Fire emergency communication system

Abstract

The application relates to the technical field of fire emergency rescue, and discloses a fire emergency communication system, which comprises an environment sensing module, a unmanned aerial vehicle relay network and a communication optimization processing module.The environment sensing module is used for collecting temperature field, smoke concentration, air flow velocity field and building structure data of a fire scene, and transmitting the data to the communication optimization processing module through a LoRaWAN protocol.The unmanned aerial vehicle relay network comprises a plurality of unmanned aerial vehicles carrying multi-frequency communication equipment, and the unmanned aerial vehicles support a hybrid link of millimeter wave, a 5G private network and satellite communication.The application adopts a multimodal cooperative collection scheme of infrared thermal imaging, laser radar and multispectral sensors, combines Kalman filtering to realize real-time fusion of temperature field, smoke concentration and building structure data, and achieves a three-dimensional environment modeling effect with a centimeter-level precision.Compared with the high misjudgment rate and slow dynamic response caused by the single sensor dependence in the prior art, the core problem of insufficient data reliability in a complex fire environment is solved.

Description

Technical Field

[0001] This invention relates to the field of fire emergency rescue technology, specifically a fire emergency communication system. Background Technology

[0002] In recent years, with the continuous increase in the height and complexity of urban buildings, the difficulty of fire rescue has significantly increased. Under the multiple threats of high temperatures, dense smoke, and building collapse, firefighters need to rely on real-time fire data and stable communication links for accurate decision-making. However, traditional fire communication systems are limited by static deployment patterns and single communication frequency bands, making it difficult to cope with the dynamic and unpredictable characteristics of the fire environment. For example, signal attenuation caused by dense smoke, equipment failure caused by high temperatures, and communication blind spots created by building structures severely restrict the real-time transmission and execution efficiency of rescue commands.

[0003] Currently, mainstream fire emergency communication solutions fall into two main categories: sensor networks based on fixed ground nodes and drone-assisted temporary relay systems. In fixed-node solutions, temperature sensor arrays are typically deployed using ZigBee or LoRa protocols, transmitting data back to the command center via multi-hop transmission. This type of solution performs reasonably well in open areas, but is prone to signal loss due to wall obstruction in high-rise building fires. The other type, drone relay solutions, utilize drones equipped with 4G / 5G communication modules to construct aerial base stations to extend coverage. However, existing drone communication links are mostly configured with fixed frequency bands, lacking the ability to adaptively adjust to interference in the fire environment. For example, when the smoke concentration in a fire increases sharply, the penetration power of millimeter-wave signals decreases drastically, and existing systems cannot autonomously switch to more interference-resistant satellite links, leading to the interruption of critical data transmission.

[0004] Existing technologies, whether fixed-node networks or drone relay systems, share a common deficiency: insufficient adaptability to the dynamic environment of a fire scene. Specifically, fixed-deployment sensor networks cannot dynamically adjust their coverage as the fire spreads, resulting in communication blind spots in high-temperature areas. Drone relay systems, due to their static frequency band configuration and single environmental perception methods, suffer severe signal attenuation under sudden increases in smoke concentration or interference from high-temperature airflows, and lack a mechanism for autonomously switching to an anti-interference link. This causes delays or loss of critical rescue commands, directly affecting the timeliness and success rate of rescue operations. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a fire emergency communication system that solves the problems of poor communication link stability, numerous coverage blind spots, and insufficient adaptive capabilities in existing fire emergency communication systems under dynamic fire conditions such as smoke and high temperatures.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a fire emergency communication system, comprising: The environmental sensing module is used to collect data on the temperature field, smoke concentration, air velocity field, and building structure at the fire scene, and transmit the data to the communication optimization and processing module via the LoRaWAN protocol. The drone relay network includes multiple drones equipped with multi-band communication devices. The drones support hybrid links of millimeter wave, 5G private network and satellite communication, and the communication link switching is dynamically triggered based on smoke concentration and temperature field data. The communication optimization processing module is used to jointly optimize the communication performance of the UAV swarm based on the data from the environmental perception module, including signal propagation modeling, UAV layout optimization, dynamic channel allocation, and UAV trajectory optimization. The emergency command center is used to receive and process real-time fire data, generate dynamic rescue instructions, and send them to firefighter terminals via fiber-optic-satellite dual-redundant links. The adaptive feedback module is used to monitor communication performance indicators in real time and is connected to the communication optimization processing module via the CAN bus. When the indicator exceeds the preset threshold, it triggers a re-optimization process.

[0007] Preferably, the environment sensing module includes: Infrared thermal imaging arrays are used to collect temperature field data at fire scenes; LiDAR is used to acquire three-dimensional point cloud data of building structures; Multispectral sensors are used to detect smoke concentration and airflow field. The data fusion unit is used to fuse temperature field, smoke concentration, air velocity field and building structure data into a unified environmental model, and connects to the LoRaWAN gateway via RS-485 bus to transmit the data to the communication optimization processing module.

[0008] Preferably, in the drone relay network, each drone includes: The millimeter-wave communication unit operates in the 28GHz band and supports beamforming and dynamic angle adjustment. The 5G private network communication unit supports dynamic adjustment of TDD time slot allocation for high-bandwidth data transmission. The satellite communication unit, with a built-in parabolic antenna and LDPC error correction encoder, is used for emergency communication in extreme environments. The link switching control unit is used to dynamically select the communication link based on smoke concentration and temperature field data. When the smoke concentration exceeds 50 mg / m³, the control unit will switch the link to the next communication link. 3 At that time, the satellite communication unit should be activated first.

[0009] Preferably, the communication optimization processing module includes: The signal propagation modeling unit is used to receive fused data from the environmental perception module via an Ethernet interface and construct a wireless signal propagation model based on the ray tracing method. The UAV deployment optimization unit is used to solve for the optimal spatial coordinates of the UAV swarm using a simulated annealing algorithm based on the signal coverage requirements of the target area. The dynamic channel allocation unit is used to allocate channel resources through a deep reinforcement learning algorithm. The algorithm's reward function includes a weighted sum of throughput and end-to-end latency. The drone trajectory optimization unit is used to plan the cooperative flight path of a drone swarm through a multi-agent game model.

[0010] Preferably, in the signal propagation modeling unit, the wireless signal propagation model is constructed using the ray tracing method, and the model parameters include: The linear correction factor for signal attenuation due to temperature gradient; The nonlinear correction coefficient for the effect of smoke concentration on signal absorption; The dynamic correction factor of air velocity field on signal propagation direction.

[0011] Preferably, in the UAV layout optimization unit, the signal coverage requirements of the target area are divided by a hexagonal grid, the grid side length is inversely proportional to the fire spread speed, and the optimal spatial coordinates of the UAV swarm are solved by a simulated annealing algorithm.

[0012] Preferably, the dynamic channel allocation unit adopts a hybrid architecture of generative adversarial networks and deep reinforcement learning, wherein: The generator uses a convolutional neural network to generate candidate channel allocation strategies; The discriminator assesses the channel collision probability through the spectrum sensing circuit and feeds it back to the generator for iterative optimization.

[0013] Preferably, the UAV trajectory optimization unit coordinates the flight paths of the UAV swarm through a multi-agent game model, and the game payoff function includes: The signal-to-noise ratio improvement rate and weighting coefficients are dynamically adjusted according to the fire temperature. The energy consumption reduction rate, with a weighting coefficient, is dynamically adjusted based on the remaining battery power of the drone.

[0014] Preferably, the communication performance indicators monitored in real time by the adaptive feedback module include: The signal-to-noise ratio is set to trigger re-optimization when it drops by more than 10%, and the threshold is set based on statistical analysis of historical fire data. The latency is set to trigger a re-optimization when it exceeds 100ms. The threshold is set according to the business requirements of the firefighter terminal. The drone's position offset will trigger a re-optimization when it exceeds 5 meters.

[0015] Preferably, the emergency command center is connected to the communication optimization processing module via a fiber-optic-satellite dual-redundant link. The fiber-optic link is used for high-bandwidth data transmission, and the satellite link is used for emergency communication in extreme environments.

[0016] This invention provides a fire emergency communication system. It has the following beneficial effects: 1. This invention employs a multimodal collaborative acquisition scheme combining infrared thermal imaging, lidar, and multispectral sensors. By integrating Kalman filtering with real-time fusion of temperature field, smoke concentration, and building structure data, it achieves centimeter-level accuracy in 3D environmental modeling. Compared to existing technologies that rely on a single sensor, resulting in high false alarm rates and slow dynamic response, this invention solves the core problem of insufficient data reliability in complex fire environments.

[0017] 2. This invention dynamically switches the millimeter-wave-5G-satellite link in real time by triggering data on smoke concentration and temperature field. Combined with LDPC coding and multipath redundancy transmission, it achieves a communication error rate of ≤1e-6 in high-interference scenarios. Compared with the shortcomings of traditional fixed link strategies, which frequently interrupt communication when smoke concentration suddenly increases, this invention solves the problem of insufficient robustness in emergency communication.

[0018] 3. This invention employs a collaborative strategy that combines simulated annealing algorithm to optimize UAV layout, GAN+DRL hybrid architecture for channel allocation, and differential game model for trajectory planning, reducing the coverage blind zone area by more than 40%. Compared to the resource allocation imbalance caused by isolated optimization of individual modules in existing technologies, this invention solves the problem of low collaborative efficiency in UAV swarms.

[0019] 4. This invention achieves millisecond-level optimized command issuance speed by real-time monitoring of signal-to-noise ratio, latency, and position offset, combined with dynamic thresholds and multi-level triggering logic. Compared to the shortcomings of traditional passive response mechanisms in terms of delayed control during sudden changes in the fire situation, this invention solves the bottleneck of insufficient system adaptability and ensures the continuous availability of the communication link. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the overall system architecture and data flow of the present invention. Figure 2 This is a schematic diagram of the internal process of the communication optimization processing module of the present invention; Figure 3 This is a schematic diagram of the control logic of the adaptive feedback module of the present invention; Figure 4 This is a schematic diagram of the emergency command center instruction issuance process of the present invention. Detailed Implementation

[0021] The technical solutions in 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.

[0022] Please see the appendix Figure 1-4 This invention provides a fire emergency communication system, comprising: In this embodiment, the environmental perception module includes an infrared thermal imaging array, a lidar, a multispectral sensor, and a data fusion unit, the technical details of which are as follows: 1. Temperature field data acquisition In general, infrared thermal imaging arrays are deployed on UAV platforms and fixed ground nodes to cover the core area of ​​the fire.

[0023] In some embodiments, the infrared sensor employs an uncooled microbolometer, operating in the 8-14 μm wavelength range, with a temperature resolution ≤0.1℃ and a frame rate of 30fps. Each frame of data generates a temperature field matrix. ,in: Two-dimensional spatial coordinates, with a grid precision of 0.5m × 0.5m; : Collect timestamps, time interval Δt=100ms; Temperature value, range -20℃ to 1000℃.

[0024] In one possible implementation, when temperature field data is transmitted via the LoRaWAN protocol, differential coding compression technology is used, with a compression ratio of ≥50%.

[0025] 2. Smoke Concentration Detection Specifically, the multispectral sensor integrates the visible light (400-700nm) and near-infrared (700-1700nm) bands, and retrieves the concentration by analyzing the spectral characteristics of smoke particle scattering. The formula for calculating smoke concentration is: ; Parameter definition: and Visible and near-infrared light intensity; =0.35, =0.45, =0.2: Calibration coefficient, fitted using measured data from a laboratory smoke chamber; Temperature gradient, calculated from the temperature difference between adjacent grids.

[0026] When smoke concentration > 100 mg / m³ 3 At that time, the sensor automatically switches to high sensitivity mode, and the sampling frequency is increased to 50Hz.

[0027] 3. Airflow field modeling air velocity field Measurements were taken using a Doppler lidar. The lidar used a 1550nm wavelength, a scanning frequency of 20Hz, and a speed resolution of 0.1m / s.

[0028] Velocity vector calculation is based on the Doppler frequency shift formula: ; Parameter definition: : Frequency shift of the echo signal; =1550nm: Laser wavelength; The incident angle of the light beam is fed back in real time by the radar gimbal angle encoder.

[0029] 4. Building structural data acquisition LiDAR acquires 3D point cloud data of buildings using the time-of-flight (TOF) principle. The point cloud density is ≥200 points / m². 2 The accuracy is ±2cm. The building structure model is stored in voxel mesh format with a mesh size of 0.1m. 3 .

[0030] When the visibility at the fire site is less than 5m, the lidar switches to high-power mode, the pulse energy is increased to 1mJ, and the maximum detection distance is extended to 50m.

[0031] 5. Multi-source data fusion The data fusion unit employs a Kalman filter algorithm to unify temperature field, smoke concentration, flow velocity field, and building structure data into a spatiotemporal coordinate system. The fusion equation is as follows: ; ; Parameter definition: : State vector, containing , , , ; : Control input (such as drone position adjustment commands); Covariance matrix of process noise and observation noise; , , The state transition matrix, control matrix, and observation matrix are determined through offline calibration.

[0032] 6. Data Transmission Protocol Specifically, the merged environment model data is transmitted to the LoRaWAN gateway via an RS-485 bus, using a custom binary format. The data packet structure includes: Packet header: 2-byte synchronization code (0x55AA); Data segments: Temperature field (4 bytes / grid), smoke concentration (2 bytes / grid), flow velocity field (4 bytes / vector); Verification code: CRC-16 checksum.

[0033] As an alternative, when the communication distance is greater than 3km, the gateway automatically switches to LoRaWAN Class C mode with a spreading factor of SF=12, increasing the link budget to 157dB.

[0034] As an aerial node in the fire emergency communication system, the UAV relay network's dynamic switching and layout optimization directly rely on smoke concentration, temperature field, and building structure data collected by the environmental perception module. This module, through multi-band communication units and intelligent switching logic, ensures a highly reliable communication link in complex fire environments and provides real-time channel status feedback to the communication optimization processing module.

[0035] In this embodiment, the drone relay network includes multiple drones equipped with multi-band communication devices, and its hardware configuration and link management logic are as follows: Communication unit hardware: The UAV is equipped with a millimeter-wave communication unit, a 5G private network unit, and a satellite communication unit, supporting dynamic switching of hybrid links. The millimeter-wave communication unit operates in the 28GHz band, uses a 64-channel phased array antenna, has an adjustable beamforming angle range of ±15°, supports MIMO 4×4 technology, and has a theoretical peak rate of 10Gbps.

[0036] The 5G private network unit supports dynamic TDD time slot allocation, with the time slot ratio dynamically adjusted according to uplink and downlink service demands. For example, when video backhaul services are dominant, the uplink time slot ratio increases to 70%, while the downlink time slot ratio increases to 60% when control command transmission is in progress.

[0037] The satellite communication unit has a built-in parabolic antenna with a gain of ≥30dBi, operates in the Ku band (12-18GHz), uses LDPC forward error correction coding with a code rate of 1 / 2 and a bit error rate of ≤1e-6.

[0038] Link switching decisions are based on smoke concentration and temperature field data, which are provided in real time by the environmental perception module.

[0039] In some embodiments, when the smoke concentration exceeds 50 mg / m³ 3 Furthermore, when the temperature exceeds 200℃, it will be forcibly switched to the satellite communication unit to avoid the absorption and attenuation of millimeter wave and 5G signals by smoke.

[0040] Specifically, the priority of millimeter wave and 5G links is evaluated using a weighted average of Received Signal Strength Indicator (RSSI), calculated as follows: ; Parameter definition: =0.5, =0.3, =0.2: Weighting coefficient, determined using historical link performance data; Signal-to-noise ratio (SNR), in dB; Interference intensity, normalized to the [0,1] interval; Link stability index, calculated based on the variance of signal fluctuations over the last 10 seconds.

[0041] If the priority difference is greater than 0.2, a link handover will be triggered with a handover delay of ≤100ms.

[0042] When the satellite communication link is interrupted, multipath redundancy transmission between millimeter wave and 5G links is enabled.

[0043] The data packet is segmented into multiple substreams and transmitted in parallel across different frequency bands. The receiving end uses the Maximum Ratio Combining (MRC) algorithm to reconstruct the original data. The formula for calculating the combining gain is: ; Parameter definition: Number of available sub-streams; : Signal-to-noise ratio of the i-th sub-stream.

[0044] The drone acquires location data through a dual-mode GPS and BeiDou positioning module with a positioning accuracy of ≤0.5m, and achieves microsecond-level time synchronization through the IEEE1588v2 protocol.

[0045] When the GPS signal fails due to electromagnetic interference from the fire scene, switch to LiDAR SLAM (Simultaneous Localization and Mapping) mode, with a positioning error of ≤1m.

[0046] The drone communication unit shell is made of aluminum nitride ceramic material with a thermal conductivity of ≥170W / (m·K) and integrates a thermoelectric cooler (TEC) to ensure that the chip junction temperature is ≤85℃ when the ambient temperature is ≤300℃.

[0047] The surface of the parabolic antenna of the satellite communication unit is coated with a high-temperature resistant zirconium oxide coating, which can withstand temperatures ≥500℃, to prevent high-temperature deformation from causing a decrease in gain.

[0048] In this embodiment, the communication optimization processing module includes a signal propagation modeling unit, a UAV layout optimization unit, a dynamic channel allocation unit, and a UAV trajectory optimization unit, the technical details of which are as follows: The signal propagation modeling unit constructs a wireless channel model under fire conditions using the ray tracing method.

[0049] The three-dimensional space of the fire scene is discretized into a 1m×1m×1m voxel grid, and the signal path loss of each grid is calculated: ; Parameter definition: Free space path loss, calculated using the following formula: ( / ),in =1m, The wavelength of the signal; Path loss index, taken under fire conditions =3.2; =0.15、 =0.08: Smoke concentration With temperature gradient Correction factor; The straight-line distance between the transmitter and receiver.

[0050] When the temperature is greater than 300℃, the path loss index Dynamic adjustment =3.5, to reflect the aggravating effect of high-temperature air ionization on signal attenuation.

[0051] The UAV deployment optimization unit solves for the optimal spatial coordinates of the UAV swarm using a simulated annealing algorithm. The objective function is defined as a weighted sum of the coverage blind spot area and the communication load balancing degree: ; Parameter definition: : No. The target signal strength of each grid; The actual coverage intensity of the drone swarm is calculated using a signal propagation model. : Standard deviation of the communication load of a drone; =0.3: Weighting coefficient.

[0052] The initial annealing temperature is set to =1000, cooling coefficient =0.95, number of iterations 1000.

[0053] Dynamic channel allocation: The dynamic channel allocation unit adopts a hybrid architecture of generative adversarial network (GAN) and deep reinforcement learning (DRL).

[0054] The generator is a 3-layer convolutional neural network (CNN). The input is the spectrum occupancy matrix and smoke concentration gradient data, and the output is the candidate channel allocation strategy matrix.

[0055] The discriminator is a two-layer Long Short-Term Memory (LSTM) network. The input is the channel allocation strategy, and the output is a collision probability assessment value. .

[0056] The reward function of DRL is defined as: ; Parameter definition: =10Gbps =200: System maximum throughput and tolerable latency; The weighting coefficients are dynamically adjusted using the online gradient descent method.

[0057] Drone trajectory optimization: The UAV trajectory optimization unit calculates the cooperative flight path using a differential game model.

[0058] Specifically, the utility function of drone i is defined as: ; Parameter definition: : Time Drone Signal-to-noise ratio; =20dB: Reference signal-to-noise ratio; Energy consumption model, including flight kinetic energy and communication power consumption; =0.4: Energy consumption weighting coefficient, based on remaining power. Dynamic adjustment ( =0.5·(1− / ).

[0059] The Nash equilibrium solution uses backward induction, with a time step of [missing information]. =0.1s.

[0060] When the environmental perception module detects a fire spread rate > 2 m / s, the communication optimization processing module triggers the emergency mode: The signal propagation modeling unit reduces the mesh precision to 2m×2m and increases the calculation speed by 4 times; The UAV layout optimization unit skips the fine search stage of the simulated annealing algorithm and directly uses a greedy algorithm to select the local optimum. The dynamic channel allocation unit employs a fixed frequency band allocation strategy to prioritize the voice communication link of firefighter terminals.

[0061] Emergency Command Center Implementation: As the decision-making hub of the fire emergency communication system, the emergency command center receives and processes real-time fire data from the drone relay network, and generates dynamic rescue commands based on the optimization results of the communication optimization processing module. This module ensures high reliability and low latency in command issuance through multi-level data fusion and redundant communication link design, and forms a closed-loop control with the firefighter terminals.

[0062] In this embodiment, the emergency command center includes a data receiving unit, a digital twin system, an instruction generation unit, and redundant communication links, the technical details of which are as follows: Data reception and preprocessing: The data receiving unit acquires fire data transmitted by the UAV relay network through a fiber-optic-satellite dual-redundant link.

[0063] Specifically, the fiber optic link uses single-mode fiber with a transmission rate of 10Gbps and a latency of ≤1ms; the satellite link uses LDPC-coded Ku-band carriers with a throughput of 50Mbps and a latency of ≤500ms.

[0064] The data packet format is a custom binary protocol, containing timestamps, drone IDs, temperature field matrices, smoke concentration matrices, and channel status information.

[0065] When the fiber optic link is interrupted, the satellite link automatically switches to the primary channel, increasing the data packet compression rate to 70% and employing the Zstandard real-time compression algorithm.

[0066] Digital twin system modeling The digital twin system uses the Unity3D engine to build a three-dimensional dynamic model of the fire scene.

[0067] The fire spread prediction algorithm uses a cellular automata model, and the update rule is as follows: ; Parameter definition: : Real-time fire state matrix; Temperature field data; Smoke concentration data; Air velocity field data; : The state transition function is obtained by training with historical fire data.

[0068] Predict the high-temperature area and escape route in the next 5 minutes, with a rendering accuracy of ≤0.5m.

[0069] Instruction generation and encryption: The rescue command generation unit employs a dynamic priority scheduling algorithm. The command priority calculation formula is as follows: ; Parameter definition: =0.5, =0.3, =0.2: Weighting coefficient; : The highest temperature in the target area (°C); Smoke concentration (mg / m³) 3 ); Number of people trapped.

[0070] The instruction data is encrypted using AES-256-GCM, and the key is updated every 30 seconds via a quantum random number generator.

[0071] Redundant link switching logic: The decision to switch between fiber optic and satellite links is based on link quality assessment metrics: ; Parameter definition: Signal-to-noise ratio (dB); Bit error rate; Delay (ms).

[0072] when When <0.1, enable dual-link parallel transmission; otherwise, select... High-value links.

[0073] Human-computer interaction interface design: The command center console is equipped with a multi-screen display system. The main screen displays a real-time 3D model of the fire scene, while the auxiliary screen displays a heat map of the channel status and the trajectory of the drone.

[0074] The color mapping rules for heatmaps are as follows: Red: Signal-to-noise ratio < 10dB; Yellow: 10dB≤SNR<20dB; Green: SNR≥20dB.

[0075] Operators can quickly delineate key monitoring areas using gesture recognition devices (such as Leap Motion) and trigger a swarm of drones to focus and scan.

[0076] Hardware fault tolerance and heat dissipation: The server rack uses a liquid cooling system with a 50% ethylene glycol aqueous solution as the coolant, with a flow rate of 10L / min, to ensure that the CPU temperature is ≤60℃.

[0077] The primary and backup servers employ a heartbeat detection mechanism, with a switching latency of ≤200ms, a data synchronization cycle of 1s, and a synchronization error of <1ms.

[0078] Adaptive feedback module: The adaptive feedback module, serving as the closed-loop control unit of the fire emergency communication system, monitors the communication performance indicators of the UAV relay network in real time and forms a dynamic adjustment mechanism with the communication optimization processing module. This module ensures rapid system response and maintains communication link stability during sudden changes in the fire scene environment through multi-threshold joint decision-making and priority triggering strategies.

[0079] In this embodiment, the adaptive feedback module includes a performance monitoring unit, a threshold decision unit, and an optimization triggering unit, the technical details of which are as follows: Performance monitoring metrics: In some embodiments, the performance monitoring unit acquires real-time data from the UAV relay network via a CAN bus interface, including: Signal-to-noise ratio (SNR): Calculated based on receiver signal strength indication (RSSI) and noise floor, with a sampling frequency of 10Hz; End-to-end latency: Measured by timestamp difference, with an accuracy of ±0.1ms; UAV position offset: Calculate the Euclidean distance deviation by combining GPS positioning data with preset trajectory coordinates.

[0080] The signal-to-noise ratio monitoring formula is: ; Parameter definition: Signal power, output by the receiver's AGC circuit; Noise power, obtained by measuring background noise when the transmitter is off.

[0081] Dynamic threshold setting: The threshold decision unit dynamically adjusts the trigger threshold based on historical fire data and real-time input from the environmental perception module.

[0082] ; Parameter definition: =20dB: Basic signal-to-noise ratio threshold; Current smoke concentration (mg / m³) 3 ); Temperature gradient (°C / m); =0.05, =0.03: Correction factor, calibrated using experimental data.

[0083] When the temperature gradient is greater than 10℃ / m The value is dynamically adjusted to 0.05.

[0084] Triggering logic and priority The optimized triggering unit adopts a multi-level priority triggering mechanism: Level 1 trigger: Signal-to-noise ratio is below the threshold for 3 consecutive sampling periods, or delay is >100ms; Level 2 trigger: The drone's position offset is greater than 5m and lasts for 10 seconds; Level 3 trigger: Communication link interruption count > 5 times / minute.

[0085] Specifically, a Level 1 trigger immediately restarts the layout optimization and channel allocation process of the communication optimization processing module; a Level 2 trigger only initiates trajectory optimization; and a Level 3 trigger switches to a backup communication frequency band.

[0086] ; Parameter definition: =0.5, =0.3, =0.2: Weighting coefficient; =300℃: The maximum temperature that the system can withstand.

[0087] when When the value is greater than 0.8, it is determined to be a system-level fault, triggering a global re-initialization process.

[0088] Hardware fault tolerance: The performance monitoring unit adopts a dual-core redundant architecture, with the primary and backup processors synchronizing data via the PCIe bus, with a synchronization period of 1ms.

[0089] The threshold decision unit integrates an FPGA chip to achieve nanosecond-level real-time response with an interrupt latency of <10μs.

[0090] Interaction with the communication optimization module: The optimization trigger command is encapsulated into a UDP packet via Ethernet protocol, which includes the trigger type, priority and associated UAV ID.

[0091] After receiving an instruction, the communication optimization processing module inserts it into the task queue according to priority, with a scheduling interval of ≤50ms.

[0092] 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 claims and their equivalents.

Claims

1. A fire emergency communication system, characterized in that, include: The environmental sensing module is used to collect data on the temperature field, smoke concentration, air velocity field, and building structure at the fire scene, and transmit the data to the communication optimization and processing module via the LoRaWAN protocol. The drone relay network includes multiple drones equipped with multi-band communication devices. The drones support hybrid links of millimeter wave, 5G private network and satellite communication, and the communication link switching is dynamically triggered based on smoke concentration and temperature field data. The communication optimization processing module is used to jointly optimize the communication performance of the UAV swarm based on the data from the environmental perception module, including signal propagation modeling, UAV layout optimization, dynamic channel allocation, and UAV trajectory optimization. The emergency command center is used to receive and process real-time fire data, generate dynamic rescue instructions, and send them to firefighter terminals via a fiber-optic-satellite dual-redundant link. The adaptive feedback module is used to monitor communication performance indicators in real time and is connected to the communication optimization processing module via the CAN bus. When the indicator exceeds the preset threshold, it triggers a re-optimization process.

2. The fire emergency communication system according to claim 1, characterized in that, The environment sensing module includes: Infrared thermal imaging arrays are used to collect temperature field data at fire scenes; LiDAR is used to acquire three-dimensional point cloud data of building structures; Multispectral sensors are used to detect smoke concentration and airflow field. The data fusion unit is used to fuse temperature field, smoke concentration, air velocity field and building structure data into a unified environmental model, and connects to the LoRaWAN gateway via RS-485 bus to transmit the data to the communication optimization processing module.

3. The fire emergency communication system according to claim 1, characterized in that, In the aforementioned drone relay network, each drone includes: The millimeter-wave communication unit operates in the 28GHz band and supports beamforming and dynamic angle adjustment. The 5G private network communication unit supports dynamic adjustment of TDD time slot allocation for high-bandwidth data transmission. The satellite communication unit, with a built-in parabolic antenna and LDPC error correction encoder, is used for emergency communication in extreme environments. The link switching control unit is used to dynamically select the communication link based on smoke concentration and temperature field data. When the smoke concentration exceeds 50 mg / m³, the control unit will switch the link to the next communication link. 3 At that time, the satellite communication unit should be activated first.

4. A fire emergency communication system according to claim 1, characterized in that, The communication optimization processing module includes: The signal propagation modeling unit is used to receive fused data from the environmental perception module via an Ethernet interface and construct a wireless signal propagation model based on the ray tracing method. The UAV deployment optimization unit is used to solve for the optimal spatial coordinates of the UAV swarm using a simulated annealing algorithm based on the signal coverage requirements of the target area. The dynamic channel allocation unit is used to allocate channel resources through a deep reinforcement learning algorithm. The algorithm's reward function includes a weighted sum of throughput and end-to-end latency. The drone trajectory optimization unit is used to plan the cooperative flight path of a drone swarm through a multi-agent game model.

5. A fire emergency communication system according to claim 1, characterized in that, In the signal propagation modeling unit, the wireless signal propagation model is constructed using the ray tracing method, and the model parameters include: The linear correction factor for signal attenuation due to temperature gradient; The nonlinear correction coefficient for the effect of smoke concentration on signal absorption; The dynamic correction factor of air velocity field on signal propagation direction.

6. A fire emergency communication system according to claim 1, characterized in that, In the UAV layout optimization unit, the signal coverage requirements of the target area are divided by a hexagonal grid. The grid side length is inversely proportional to the fire spread speed, and the optimal spatial coordinates of the UAV swarm are solved by a simulated annealing algorithm.

7. A fire emergency communication system according to claim 1, characterized in that, The dynamic channel allocation unit adopts a hybrid architecture of generative adversarial networks and deep reinforcement learning, wherein: The generator uses a convolutional neural network to generate candidate channel allocation strategies; The discriminator assesses the channel collision probability through the spectrum sensing circuit and feeds it back to the generator for iterative optimization.

8. A fire emergency communication system according to claim 1, characterized in that, The UAV trajectory optimization unit coordinates the flight paths of the UAV swarm through a multi-agent game model, and the game payoff function includes: The signal-to-noise ratio improvement rate and weighting coefficients are dynamically adjusted according to the fire temperature. The energy consumption reduction rate, with a weighting coefficient, is dynamically adjusted based on the remaining battery power of the drone.

9. A fire emergency communication system according to claim 1, characterized in that, The communication performance metrics monitored in real time by the adaptive feedback module include: The signal-to-noise ratio is set to trigger re-optimization when it drops by more than 10%, and the threshold is set based on statistical analysis of historical fire data. The latency is set to trigger a re-optimization when it exceeds 100ms. The threshold is set according to the business requirements of the firefighter terminal. The drone's position offset will trigger a re-optimization when it exceeds 5 meters.

10. A fire emergency communication system according to claim 1, characterized in that, The emergency command center is connected to the communication optimization and processing module via a fiber-optic-satellite dual-redundant link. The fiber-optic link is used for high-bandwidth data transmission, and the satellite link is used for emergency communication in extreme environments.

Images