecting multimodal data about the tunnel environment, the multiple intelligent monitoring terminals have equal communication privileges on the RS485 bus.

[0061] Furthermore, the evacuation direction display module 20 is used to perform the following operation steps:

[0062] The system reads the tunnel construction progress and analyzes the distribution locations of safety exits. Based on the relative positions of their own locations and alarm event locations, the multiple intelligent monitoring terminals first calculate the longitudinal evacuation direction along the tunnel axis pointing to the distribution locations of the safety exits. It then reads the safety attributes and alarm trigger mode data of the alarm events, combines them with real-time wind speed and direction data, identifies the risk diffusion probability, and establishes the diffusion probability map. Using the longitudinal evacuation direction as a constraint, it identifies progressively decreasing evacuation directions based on the diffusion probability map and establishes the emergency evacuation route. Finally, it generates corresponding control commands based on the emergency evacuation route, driving the arrow LEDs of the escape indicator device to light up in sequence, forming a light flow guide.

[0063] Furthermore, the alarm event situation awareness module 30 is used to perform the following operation steps:

[0064] Based on the security attributes of the alarm events and the diffusion probability map, the multiple intelligent monitoring terminals in the tunnel are dynamically divided into different communication priority groups, including a core disaster group, a diffusion monitoring group, and a security background group; differentiated scheduling of communication permissions is performed for the core disaster group, the diffusion monitoring group, and the security background group, and the polling scheduling strategy of the RS485 bus is adjusted; according to the polling scheduling strategy, the multiple intelligent monitoring terminals are controlled to complete alarm event situational awareness based on dynamic updates of multimodal data.

[0065] Furthermore, the polling scheduling strategy specifically includes: actively polling terminals within the core disaster recovery group at a first interval; periodically querying terminals within the diffusion monitoring group at a second interval, wherein the first interval is shorter than the second interval; and switching terminals within the security background group to a mode that actively reports only when data changes exceed a threshold.

[0066] Furthermore, the alarm event situation awareness module 30 is used to perform the following operation steps:

[0067] The system continuously aggregates multimodal data streams uploaded from intelligent monitoring terminals of the core disaster response group, the diffusion monitoring group, and the safety background group; performs spatiotemporal evolution analysis of alarm events based on the multimodal data streams to generate spatiotemporal safety distribution data; generates a dynamic isosurface model of alarm event diffusion based on the spatiotemporal safety distribution data through an interpolation algorithm; and overlays the dynamic isosurface model with a preset tunnel 3D geographic information model to visualize the risk prediction map.

[0068] Furthermore, the alarm event situation awareness module 30 is used to perform the following operation steps:

[0069] Based on the dynamic isosurface model and the risk prediction map, the effectiveness of the emergency evacuation route is automatically evaluated. If the emergency evacuation route is identified as invalid, an alternative optimized evacuation route is automatically calculated and generated. Through the RS485 bus, an instruction is sent to the intelligent monitoring terminal located on the emergency evacuation route to update the guidance direction of the escape indicator device in order to switch to the alternative optimized evacuation route.

[0070] Furthermore, the multiple intelligent monitoring terminals broadcast their own status information on the RS485 bus according to a preset period, and listen to the status broadcasts of adjacent terminals. When the first terminal goes offline abnormally, the adjacent second terminal detects the abnormality and reports the fault information of the first terminal to the gateway through the RS485 bus.

[0071] Through the foregoing detailed description of a tunnel safety monitoring and early warning management method, those skilled in the art can clearly understand the tunnel safety monitoring and early warning management system in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0072] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for tunnel safety monitoring, early warning, and management, characterized in that, The method includes: Multiple intelligent monitoring terminals connected to the RS485 bus collect multimodal data on the tunnel environment, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video. Based on the multimodal data, environmental anomaly alarm triggering analysis is performed. When any intelligent monitoring terminal detects an alarm event, it broadcasts the alarm event to other terminals via the RS485 bus. Other terminals that receive the alarm event autonomously analyze the diffusion probability spectrum based on the relative relationship between their own location and the location of the alarm event, calculate the emergency evacuation route, and control the integrated escape indicator device to display the corresponding evacuation direction. Based on the security attributes of the alarm event and the diffusion probability map, for the multiple intelligent monitoring terminals, differentiated scheduling of communication permissions based on the RS485 bus is performed to complete the alarm event situational awareness based on dynamic updates of multimodal data. Based on the security attributes of the alarm event and the diffusion probability map, differentiated communication permission scheduling based on the RS485 bus is performed for the multiple intelligent monitoring terminals, including: Based on the security attributes of the alarm events and the diffusion probability map, the multiple intelligent monitoring terminals in the tunnel are dynamically divided into different communication priority groups, including the core disaster group, the diffusion monitoring group, and the security background group. Differentiated scheduling of communication permissions is implemented for the core disaster response group, the spread monitoring group, and the security background group, and the polling scheduling strategy of the RS485 bus is adjusted accordingly. According to the polling scheduling strategy, the multiple intelligent monitoring terminals are controlled to complete the situational awareness of alarm events based on dynamic updates of multimodal data. According to the polling scheduling strategy, the multiple intelligent monitoring terminals are controlled to complete the alarm event situation awareness based on dynamic updates of multimodal data, including: It continuously aggregates multimodal data streams uploaded from intelligent monitoring terminals of the core disaster response team, the spread monitoring team, and the security background team; Spatiotemporal evolution analysis of alarm events is performed based on multimodal data streams to generate spatiotemporal safety distribution data; Based on the aforementioned spatiotemporal security distribution data, a dynamic isosurface model for the propagation of alarm events is generated using an interpolation algorithm; The dynamic isosurface model is overlaid with a preset three-dimensional geographic information model of the tunnel to visualize the risk prediction map.

2. The tunnel safety monitoring and early warning management method as described in claim 1, characterized in that, Multiple intelligent monitoring terminals connected to an RS485 bus collect multimodal data about the tunnel environment. Based on this multimodal data, environmental anomaly alarm triggering analysis is performed, including: Historical tunnel safety events are retrieved based on multimodal data types, and extracted according to safety attributes, establishing M safety event attributes and M associated data types; Based on the M associated data types, coupled alarm training is performed on the M security event attributes to establish a multi-mode coupled alarm analysis model; The multimodal coupled alarm analysis model is embedded in the multiple intelligent monitoring terminals to perform alarm analysis on the multimodal data collected by the multiple intelligent monitoring terminals. When an alarm is triggered, the security attributes of the alarm event and the alarm trigger modal data are output.

3. The tunnel safety monitoring and early warning management method as described in claim 2, characterized in that, When collecting multimodal data on the tunnel environment, the multiple intelligent monitoring terminals have equal communication permissions on the RS485 bus.

4. The tunnel safety monitoring and early warning management method as described in claim 2, characterized in that, Other terminals receiving the alarm event, based on the relative relationship between their own location and the location of the alarm event, autonomously analyze the diffusion probability map, calculate emergency evacuation routes, and control the integrated escape indicator device to display the corresponding evacuation direction, including: Read the tunnel construction progress and analyze to obtain the distribution locations of safety exits; Based on the relative relationship between their own position and the location of the alarm event, the multiple intelligent monitoring terminals first calculate the longitudinal evacuation direction along the tunnel axis pointing to the distribution location of the safety exit; Read the security attributes and alarm trigger mode data of the alarm event, combine them with real-time wind speed and direction data, identify the risk spread probability, and establish the spread probability map. Using the longitudinal evacuation direction as a constraint, the emergency evacuation route is established based on the diffusion probability map to identify evacuation directions that decrease step by step. Based on the emergency evacuation route, corresponding control commands are generated to drive the arrow LEDs of the escape indicator device to light up in sequence, forming a light flow guide.

5. The tunnel safety monitoring and early warning management method as described in claim 1, characterized in that, The polling scheduling strategy specifically includes: actively polling terminals within the core disaster recovery group at a first interval; periodically polling terminals within the diffusion monitoring group at a second interval, wherein the first interval is shorter than the second interval; and switching terminals within the security background group to a mode that actively reports only when data changes exceed a threshold.

6. The tunnel safety monitoring and early warning management method as described in claim 1, characterized in that, Following the visualization of the risk prediction map, it also includes: Based on the dynamic isosurface model and the risk prediction map, the effectiveness of the emergency evacuation route is automatically evaluated. If the emergency evacuation route is identified as invalid, alternative optimized evacuation routes are automatically calculated and generated. The RS485 bus sends a command to the intelligent monitoring terminal located on the emergency evacuation route to update the direction of the escape indicator device, so as to switch to the alternative optimized evacuation route.

7. The tunnel safety monitoring and early warning management method as described in claim 1, characterized in that, The multiple intelligent monitoring terminals broadcast their own status information on the RS485 bus according to a preset period and listen to the status broadcasts of adjacent terminals. When the first terminal goes offline abnormally, the adjacent second terminal detects the abnormality and reports the fault information of the first terminal to the gateway through the RS485 bus.

8. A tunnel safety monitoring and early warning management system, characterized in that, The system is used to implement the tunnel safety monitoring and early warning management method according to any one of claims 1-7, the system comprising: The multimodal data acquisition module is used to acquire multimodal data of the tunnel environment from multiple intelligent monitoring terminals connected to the RS485 bus, including gas, dust, wind speed, temperature, humidity, radioactive rays, and video, and to perform environmental anomaly alarm triggering analysis based on the multimodal data; The evacuation direction display module is used to broadcast the alarm event to other terminals via the RS485 bus when any intelligent monitoring terminal detects an alarm event. Other terminals that receive the alarm event can autonomously analyze the diffusion probability spectrum based on the relative relationship between their own position and the location of the alarm event, calculate the emergency evacuation route, and control the integrated escape indicator device to display the corresponding evacuation direction. The alarm event situation awareness module is used to perform differentiated scheduling of communication permissions based on the RS485 bus for the multiple intelligent monitoring terminals according to the security attributes of the alarm event and the diffusion probability map, so as to complete the alarm event situation awareness based on dynamic updates of multimodal data.

Images