Railway tunnel disaster prevention lighting and evacuation indication internet of things networking scheduling method

Abstract

The application discloses a railway tunnel disaster prevention lighting and evacuation indication internet of things networking scheduling method and relates to the technical field of railway tunnel disaster prevention. The method comprises the following steps: S1, a tunnel end monitoring device initiatively initiates environment channel detection through a power line carrier communication network, scans impedance characteristics, noise levels and attenuation characteristics of different line sections in the tunnel, establishes a dynamic channel fingerprint library, and records an optimal communication frequency point and a channel quality threshold; and S2, a local center monitoring master station is linked with a fire alarm system in real time, and when a fire alarm signal is received, a dynamic evacuation path is generated through a disaster evolution prediction model according to a fire alarm position, a smoke diffusion state and a real-time channel state. The application actively performs channel detection and electrical parameter acquisition through a power line carrier communication module, establishes a dynamic channel fingerprint library, and performs fault diagnosis and state grading based on the dynamic channel fingerprint library, thereby improving the reliability and maintainability of the whole scheme and ensuring the availability of the equipment in a disaster.

Description

Technical Field

[0001] This invention relates to the field of railway tunnel disaster prevention technology, specifically to an Internet of Things (IoT) network scheduling method for railway tunnel disaster prevention lighting and evacuation guidance. Background Technology

[0002] Railway tunnels are enclosed environments with complex structures, making personnel evacuation and emergency rescue difficult during emergencies such as fires. Disaster prevention lighting and evacuation guidance systems are crucial for ensuring life safety. However, existing monitoring and intelligent management technologies have shortcomings: in terms of communication, wired methods are costly and inconvenient to deploy, while wireless methods have unstable signals; in terms of monitoring, it is difficult to accurately perceive equipment status in real time; in terms of scheduling and coordination, control commands and status feedback chains are disconnected, lacking a hierarchical collaborative mechanism, making it difficult to support efficient and dynamic emergency evacuation scheduling during disasters.

[0003] For example, patent publication number CN117119645A discloses a wireless sensing intelligent emergency lighting evacuation system, which consists of an intelligent central monitoring station, lighting fixtures, and indicator lights. This system monitors and collects fire information in real time through the lighting fixtures and indicator lights, issuing alarms and simultaneously transmitting the information to the central monitoring station. Through the collection, transmission, and conversion of node information, the system achieves dynamic monitoring, significantly shortening evacuation time and improving evacuation efficiency. It can automatically calculate the optimal escape route based on the fire situation, effectively solving the problem of traditional systems relying on fixed indicators and being unable to dynamically adjust escape routes.

[0004] However, the above and similar technical solutions still have the following shortcomings: lack of equipment health monitoring, which may lead to the ideal assumption that the overall reliability of the technical solution depends on the continuous integrity of the equipment when a disaster occurs; its decision-making and control are concentrated in the central monitoring station, lacking an intermediate layer with local decision-making and execution capabilities, and once communication with a certain area is interrupted, that area may be paralyzed; lack of dynamic perception and prediction capabilities of the spatial evolution of disasters, resulting in static and one-sided decision-making, making it difficult to effectively guide people to avoid the spreading fire and smoke. Summary of the Invention

[0005] The purpose of this invention is to provide an Internet of Things (IoT) network scheduling method for disaster prevention lighting and evacuation guidance in railway tunnels, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an Internet of Things (IoT) network scheduling method for disaster prevention lighting and evacuation guidance in railway tunnels, comprising:

[0007] S1. The tunnel end monitoring equipment actively initiates environmental channel detection through the power line carrier communication network, scans the impedance characteristics, noise level and attenuation characteristics of different line sections in the tunnel, establishes a dynamic channel fingerprint database, and records the preferred communication frequency and channel quality threshold.

[0008] S2. The central monitoring station of the bureau is linked with the fire alarm system in real time. When a fire alarm signal is received, it immediately generates a dynamic evacuation route based on the fire alarm location, smoke diffusion status and real-time channel status through the disaster evolution prediction model.

[0009] S3. Activate the emergency communication dispatch mechanism. The tunnel-end monitoring equipment executes cooperative optical flow guidance control based on dynamic evacuation paths through the power line carrier communication network.

[0010] S4. Real-time collection of electrical parameters and environmental data of each lighting and indicator light fixture location, combined with channel fingerprint database for fault diagnosis, and real-time reporting of lighting fixture status data, local ambient temperature and communication quality data to the central monitoring station.

[0011] S5. In emergency mode, communication dispatch adopts a hybrid multiple access mechanism that prioritizes fire events: the tunnel end monitoring equipment broadcasts instructions through a dynamic time division multiple access mechanism, and the lighting and indicator lights report critical statuses using a carrier sense multiple access mechanism based on fault status priority. Instructions can interrupt regular communication at any time to ensure that the instruction channel has priority.

[0012] Furthermore, in S2, dynamic evacuation routes are generated through a disaster evolution prediction model. Specific methods include:

[0013] Using the fire alarm location as the initial fire source, and combining real-time air volume and wind speed data, a real-time smoke diffusion simulation model based on computational fluid dynamics is used to predict the spatial distribution of smoke concentration and visibility attenuation characteristics within the first preset time period.

[0014] Based on the prediction results and combined with the channel fingerprint database, a map of areas where power line carrier communication quality deteriorates due to temperature and smoke particle adhesion during a disaster is predicted.

[0015] The tunnel topology is discretized into an evacuation network with each lighting and indicator light node as the basic unit, and each edge in the network is dynamically assigned a passage cost weight.

[0016] Using a preset path search algorithm, starting from the main distribution area of ​​trapped personnel or a preset fixed emergency assembly point, and ending at a safety exit or tunnel entrance, several candidate evacuation paths are calculated in the evacuation network.

[0017] The total travel cost, path overlap, and communication risk coefficient of each evacuation route are evaluated, and a weighted fusion is performed to obtain a comprehensive cost score for the route. Based on this score, the sequence of primary and backup evacuation routes is determined.

[0018] Furthermore, S3 specifically includes:

[0019] The tunnel-end monitoring equipment sends real-time control commands to the lighting and indicator lights in the relevant sections of the current main evacuation route, dynamically adjusting their light intensity, flashing frequency and direction of indication to form a guiding light flow;

[0020] The backup control commands corresponding to the backup evacuation routes are preloaded into the local cache of the lighting and indicator lights in the affected section;

[0021] When the actual disaster evolution is detected to reach the preset switching threshold, the tunnel end monitoring equipment sends a switching trigger command to the relevant lighting and indicator lights, immediately activating and executing the pre-loaded backup control command.

[0022] Furthermore, the electrical parameters include the active impedance spectrum, time-domain reflection waveform, operating current harmonic components, and leakage current value of the circuit; the environmental data includes the surface temperature of the lamp housing and the local ambient temperature gradient; the critical states include normal state, emergency state, and warning state; the fault diagnosis methods include:

[0023] The collected active detection impedance spectrum is convolved and compared with the health fingerprints in the channel fingerprint database under the same location and similar environmental conditions to calculate the feature difference degree. If the feature difference degree exceeds the threshold, the fault type is located by combining the current harmonic abrupt component and the abnormal reflection point of the time domain reflection waveform, and the state is marked as an emergency state.

[0024] Trend analysis is performed on the long-term time-series data of leakage current and the relationship curve between the casing temperature and the ambient temperature gradient. When the leakage current exceeds the safety threshold within the third preset time period or the slope of the casing temperature-ambient temperature gradient curve remains abnormal, the status is marked as an early warning status, and the remaining effective working time is estimated.

[0025] Furthermore, the hybrid multiple access mechanism in S5 includes:

[0026] In emergency mode, the tunnel-end monitoring equipment divides the communication cycle into instruction issuance time slots and status reporting time slots;

[0027] The instruction issuance time slot adopts a time division multiple access mechanism, and its time slot length is adjusted in real time according to the urgency of the broadcast instruction and the current broadcast area channel status evaluated in the channel fingerprint database;

[0028] The status reporting time slot adopts a carrier sense multiple access mechanism, and its time slot length is dynamically allocated according to the total number of lighting and indicator lights marked as warning status and emergency status in the network.

[0029] During the status reporting time slot, the carrier sense multiple access behavior of each lighting and indicator light fixture is controlled by the fault diagnosis status determined by S4. For lighting and indicator light fixtures in normal state, a standard random backoff contention mechanism is used for reporting; for lighting and indicator light fixtures in warning state, priority contention right is granted; and for lighting and indicator light fixtures in emergency state, the highest level of preemption right is granted.

[0030] Furthermore, when the lighting status data is reported, the status level identifier and physical location code corresponding to the lighting and indicator lights need to be embedded in the data frame.

[0031] Furthermore, the method also includes, in order to address the failure of monitoring equipment at the tunnel end, designing lighting or indicator lights that allow direct environmental perception capabilities, and upgrading them to temporary master nodes when preset conditions are met, organizing nearby lighting and indicator lights for emergency evacuation guidance.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] The IoT-based network scheduling method for disaster prevention lighting and evacuation instructions in railway tunnels solves the problems of unstable wireless signals, susceptibility to interference, and poor penetration in tunnels by directly utilizing existing power lines within the tunnel to build a power line carrier communication network. This provides a communication channel that eliminates the need for additional wiring, ensuring reliable transmission of emergency commands. By establishing a hybrid multiple access communication scheduling mechanism that prioritizes fire incidents, differentiated network resource access permissions are allocated to services with different levels of urgency. This effectively avoids network congestion and conflicts in high-concurrency emergency scenarios, ensuring the timely issuance of critical commands and the priority reporting of critical status information, thus guaranteeing the real-time and orderly nature of emergency response.

[0034] Meanwhile, by actively probing channels and collecting electrical parameters using power line carrier communication modules, a dynamic channel fingerprint database was established. Based on this database, fault diagnosis and status classification were performed, improving the overall reliability and maintainability of the solution and ensuring equipment availability during disasters. By constructing a hierarchical coordination architecture between the central monitoring station and tunnel-end monitoring equipment, and establishing a temporary master node self-organizing mechanism, the continuous service capability in the event of equipment failure was enhanced. The central monitoring station predicted the spatial evolution of the disaster using a real-time smoke diffusion simulation model based on computational fluid dynamics, and combined this with the dynamic channel fingerprint database to predict communication quality degradation. A dynamically weighted evacuation network was constructed, generating dynamic primary / backup evacuation paths to effectively guide personnel evacuation. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the Internet of Things (IoT) network scheduling method for railway tunnel disaster prevention lighting and evacuation indication according to the present invention.

[0036] Figure 2This is a schematic diagram of the dynamic evacuation path generation method of the present invention;

[0037] Figure 3 This is a schematic diagram of the optical flow guidance control performed by the tunnel-end monitoring equipment of the present invention;

[0038] Figure 4 This is a schematic diagram of the fault diagnosis method of the present invention;

[0039] Figure 5 This is a schematic diagram of the emergency response method for the failure of the tunnel end monitoring equipment according to the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] like Figure 1 As shown, this invention provides a technical solution: an IoT network scheduling method for disaster prevention lighting and evacuation guidance in railway tunnels. The method is based on a two-level monitoring architecture consisting of a central monitoring station and tunnel-end monitoring equipment, and utilizes existing power lines within the tunnel to establish a power line carrier communication network. The method includes:

[0042] S1. The tunnel end monitoring equipment actively initiates environmental channel detection through the power line carrier communication network, scans the impedance characteristics, noise level and attenuation characteristics of different line sections in the tunnel, establishes a dynamic channel fingerprint database, and records the preferred communication frequency and channel quality threshold.

[0043] This step is primarily executed by tunnel-end monitoring equipment deployed at both ends of the railway tunnel. Utilizing existing power lines within the tunnel, it proactively and periodically conducts comprehensive detection and feature analysis of the communication channels, establishing a dynamic channel fingerprint database that reflects changes in the physical environment and disaster conditions of the railway tunnel.

[0044] It is important to note that the tunnel-end monitoring equipment integrates a high-performance power line carrier (PLC) communication module, a main control unit, and a local storage unit. This equipment is located in the tunnel's power distribution box or monitoring chamber and is directly connected to the tunnel's main power line.

[0045] The detection is conducted using the existing power line network within the tunnel that supplies power to lighting and indicator lights, without adding any new communication cables. The PLC communication module supports wideband operation (e.g., 2-30MHz) and has programmable signal transmission and reception capabilities. The detection is initiated proactively by the main control unit of the tunnel-end monitoring equipment according to a preset strategy (e.g., timed or event-triggered).

[0046] The tunnel-end monitoring equipment scans all power line branches originating from itself according to the following procedure:

[0047] Segmented scanning: The power line circuits within the tunnel are logically divided into multiple line segments, based on physical power distribution branches and logical segments set at fixed intervals (e.g., every 100 meters). The tunnel-end monitoring equipment sequentially sends a set of preset detection signal sequences to each target segment. This sequence consists of multiple standard sine wave or pseudo-random code signals with different center frequencies (e.g., 2MHz, 5MHz, 12MHz, etc.) and different power levels.

[0048] Key channel parameter acquisition: Impedance characteristics: By transmitting a probe signal with a known voltage and accurately measuring the amplitude and phase difference between the reflected and incident signals, the line input impedance of this section at multiple frequency points is calculated and recorded to reflect the electrical characteristics of the line and the end load conditions. Noise level: During the intervals between transmitting probe signals, the high-sensitivity receiving circuit of the PLC module monitors the background noise on the line and records the noise power spectral density of each frequency band, paying particular attention to power frequency harmonics (such as 50Hz, 150Hz, etc.) noise and possible impulse noise characteristics. Attenuation characteristics: By comparing the transmitted signal power with the signal power reflected / retransmitted from the end equipment of the target section, the path attenuation value (in dB) of the signal after passing through the section at a specific frequency point is calculated, and the attenuation curves at multiple frequency points are recorded. To provide a data foundation for predicting channel changes during disasters, this involves linking with the data interface of fire alarm systems within tunnels (such as heat / smoke detectors) or conducting simulation tests during system commissioning. The typical change patterns of channel parameters in a specific section, particularly noise levels and impedance, are recorded when a fire alarm is reported or a simulated high temperature occurs. These records are then stored as feature values ​​in a fingerprint database to update the disaster feature vector. For example, high temperatures may cause changes in line insulation parameters, manifesting as a regular drift in impedance within a specific frequency band; smoke may have a slight impact on signal attenuation.

[0049] The collected data is processed in the main control unit to form and update the dynamic channel fingerprint database. This channel fingerprint database is stored in the local non-volatile memory of the tunnel end monitoring equipment and can optionally be uploaded to the central monitoring station for backup. This channel fingerprint database uses line segments as the basic index unit. Each segment entry includes the following fields: segment identifier, unique ID, associated physical location; basic channel fingerprint, multi-frequency impedance matrix, multi-frequency background noise baseline spectrum, and multi-frequency path attenuation baseline value; optimal communication parameters, preferred communication frequencies (1-3 frequencies with the highest overall signal-to-noise ratio and relatively low attenuation, including preferred communication frequencies for normal and emergency situations), channel quality thresholds (the lowest usable signal-to-noise ratio threshold and the maximum allowable attenuation value calculated based on baseline noise and attenuation); and disaster feature vectors, pre-recorded or machine learning-generated feature data reflecting the channel parameter variation patterns under typical fire scenarios under associated fire alarm conditions (such as heat / smoke detector alarms or simulated fire tests), including but not limited to: high-temperature characteristic response (impedance drift mode), smoke adhesion effect (attenuation increment characteristics), noise disturbance mode (fire noise frequency domain distribution), and comprehensive degradation spectrum (signal-to-noise ratio spatiotemporal evolution pattern).

[0050] S2. The central monitoring station of the bureau is linked with the fire alarm system in real time. When a fire alarm signal is received, it immediately generates a dynamic evacuation route based on the fire alarm location, smoke diffusion status and real-time channel status through the disaster evolution prediction model.

[0051] The central monitoring station of the bureau achieves real-time linkage with the tunnel fire alarm system through a dedicated, highly reliable data interface (such as OPC UA). When the fire alarm system confirms and reports a fire alarm signal (including the alarm detector address, alarm type, and alarm time), the signal will immediately interrupt the regular polling task of the central monitoring station and trigger the highest priority emergency processing thread.

[0052] After the emergency response thread is activated, the central monitoring station synchronously calls the following data: disaster data, obtaining the precise fire alarm location from the tunnel fire alarm system (corresponding to the specific detector number, which can be mapped to the tunnel kilometer marker); environmental data, obtaining real-time wind speed and direction data of the alarm point and upstream and downstream from the tunnel environmental monitoring system; channel status data, retrieving the complete and latest dynamic channel fingerprint database stored locally from the tunnel-end monitoring equipment corresponding to the target tunnel, specifically calling the basic channel fingerprint and optimal communication parameters for each line segment; and equipment status data, obtaining the current online status of all lighting and indicator lights in the tunnel to determine the set of available equipment.

[0053] Then, dynamic evacuation routes are generated using a disaster evolution prediction model, such as... Figure 2 As shown, specifically:

[0054] Using the fire alarm location as the initial fire source, and combining real-time air volume and speed data, a real-time smoke diffusion simulation model based on computational fluid dynamics is used to predict the spatial distribution of smoke concentration and visibility attenuation characteristics within the first preset time period.

[0055] It is important to note that the real-time smoke diffusion simulation model based on computational fluid dynamics uses the tunnel's three-dimensional geometry and ventilation system layout as fixed boundary conditions. Taking the fire alarm location as the starting point for both heat and smoke sources, the model inputs the initial heat release rate (HRR) and smoke generation rate estimated based on the alarm type (e.g., heat or smoke detection), and uses real-time acquired airflow and velocity data as core dynamic boundary conditions. The model uses second-level time steps to simulate the smoke diffusion process within the tunnel over a first preset time period (T1, e.g., 5-10 minutes in the future). The output is a time-varying spatial distribution cloud map of smoke concentration, and simultaneously generates a corresponding visibility attenuation feature map based on a preset smoke concentration-visibility conversion relationship (e.g., a simplified model based on Lambert-Beer's law).

[0056] Prior to deployment, considering the tunnel's geometry and ventilation conditions, full 3D CFD simulations were performed on various typical fire scenarios (different fire source locations, different HRRs, and different combinations of ventilation velocities). The spatiotemporal evolution data of the simulated flow field were stored in the database of the central monitoring station. During a disaster, the model quickly matched the closest pre-calculated scenarios from the database based on the real-time alarm location, estimated HRR, and measured wind speed. It then used interpolation algorithms to generate the initial flow field for the current scenario and predicted the location and concentration distribution of the smoke front based on this.

[0057] Based on the above predictions, and combined with the channel fingerprint database, a map of areas where power line carrier communication quality deteriorates due to temperature and smoke particle adhesion during a disaster is predicted.

[0058] It is important to note that the predicted smoke concentration distribution is matched with the selectable disaster feature vectors under each line segment entry in the channel fingerprint database. For example, if the disaster feature vector of a certain segment records a regular drift in impedance at a specific frequency band and a typical change pattern in noise level under simulated fire conditions, then the predicted change in channel parameters for that segment under the current actual disaster situation is calculated proportionally based on the currently predicted smoke concentration value. The calculated predicted changes in channel parameters (impedance change, noise increment) for each line segment are then superimposed onto the baseline value of the basic channel fingerprint record for that segment to recalculate the predicted signal-to-noise ratio for the future time period T1. By traversing all segments, a map of the power line carrier communication quality degradation areas for the entire tunnel power line network is generated, identifying risky segments where the communication quality (signal-to-noise ratio) is below the channel quality threshold.

[0059] The tunnel topology is discretized into an evacuation network with each lighting and indicator light node as the basic unit, and a passage cost weight is dynamically assigned to each edge in the network (i.e., the tunnel segment connecting adjacent nodes).

[0060] It is important to note that the passage cost weight is composed of a weighted sum of normalized values ​​from multiple indicators: Basic length cost, which is the physical length of the tunnel segment corresponding to the edge, and then normalized; Disaster cost, calculated based on the predicted minimum visibility of the segment within time period T1, where disaster cost = (safe visibility threshold - predicted visibility) / safe visibility threshold. The safe visibility threshold can be set empirically; if the predicted visibility ≥ the safe visibility threshold, the disaster cost is 0; Communication cost, calculated based on the predicted communication quality of the segment. If the segment is identified as a risky area, the communication cost is a fixed penalty constant (e.g., 0.5, used to increase the passage cost of the path during the path generation stage), otherwise it is 0. The formula for calculating the passage cost weight is: W = β × normalized value of basic length cost + γ × disaster cost + λ × communication cost, where β, γ, and λ are adjustable weight coefficients, defaulted to β = 0.3, γ = 0.5, and λ = 0.2.

[0061] Using a preset path search algorithm, starting from the main distribution area of ​​trapped personnel or a preset fixed emergency assembly point, and ending at a safety exit or tunnel entrance, several candidate evacuation paths are calculated in the evacuation network.

[0062] It should be noted that the main distribution area of ​​the trapped personnel is the thermal distribution area of ​​the trapped personnel determined by the tunnel personnel positioning system. The destination is the nearest safe exit or tunnel entrance that has not been affected by the disaster. The path search algorithm can use a heap-optimized version of Dijkstra's algorithm or the A* algorithm. In the evacuation network considering dynamic weights, L shortest cost paths (L is usually 3-5) from the starting point to the destination are calculated as a set of candidate evacuation paths.

[0063] The total travel cost, path overlap, and communication risk coefficient of each evacuation route are evaluated, and a weighted fusion is performed to obtain a comprehensive cost score for the route. Based on this score, the sequence of primary and backup evacuation routes is determined.

[0064] It is important to note that the total travel cost is the sum of the weights of all edges (tunnel segments) traversed by the path. This value directly reflects the overall risk level of evacuation along this path; a higher value indicates a greater risk. The path overlap is the proportion of shared edges between this path and other paths to the total number of edges in the path (taking the maximum value); a higher value indicates a worse diversion effect. The communication risk coefficient is a communication reliability threshold (this value is derived from the channel quality threshold defined in S1 or set empirically), taken as the lowest predicted signal-to-noise ratio (SNR) of all edges traversed by the path. min If SNRmin If the SNR is greater than or equal to the communication reliability threshold, then the communication risk coefficient is 0, indicating reliable communication; if the SNR is greater than or equal to the SNR threshold, then the communication risk coefficient is 0, indicating reliable communication. min If the communication reliability threshold is less than the threshold value, the communication risk coefficient is the communication interruption penalty value (e.g., 10). This penalty value is a preset constant, much larger than the sum of the normalized values ​​of other indicators, used to exclude unreliable communication paths from the score, ensuring command reachability. The total travel cost and path overlap of each candidate path are normalized, and then a linear weighted method is used to calculate the comprehensive path cost score. The calculation formula is: Comprehensive Path Cost Score = W1 × Normalized Total Travel Cost + W2 × Normalized Path Overlap + Communication Risk Coefficient, where W1 and W2 are preset weights, such as W1 = 0.7 and W2 = 0.3.

[0065] All paths with a communication risk coefficient equal to the communication interruption penalty value are excluded. The remaining paths are sorted in ascending order of their comprehensive cost score, with the path ranked first serving as the primary evacuation path, and the subsequent paths forming a sequence of backup evacuation paths in sequence.

[0066] Simultaneously, the central monitoring station generates a "Dynamic Path Switching Contingency Plan" that can be directly executed by the tunnel-end monitoring equipment. This plan is generated by extracting key threat parameters and their spatiotemporal evolution thresholds that may affect the availability of the main evacuation route from smoke diffusion projections and communication quality degradation area maps. For example, in main evacuation route segment A1, visibility is predicted to be below 5 meters after time T (starting from the alarm sounding); in main evacuation route node N2, due to heat radiation from the fire source, its communication signal-to-noise ratio is predicted to be below 10 dB after 60 seconds. These predicted threat parameters are then mapped to specific physical monitoring points that can be monitored in real-time by equipment within the tunnel. These physical monitoring points are typically lamp IDs or dedicated sensor IDs. For example, the visibility of segment A1 is mapped to the ambient light sensor integrated in all lamps within that segment; the communication signal-to-noise ratio of node N2 is mapped to the link quality index reported in real-time by the power line carrier communication module of the lamp at that node. The predicted threats are compiled into executable trigger rules, such as: if the local temperature reported by lighting node N5 on the main evacuation path is >70℃, switch to backup evacuation path 1; if the average signal-to-noise ratio of communication in segment B of the main evacuation path is <15dB for 10 seconds, switch to backup evacuation path 2. Multiple rules can be assigned execution priorities to handle conflicts caused by multiple conditions triggering simultaneously. The generated contingency plan table is a structured data set, with core fields including: rule ID, trigger condition, execution action, and rule priority.

[0067] Finally, step S2 outputs a structured evacuation control instruction package, which includes: the main evacuation route node sequence and control parameters, the backup evacuation route sequence and control parameters, and a dynamic route switching plan table.

[0068] S3. Activate the emergency communication dispatch mechanism. The tunnel-end monitoring equipment executes coordinated optical flow guidance control based on dynamic evacuation paths through the power line carrier communication network.

[0069] Once the central monitoring station completes the dynamic evacuation route planning, it immediately sends an "emergency mode activation" command and evacuation control instruction package to the tunnel-end monitoring equipment in the affected tunnel via the redundant network between the secondary monitoring architectures (such as a self-healing fiber optic ring network). Upon receiving the instruction package, the tunnel-end equipment verifies the current online status of all lighting fixtures along the route based on the locally stored equipment ledger. This equipment ledger is a pre-generated and dynamically maintained electronic list of all lighting and indicator lights within the tunnel, recording at least the unique physical ID, logical installation location (associated with tunnel mileage), power line circuit, and communication address of each fixture. The recorded online status is continuously and actively monitored and updated in real-time by the status reporting data from steps S1 and S4. After confirmation, a high-priority "system clock synchronization and emergency mode entry" command is broadcast via the power line carrier communication network to ensure that the controller clock references of all lighting fixtures within the network are consistent.

[0070] Then, as Figure 3 As shown, the tunnel-end monitoring equipment performs cooperative optical flow guidance control based on dynamic evacuation paths. Specifically:

[0071] The tunnel-end monitoring equipment sends real-time control commands to the lighting and indicator lights in the relevant sections of the current main evacuation route, dynamically adjusting their light intensity, flashing frequency, and direction of indication to form a guiding light flow.

[0072] It is important to note that the tunnel-end monitoring equipment selects the optimal communication frequency and transmission power for each target line section based on a dynamic channel fingerprint database. Through the power line carrier communication network, it issues real-time control commands to each light fixture in the main evacuation path sequence. These commands include specific action parameters (brightness value, flashing code, directional angle) and the absolute timestamp of execution or the delay time relative to the synchronization signal. The lights on the control path execute their lighting logic sequentially according to the evacuation direction, visually forming a guiding light strip flowing towards the safety exit. The lighting logic includes cycles of "on-hold-off" and "high brightness-low brightness." The light intensity parameters in the control commands can be combined with environmental noise baseline data associated with the dynamic channel fingerprint database to automatically increase the output power of the lights in sections with predicted low visibility, counteracting visual attenuation caused by smoke and maintaining necessary lighting guidance effects.

[0073] The backup control commands corresponding to the backup evacuation routes are preloaded into the local cache of the lighting and indicator lights in the affected section.

[0074] It is important to note that when the tunnel-end monitoring equipment receives the main evacuation route control command, it simultaneously receives and parses the "Dynamic Route Switching Plan Table" issued by the central monitoring station. Based on this plan table, the tunnel-end monitoring equipment pre-compiles the control command packets corresponding to the backup evacuation routes and then reliably sends these backup control commands to the relevant lighting fixtures via the power line carrier communication network. Upon receiving the commands, the local control units of each lighting fixture store them in their local reserved buffer and send a "preload confirmation" signal back to the tunnel-end monitoring equipment.

[0075] When the actual disaster evolution is detected to reach the preset switching threshold, the tunnel end monitoring equipment sends a switching trigger command to the relevant lighting and indicator lights, immediately activating and executing the pre-loaded backup control command.

[0076] It is important to note that the tunnel-end monitoring equipment continuously monitors the following information to determine whether the path switching threshold has been reached: local sensor data, receiving real-time environmental data (such as smoke concentration and temperature) and equipment status data (such as communication signal-to-noise ratio) reported by lighting fixtures and sensors via the power line carrier communication network; and mandatory commands from higher authorities, receiving mandatory switching instructions from the central monitoring station via the redundant backbone network. The tunnel-end monitoring equipment quickly compares the real-time monitoring data with the preset conditions in the "Dynamic Path Switching Contingency Plan Table". Once the monitoring data meets any preset condition in the plan table, the tunnel-end monitoring equipment immediately broadcasts a switching trigger command to the specific lighting fixture group corresponding to the condition, which has been pre-loaded with backup control instructions, via the power line carrier communication network. Upon receiving the trigger command, the controller of the relevant lighting fixture immediately reads and executes the pre-stored backup control instructions from its local cache, maximizing the continuity and reliability of evacuation visual guidance.

[0077] S4. Real-time collection of electrical parameters and environmental data of each lighting and indicator light fixture location, combined with channel fingerprint database for fault diagnosis, and real-time reporting of lighting fixture status data, local ambient temperature and communication quality data to the central monitoring station.

[0078] It should be noted that each lighting and indicator light fixture, in addition to the light source, driver power supply and PLC module, also integrates a multi-functional environmental sensing and electrical diagnostic module.

[0079] The electrical parameters include: active impedance spectrum detection, where the PLC communication module of the lamp periodically injects a set of sweep frequency detection signals into the power line circuit during non-business communication periods, and synchronously measures the reflection response of the line through a high-precision analog-to-digital converter. After fast Fourier transform by the local processor, the input impedance-frequency characteristic spectrum of the current line is obtained; time-domain reflection waveform, where the PLC module emits a narrow pulse signal at a specific moment and captures the reflected waveform. By analyzing the amplitude, polarity, and time delay of the reflected pulse, the impedance discontinuities in the line (such as joint oxidation, cable damage, insulation degradation points) and their approximate distance from the lamp are determined; operating current harmonic components, where the input current waveform is sampled in real time by a built-in current sensor, harmonic analysis is performed, and the total harmonic distortion and the content of each harmonic are extracted to assess the health status of the power supply; and leakage current value, where the leakage current value is obtained by monitoring the grounding loop current to assess the electrical insulation safety status of the lamp and its circuit.

[0080] The environmental data includes: surface temperature of the lamp housing, with digital temperature sensors installed inside the lamp housing or at key locations on the heat sink; and local ambient temperature gradient, with temperature sensors installed outside the lamp housing and at a certain distance (e.g., 20 cm) from the lamp, the difference between which constitutes the temperature gradient, used to assess the lamp's own heat dissipation rate and the impact of external abnormal heat sources.

[0081] like Figure 4 As shown, the present invention provides a fault diagnosis method;

[0082] Specifically:

[0083] The collected active detection impedance spectrum is convolved and compared with the health fingerprints in the channel fingerprint database under the same location and similar environmental conditions to calculate the feature difference degree. If the feature difference degree exceeds the threshold, the fault type is located by combining the current harmonic abrupt component and the abnormal reflection point of the time domain reflection waveform, and the state is marked as an emergency state.

[0084] It is important to note that the real-time actively detected impedance spectrum is convolved with the health impedance spectrum benchmark stored in the dynamic channel fingerprint database under similar environmental conditions (such as similar ambient temperature and the same brightness level of the lamps). The difference between the two in the characteristic frequency band is then calculated (e.g., by calculating the normalized cross-correlation coefficient or the frequency-domain weighted Euclidean distance).

[0085] If the impedance spectrum characteristic difference is below a threshold (e.g., correlation coefficient below 0.85), the luminaire status is marked as normal; if it exceeds the threshold, a deep diagnostic process is triggered. This deep diagnostic process includes: immediately checking the harmonic components of the operating current within the same acquisition cycle; if a specific harmonic (e.g., the 3rd harmonic) content shows an abnormal surge relative to the historical baseline, it indicates that the filter capacitor in the drive power supply is dry, the inductor is saturated, or the power switching device performance is degraded. Simultaneously analyze the time-domain reflection waveform acquired this time; if a new abnormal reflection peak appears on the waveform other than at the luminaire's own location, and the position of this reflection peak is fixed, it indicates a local impedance fault point (e.g., loose connection) on the line between the luminaire and the tunnel-end monitoring equipment. By calculating the time delay of the emission peak, the distance to the fault point can be estimated. If any of the above abnormal characteristics are met, the luminaire status is marked as emergency, and the diagnosed fault type code is recorded, such as "F1: Power supply harmonic abnormality" or "F2: Impedance fault at a distance of 3 meters from the line".

[0086] Trend analysis is performed on the long-term time-series data of leakage current and the relationship curve between the casing temperature and the ambient temperature gradient. When the leakage current exceeds the safety threshold within the third preset time period or the slope of the casing temperature-ambient temperature gradient curve remains abnormal, the status is marked as an early warning status, and the remaining effective working time is estimated.

[0087] It is important to establish a long-term (e.g., the past 24 hours) time-series database of leakage current values ​​and use trend analysis algorithms (e.g., exponentially weighted moving average) to analyze their changing trends and rates. If the leakage current value continuously exceeds the preset safety threshold (e.g., 0.5mA) within the third preset time period (T3, e.g., 10 consecutive minutes), or if its value does not exceed the safety threshold but its rate of increase exceeds the preset degradation rate threshold (0.05mA / min), it indicates that the insulation performance is deteriorating rapidly and there is a potential risk of breakdown. In this case, the lighting fixture status should be marked as a warning state.

[0088] Establish a historical relationship model between the luminaire housing temperature and the ambient temperature gradient, and generate a housing temperature-ambient temperature gradient relationship curve. Under the same operating power and ambient temperature, a healthy luminaire should maintain a relatively stable relationship between its temperature rise and temperature gradient. If an abnormal change in the slope of this curve is detected, such as an abnormal increase in housing temperature leading to a decrease in the gradient under the same ambient temperature, it indicates a decline in the efficiency of the luminaire's heat dissipation system, and the luminaire status should be marked as a warning state.

[0089] At the same time, the remaining effective operating time of the luminaire or the recommended next maintenance time window can be extrapolated based on the degradation rate model (such as the accelerated life model based on the Arrhenius equation).

[0090] The luminaire status data includes: online / offline status, determined by periodic link detection and heartbeat response mechanisms to assess the communication connection status of lighting and evacuation indicator lights. If a luminaire fails to respond to polling or beacon frames from the tunnel-end monitoring equipment within multiple communication cycles, its status is marked as "offline"; otherwise, it is marked as "online." Fault diagnosis status, such as the "normal status," "warning status," and "emergency status" defined above; and real-time operating parameters, including the luminaire's current operating current, voltage, power, and cumulative operating time of the light source. When luminaires report alarms or periodic status reports via power line carrier communication networks, the above status data must be encapsulated into a frame format conforming to the protocol. Key fields include: frame header and address, start character, source light fixture physical location code, and binding to tunnel mileage location; status and diagnostics section, status level identifier (00-normal, 01-warning, 10-emergency, 11-offline / communication interruption), diagnostic result code (specific fault / warning type in emergency / warning state), key data (emergency state with abnormal parameter values, warning state with current value, trend rate and estimated remaining time); environmental data section, shell temperature, ambient temperature, and calculated temperature gradient at the time of reporting; communication quality section, measured signal-to-noise ratio and packet error rate after forward error correction during this data reporting communication process; frame check sequence and CRC check code.

[0091] S5. In emergency mode, communication dispatch adopts a hybrid multiple access mechanism that prioritizes fire events: the tunnel end monitoring equipment broadcasts instructions through a dynamic time division multiple access mechanism, and the lighting and indicator lights report critical statuses using a carrier sense multiple access mechanism based on fault status priority. Instructions can interrupt regular communication at any time to ensure that the instruction channel has priority.

[0092] It is important to note that when the tunnel-end monitoring equipment receives the "Emergency Mode Activation" command from the central monitoring station, its main control unit immediately controls the PLC communication module to switch from normal polling to emergency hybrid multiple access scheduling mode. The tunnel-end monitoring equipment re-divides the continuous communication timeline into periodically repeating superframes. Each superframe contains two parts: a command issuance time slot, dedicated to broadcasting or multicasting control commands from the tunnel-end monitoring equipment to all lighting fixtures; and a status reporting time slot, dedicated to the lighting fixtures reporting their own status data to the tunnel-end monitoring equipment. At the beginning of each superframe, the tunnel-end monitoring equipment first broadcasts a mandatory synchronization beacon frame through the power line carrier communication network. This frame includes a superframe start flag, the precise length and start and end times of the current command issuance time slot and status reporting time slot, and network-wide clock calibration information.

[0093] The length of the instruction issuance time slot is adjusted based on several factors, including: the urgency of the broadcast instruction (e.g., initial evacuation instructions and route switching trigger instructions are of the highest urgency, while single-point dimming instructions for lighting fixtures have lower urgency); and the channel status of the broadcast area. Based on a dynamic channel fingerprint database, the channel quality of the line segment to be covered by the current instruction is assessed. If a segment is found to have high attenuation or high noise, the transmission time slot for that instruction needs to be extended, or a lower-rate, stronger error-correction encoding method should be used for retransmission to ensure reception reliability. Before each instruction is sent, the tunnel-end monitoring equipment calculates the required time slot length based on the above factors. This length information is included in the synchronization beacon frame or instruction frame header, and the start time of the next status reporting time slot for all lighting fixtures will be delayed accordingly.

[0094] The length of the status reporting time slot is dynamically adjusted based on the total number n of lights in the network that are in warning and emergency states. The larger n is, the more urgent information needs to be reported, and the longer the allocated reporting time slot. The calculation formula is: Status reporting time slot length = Base length + k × n, where the base length can be set to 20ms, and the weighting coefficient k is 2ms. This calculation is performed on the monitoring equipment at the tunnel end, and the result is published via synchronization beacon frames.

[0095] The basis for dynamic adjustment of the communication cycle includes: (1) Emergency stage and command characteristics: In the initial stage of fire alarm, it is necessary to broadcast global evacuation commands quickly and multiple times, and collect environmental data intensively to confirm the disaster. At this time, the superframe cycle should be shortened (e.g., from the normal 1 second to 200 milliseconds) to achieve high-frequency command refresh and status polling. In the stable execution stage of evacuation path, after the main evacuation path is determined and executed stably, the superframe cycle can be appropriately extended (e.g., 500 milliseconds) to reserve more time slot resources for status reporting, while still ensuring timely response to switching commands. When the path is switched or the disaster situation changes suddenly, once the path switching condition is triggered, or a large number of emergency status reports are received, the shortest cycle mode is immediately switched to issue switching commands at the highest frequency and collect the latest situation information. (2) Network load and service density: When the number of lamps n in the early warning or emergency state surges, it is necessary not only to increase the length of the status reporting time slot, but also to shorten the cycle to avoid data accumulation at the terminal. (3) Overall assessment of channel quality: If the channel quality of the entire network is generally deteriorated through the dynamic channel fingerprint database or the data reported by S4, such as the background noise being greatly increased due to a fire, the period should be appropriately extended to allow sufficient time for a single transmission, so as to avoid a large number of transmission timeout failures due to the period being too short.

[0096] Within the status reporting time slot, luminaires in different health states employ a differentiated contention strategy, with their priority directly determined by the diagnostic results of step S4. Specifically: Level 1, emergency state luminaires, possess the highest priority for preemption. Once data to be reported is available, they immediately interrupt any other communication process in the current channel, forcibly inserting and reporting their emergency state frame. Level 2, warning state luminaires, possess priority contention. When an idle channel is detected, they can directly preempt the next available communication time slot for reporting, without undergoing a complete random backoff process. Level 3, normal state luminaires, employ a standard carrier sense and random backoff mechanism for contention in reporting.

[0097] The command issuance time slot has the highest network priority. At any time, the tunnel-end monitoring equipment can immediately interrupt the ongoing status reporting time slot to issue emergency control commands (such as evacuation route switching commands), such as by sending a special channel clearing signal or a forced synchronization beacon frame to clear the channel and switch to the command issuance time slot.

[0098] In addition, such as Figure 5 As shown, the present invention also includes, in order to deal with the failure of the monitoring equipment at the tunnel end, designing lighting or indicator lights that are allowed to have direct environmental perception capabilities, and when preset conditions are met, upgrading them to temporary master nodes to organize nearby lighting and indicator lights for emergency evacuation guidance.

[0099] Specifically, each lighting or indicator light fixture with direct environmental perception capabilities is pre-programmable for autonomous upgrade logic. This logic resides in the secure storage area of ​​its controller after the fixture is powered on and is activated only when all of the following preset conditions are met: communication failure determination: the fixture has not received a synchronization beacon frame or any valid command from the tunnel-end monitoring equipment for several consecutive superframe cycles (e.g., 10 cycles); direct disaster perception: the data detected by the environmental sensors integrated into the fixture simultaneously exceeds a preset emergency threshold (e.g., temperature > 70°C and smoke concentration > 5%obs / s); power supply status: the fixture's own power supply is normal. Once a fixture determines that its conditions meet the above preset conditions, it will broadcast a "temporary master node election request" frame on its power line branch via the power line carrier communication network. This frame contains the fixture ID, current location code, and perceived environmental data. Other fixtures that receive this request, if they also meet or partially meet the above preset conditions, will reply with their own election information after a random backoff. The election rules follow a predefined priority algorithm, such as: Priority 1, the luminaire with the highest perceived environmental hazard level takes priority; Priority 2, if the hazard levels are the same, the luminaire with a better physical location takes priority. The luminaire that wins the election will promote itself to a temporary master node and broadcast a "temporary master node ready" declaration frame to other luminaires within its communication range.

[0100] The temporary master node continuously collects its own status and environmental perception data reported by neighboring lights. Based on the preset tunnel topology map and its own location, it determines the directions of at least two nearest safe exits. Simultaneously, based on its own and neighboring light's reported temperature and smoke data, it assesses the hazard level in each direction. The direction away from the hazard source and pointing towards the nearest safe exit is selected as the initial evacuation direction. A guidance command based on unidirectional light flow is generated. The command is as follows: starting from the temporary master node, control the synchronous high-frequency flashing (e.g., 3Hz) of M consecutive lights along the selected evacuation direction (all lights within a 20m downstream), forming a clear indicator light strip. For lights in the opposite direction, they can be controlled to illuminate a solid red light as a no-passage warning.

[0101] The temporary master node broadcasts the aforementioned guidance instructions to the lights in the affected area via the power line carrier communication network. The lights receiving the instructions execute them immediately and send a confirmation reply to the temporary master node. The temporary master node then instructs one light at the end of the evacuation route to become the next-hop temporary sub-master node, responsible for organizing the guidance of its downstream area, if communication is lost and danger is detected.

[0102] The temporary master node will take over the communication scheduling of the local network. It can use a simplified version of the S5 hybrid multiple access communication mechanism. For example, it periodically broadcasts simplified synchronization beacons, divides short command slots and status reporting slots, and prioritizes emergency status reports. To prevent interference, the temporary master node listens for channels before broadcasting. If it detects a valid signal from another temporary master node or a potentially recoverable original tunnel-end monitoring device, it will back off or become a slave node according to a preset priority (the original central device has the highest priority). All lights, including the temporary master node, will still attempt to report critical status data (emergency / warning status, environmental data) to possible higher-level systems via the power line carrier communication network to provide information for possible global rescue efforts.

[0103] The temporary master node periodically listens for beacons from the original tunnel-end monitoring equipment. Once it receives a valid synchronization beacon from the original system, the temporary master node immediately stops broadcasting self-organizing boot instructions, reports its status and data from its time as a temporary master node via the power line carrier communication network, and then automatically degrades to a regular slave node, resuming centralized control.

[0104] The method may also include a mobile access gateway for authorizing secure access to mobile terminals, enabling status monitoring and emergency command.

[0105] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A method for IoT-based network scheduling of disaster prevention lighting and evacuation guidance in railway tunnels, characterized in that, include: S1. The tunnel-end monitoring equipment actively initiates environmental channel detection through the power line carrier communication network, scans the impedance characteristics, noise levels and attenuation characteristics of different line sections in the tunnel, establishes a dynamic channel fingerprint database, and records the preferred communication frequency and channel quality threshold; the preferred communication frequency is the communication frequency determined based on the signal-to-noise ratio and path attenuation. S2. The central monitoring station of the bureau is linked with the fire alarm system in real time. When a fire alarm signal is received, it immediately generates a dynamic evacuation route based on the fire alarm location, smoke diffusion status and real-time channel status through the disaster evolution prediction model. S3. Activate the emergency communication dispatch mechanism. The tunnel-end monitoring equipment executes cooperative optical flow guidance control based on dynamic evacuation paths through the power line carrier communication network. S4. Real-time collection of electrical parameters and environmental data of each lighting and indicator light fixture location, combined with channel fingerprint database for fault diagnosis, and real-time reporting of lighting fixture status data, local ambient temperature and communication quality data to the central monitoring station. S5. In emergency mode, communication dispatch adopts a hybrid multiple access mechanism that prioritizes fire events: the tunnel end monitoring equipment broadcasts instructions through a dynamic time division multiple access mechanism, and the lighting and indicator lights report critical statuses using a carrier sense multiple access mechanism based on fault status priority. Instructions can interrupt regular communication at any time to ensure that the instruction channel has priority. The electrical parameters include the active detection impedance spectrum, time-domain reflection waveform, operating current harmonic components, and leakage current value of the circuit; the environmental data include the surface temperature of the lamp housing and the local ambient temperature gradient; the critical states include normal state, emergency state, and warning state. The fault diagnosis method includes: S41. Perform convolutional comparison between the collected active detection impedance spectrum and the healthy fingerprints in the channel fingerprint database under the same location and similar environmental conditions, and calculate the feature difference degree. If the characteristic difference exceeds the threshold, the fault type is located by combining the current harmonic mutation component with the abnormal reflection point of the time domain reflection waveform, and the state is marked as an emergency state. S42. Perform trend analysis on the long-term time series data of leakage current and the relationship curve between shell temperature and ambient temperature gradient. When the leakage current exceeds the safety threshold within the third preset time period or the slope of the relationship curve between shell temperature and ambient temperature gradient continues to be abnormal, mark the status as a warning status and estimate the remaining effective working time.

2. The IoT networking scheduling method for railway tunnel disaster prevention lighting and evacuation indication according to claim 1, characterized in that: In S2, dynamic evacuation routes are generated through a disaster evolution prediction model. Specific methods include: S21. Taking the fire alarm location as the initial fire source point, and combining real-time air volume and wind speed data, predict the spatial distribution of smoke concentration and visibility attenuation characteristics within the first preset time period using a real-time smoke diffusion simulation model based on computational fluid dynamics. S22. Based on the prediction results of S21, and combined with the channel fingerprint database, predict the area map of power line carrier communication quality degradation caused by temperature and smoke particle adhesion during the disaster. S23. Discretize the tunnel topology into an evacuation network with each lighting and indicator light node as the basic unit, and dynamically assign passage cost weights to each edge in the network. S24. Using a preset path search algorithm, starting from the main distribution area of ​​trapped personnel or a preset fixed emergency assembly point, and ending at a safety exit or tunnel entrance, calculate several candidate evacuation paths in the evacuation network. S25. Evaluate the total travel cost, path overlap, and communication risk coefficient of each evacuation route, and obtain a comprehensive cost score for the route by weighted fusion. Based on this score, determine the sequence of primary evacuation routes and backup evacuation routes.

3. The IoT networking scheduling method for disaster prevention lighting and evacuation indication in railway tunnels according to claim 1, characterized in that: S3 specifically includes: S31. The tunnel end monitoring equipment sends real-time control commands to the lighting and indicator lights in the relevant sections of the current main evacuation route, dynamically adjusting their light intensity, flashing frequency and indication direction to form a guiding light flow. S32. Preload the backup control commands corresponding to the backup evacuation routes into the local cache of the lighting and indicator lights in the affected section; S33. When the actual disaster evolution is detected to reach the preset switching threshold, the tunnel end monitoring equipment sends a switching trigger command to the relevant lighting and indicator lights, immediately activating and executing the pre-loaded backup control command.

4. The IoT networking scheduling method for disaster prevention lighting and evacuation indication in railway tunnels according to claim 1, characterized in that: The hybrid multiple access mechanism in S5 includes: S51. In emergency mode, the tunnel end monitoring equipment divides the communication cycle into instruction issuance time slots and status reporting time slots. The instruction issuance time slot adopts a time division multiple access mechanism, and its time slot length is adjusted in real time according to the urgency of the broadcast instruction and the current broadcast area channel status evaluated in the channel fingerprint database; The status reporting time slot adopts a carrier sense multiple access mechanism, and its time slot length is dynamically allocated according to the total number of lighting and indicator lights marked as warning status and emergency status in the network. S52. During the status reporting time slot, the carrier sense multiple access behavior of each lighting and indicator light fixture is controlled by the fault diagnosis status determined in S4. For lighting and indicator light fixtures in normal status, a standard random backoff contention mechanism is used for reporting; for lighting and indicator light fixtures in warning status, priority contention is granted; and for lighting and indicator light fixtures in emergency status, the highest level of preemption is granted.

5. The IoT networking scheduling method for railway tunnel disaster prevention lighting and evacuation guidance according to claim 1, characterized in that: When the lighting status data is reported, the status level identifier and physical location code corresponding to the lighting and indicator lights must be embedded in the data frame.

6. The IoT networking scheduling method for disaster prevention lighting and evacuation indication in railway tunnels according to claim 1, characterized in that: The method also includes, in order to deal with the failure of monitoring equipment at the tunnel end, designing lighting or indicator lights that allow direct environmental perception capabilities, and upgrading them to temporary master nodes when preset conditions are met, organizing nearby lighting and indicator lights for emergency evacuation guidance.

Images