A built-in intelligent AI fire scene command search and rescue collaborative control method and system for a fire rescue mask

Abstract

The application relates to the technical field of intelligent fire-fighting emergency rescue intelligent equipment, and discloses a fire rescue mask built-in intelligent AI fire scene command search and rescue collaborative control method and system, which comprises the following steps: step one, device power-on initialization and module self-checking; step two, real-time collection of multi-source sensing data; step three, edge AI data fusion and feature extraction; step four, AI intelligent research and decision making; step five, decision result pushing screen display visualization; and step six, risk grading voice early warning and interactive broadcast. The fire rescue mask built-in intelligent AI fire scene command search and rescue collaborative control method and system realize local edge AI calculation and multi-source sensing data fusion at the mask end, do not need to rely on the cloud and the public network, can stably complete intelligent research and decision making such as life heat source identification, environment risk assessment and path planning under a network interruption environment, and effectively solve the problems of high communication dependence and decision lag of a traditional system.

Description

Technical Field

[0001] This invention relates to the field of intelligent equipment technology for smart fire emergency rescue, specifically to a method and system for intelligent AI-based fire scene command, search and rescue collaborative control built into a fire rescue mask. Background Technology

[0002] In modern fire rescue operations, the increasing complexity of building structures, the diversification of fire scenarios, and the extreme nature of fire environments have become the norm. Multiple dangerous factors such as dense smoke, high temperatures, toxic gases, and building collapses are intertwined, placing extremely high demands on firefighters' on-site perception, route decision-making, life search and rescue, and coordinated command capabilities.

[0003] The existing equipment and command system have significant shortcomings: the on-site perception dimension is limited, making it difficult to simultaneously collect multi-source data such as temperature and humidity, gas concentration, heat source distribution, and positioning attitude; intelligent judgment relies heavily on cloud computing and public network communication, and cannot operate independently after network outages at fire sites, resulting in delayed and unreliable decision-making; face masks lack visual displays and noise-resistant voice interaction, leading to untimely risk warnings and slow emergency response; there is a lack of unified coordination and dispatch among rescue personnel, which can easily lead to problems such as repeated search and rescue, missed areas, and route conflicts; positioning is prone to drift, communication lacks self-organizing network capabilities, and data and command transmission stability is poor.

[0004] These problems lead to firefighters relying on experience for on-site decision-making, incomplete information acquisition, low coordination efficiency, and untimely evacuation, seriously affecting rescue effectiveness and personnel safety. Existing technologies cannot achieve local edge AI computing on masks, multi-sensor data fusion, intelligent route planning, voice command control, Mesh self-organizing network collaboration, and highest priority emergency evacuation control, making it difficult to adapt to the real-time command, precise search and rescue, and safety assurance needs of complex fire scenes.

[0005] Based on this, it is necessary to propose a method and system for fire rescue mask with built-in intelligent AI for fire command, search and rescue collaborative control. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for integrated intelligent AI fire command and search and rescue coordination control in a fire rescue mask, which solves the problems mentioned in the background technology.

[0007] This invention provides the following technical solution: a method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask, comprising the following steps:

[0008] Step 1: Power-on initialization and module self-test;

[0009] Step 2: Real-time acquisition of multi-source sensing data;

[0010] Step 3: Edge AI Data Fusion and Feature Extraction;

[0011] Step 4: AI intelligently analyzes and generates decisions;

[0012] Step 5: Visualize the decision results displayed on the screen;

[0013] Step Six: Risk-based voice alerts and interactive broadcasts;

[0014] Step 7: Voice command recognition and execution control;

[0015] Step 8: Wireless network transmission and command sending / receiving;

[0016] Step 9: Coordinated scheduling of multiple rescue nodes;

[0017] Step 10: Prioritize emergency evacuation and control;

[0018] Step four includes fire type determination, automatic calculation of fire extinguishing water volume, generation of fire extinguishing plan, planning of attack route, generation of emergency evacuation route, and division of search and rescue area. Step ten is the highest priority execution logic of the system.

[0019] Preferably, the data collected in step two includes ambient temperature and humidity, smoke concentration, toxic gas concentration, oxygen concentration, laser ranging, infrared thermal imaging, audio and video, heat source distribution, fused positioning location, attitude, and voice commands.

[0020] Preferably, in step three, the edge AI performs human life heat source recognition, obstacle and passage recognition, environmental risk level calculation, speech noise reduction and semantic feature extraction. All analyses are completed locally on the mask and do not rely on the cloud.

[0021] Preferably, the voice commands in step seven include route planning, life search, water volume calculation, evacuation navigation, display switching, intercom control, and equipment adjustment. After AI parsing, the corresponding modules are directly driven to execute and feedback is provided to both ends.

[0022] Preferably, step eight uses a fire-fighting dedicated Mesh self-organizing network, which supports multi-hop relay, automatic link switching, encrypted transmission, and breakpoint resume transmission to realize data backhaul and command issuance.

[0023] Preferably, step nine, multi-rescue node collaboration, includes automatic allocation of search and rescue areas, real-time sharing of personnel locations, synchronization of task progress, and coordination of team action sequences to avoid repeated search and rescue and omission of areas.

[0024] Preferably, the emergency evacuation mode triggering conditions in step ten include high temperature, open flame, collapse, excessive toxic gas, and abnormal vital signs of personnel. After triggering, the evacuation route is forcibly displayed, continuous alarms are issued, danger information is broadcast across the network, and the transmission location is locked.

[0025] A fire rescue mask with built-in intelligent AI fire command and search and rescue collaborative control system includes a fire mask body, a main control unit, an edge AI module, a multi-source sensing module, an intelligent display module, a voice interaction module, a wireless Mesh communication module, a positioning module, a storage module, a power supply module, and an audible and visual warning module. All modules are embedded and sealed inside the mask without damaging the original protective structure.

[0026] Preferably, the multi-source sensing module includes an infrared thermal imaging unit, a laser ranging unit, a gas sensor, a temperature and humidity sensor, a smoke sensor, and an audio and video acquisition unit; the voice interaction module includes noise reduction and sound pickup, voice recognition, voice broadcasting, and two-way intercom units.

[0027] Preferably, the positioning module adopts UWB + inertial navigation fusion positioning, the wireless mesh communication module supports multi-level relay and operation without public network, and the screen display module supports high-brightness enhanced display in dense smoke environment.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. This type of fire rescue mask incorporates an intelligent AI-powered fire scene command and search and rescue collaborative control method and system. It achieves local edge AI computing and multi-source perception data fusion at the mask end, without relying on the cloud or public network. It can stably complete intelligent judgments such as life heat source identification, environmental risk assessment, and path planning in a network-off environment, effectively solving the problems of high communication dependence and decision lag in traditional systems. By integrating temperature and humidity, gas, thermal imaging, laser ranging, positioning and other sensing units, it comprehensively collects multi-dimensional information of the fire scene. With the help of local AI, it can quickly complete feature extraction and risk calculation, significantly improving firefighters' real-time perception capabilities of complex fire scenes and providing reliable data and decision support for safe operations and precise search and rescue.

[0030] 2. This type of fire rescue mask incorporates an intelligent AI-powered fire scene command and search and rescue collaborative control method and system, constructing an efficient human-computer interaction system with voice interaction, high-brightness screen display, and tiered early warning. It supports offline voice command control for route planning, life search, water volume calculation, etc. The execution results are displayed on the screen and provided with feedback from both the voice and the screen. The screen display is automatically enhanced in dense smoke environments. A four-level voice broadcast strategy is adopted for risk levels, and emergency evacuation commands are responded to with the highest priority, significantly improving information transmission efficiency and ease of operation. At the same time, it adopts UWB and inertial navigation fusion positioning, ensuring continuous and drift-free positioning, solving the problems of easy interruption and low accuracy of traditional positioning in indoor fire scenes.

[0031] 3. This type of fire rescue mask incorporates an intelligent AI-powered fire scene command and search and rescue collaborative control method and system. Through a dedicated fire-fighting mesh self-organizing network, it achieves full-domain collaboration among multiple rescue nodes, supports multi-hop relay, encrypted transmission, and breakpoint resume. It can automatically allocate search and rescue areas, share personnel locations, and synchronize task progress, avoiding repeated searches and omissions. The emergency evacuation mode has the highest execution priority. When dangers such as high temperature or excessive toxic gases are triggered, it forcibly pushes evacuation routes, broadcasts alarms across the entire network, and locks and transmits location data, providing comprehensive protection for firefighters. All modules are embedded and sealed, without damaging the original protective structure of the mask, balancing intelligence and protective reliability. Attached Figure Description

[0032] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0034] Figure 2 This is a schematic diagram of the system flow of the present invention;

[0035] Figure 3 This is a schematic diagram of the hardware structure of the face mask of the present invention. Detailed Implementation

[0036] 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.

[0037] Please see Figure 1-3 A method for integrated intelligent AI-powered fire scene command and search and rescue collaborative control in a fire rescue mask includes the following steps:

[0038] Step 1: Power-on initialization and module self-test of the device.

[0039] Specifically, after the built-in system of the fire mask is powered on, the main control unit sequentially performs hardware continuity tests, function validity verification, communication link connectivity tests, and AI model integrity verification on the display module, detection and search module, voice interaction module, wireless communication module, power supply module, and storage module. If all self-tests pass, the system enters the normal operation mode. If any module malfunctions, the corresponding audio and visual prompts are activated according to the fault level, and the system is prohibited from entering the fire scene operation state until the fault is cleared and the self-test passes again.

[0040] Step 2: Real-time acquisition of multi-source sensing data.

[0041] Specifically, data is collected synchronously through a sensor array built into the mask:

[0042] Environmental data: on-site temperature, relative humidity, smoke concentration, concentration of toxic and harmful gases, and oxygen concentration; Detection data: laser ranging values, infrared thermal imaging images, real-time audio and video streams, and heat map of heat source distribution; Positioning data: UWB / inertial navigation fusion positioning coordinates, firefighter's body posture, direction of travel, and speed; Human-computer interaction data: voice control commands, physical button operations, and two-way intercom audio signals. All sensing data is continuously collected at a fixed frequency to ensure real-time and complete fire scene information.

[0043] Step 3: Edge AI Data Fusion and Feature Extraction

[0044] Specifically, the face mask's built-in edge AI module performs local real-time fusion processing and feature extraction of multi-source sensing data, all completed on the face mask itself, without relying on cloud computing or a public network environment, and can still operate stably even when offline.

[0045] Based on infrared thermal imaging, intelligent identification of human life heat sources is performed to accurately locate the coordinates and number of trapped personnel; structural elements such as obstacles, walls, passages, doors, windows, and collapsed areas are identified from visual images; environmental parameters are comprehensively calculated to output the fire scene environmental risk level; and noise reduction, endpoint detection, and semantic feature extraction are performed on voice signals to improve the robustness of command recognition.

[0046] Furthermore, a comprehensive fire scene environmental risk assessment algorithm and a life heat source identification confidence algorithm are used to achieve risk assessment and life identification. The assessment algorithm formula is as follows: In the formula: This is a comprehensive risk index; This is the normalized value for temperature; This represents the normalized value of the toxic gas concentration. This represents the normalized value of the smoke concentration. This is the normalized value for oxygen concentration; These are weighting coefficients, all of which are positive numbers between 0 and 1, and the sum of the coefficients is 1. They can be preset and adjusted according to scenarios such as building fires, chemical fires, and forest fires. The confidence algorithm formula is as follows: In the formula: Confidence in life; For thermal imaging human feature matching degree; For distance attenuation compensation; Weights for motion features; These are weighting coefficients, all of which are positive numbers between 0 and 1, and the sum of the coefficients is 1. They can be dynamically adjusted according to the recognition distance and environmental interference.

[0047] Step 4: AI intelligently analyzes and generates decisions.

[0048] Specifically, based on the data fusion results, edge AI automatically performs multi-dimensional intelligent analysis and outputs standardized decision results:

[0049] The system determines the fire type, combustion intensity, high-temperature area, and dangerous spread range; automatically calculates and recommends the amount of water needed for firefighting based on the burning area, fire temperature, and space volume; matches the fire type with on-site conditions to generate the optimal intelligent firefighting plan; dynamically plans safe attack routes based on building structure, real-time obstacles, and risk distribution; generates multiple switchable emergency evacuation routes based on the nearest safe exit and passageway accessibility; and automatically divides search and rescue responsibility areas according to the location distribution of rescue personnel across the network to avoid duplication and omissions.

[0050] Furthermore, a fire extinguishing water consumption calculation algorithm, a dynamic path planning cost function algorithm, and a multi-search and rescue node collaborative area division algorithm are used to realize water consumption calculation, path planning, and area collaboration. The calculation algorithm formula is as follows: In the formula: Recommended water consumption; This is an estimated value for the burning area; Temperature at the fire scene; To estimate the time required for firefighting; The scenario coefficient can be set from 0.1 to 1.0 based on the fire type, building structure, and combustible material. The cost function algorithm formula is as follows: In the formula: This represents the total cost of the path. This is the path length; This represents the path risk value. This represents the blocking probability. The weighting coefficients are used to determine the region partitioning algorithm formula. In the formula: Region compatibility; Personnel location density; As a cost of distance; For environmental risks; For weight coefficients, the weight coefficients in the cost function algorithm formula and the region partitioning algorithm formula are all positive numbers between 0 and 1, and the sum of the coefficients is 1;

[0051] Step 5: Decision results are pushed to the screen and visualized.

[0052] Specifically, AI-generated attack / evacuation routes, real-time locations, fire-fighting facilities, environmental parameters, trapped personnel markers, teammate locations, and command and dispatch information are uniformly pushed to the miniature high-brightness screen built into the mask. The screen dynamically refreshes, highlights key information, and displays content in a hierarchical manner. When dense smoke or low-light environments are detected, the screen automatically increases brightness and contrast to ensure that firefighters can clearly read information under low visibility conditions.

[0053] Step Six: Risk-based voice alerts and interactive broadcasts.

[0054] Specifically, the system uses a four-level voice broadcast strategy based on the comprehensive risk index:

[0055] Safety: Status prompts; Caution: Intermittent reminders; Danger: Continuous warnings; Emergency: Mandatory evacuation announcements, along with voice broadcasts of command instructions, task information, and Q&A results;

[0056] Step 7: Voice Command Recognition and Execution Control

[0057] Specifically, the system receives voice commands from firefighters in real time through a noise-reducing audio pickup unit. After semantic parsing by edge AI, it directly executes the corresponding operations, including: planning evacuation routes, initiating life search, displaying safety exits, calculating fire extinguishing water volume, switching display screens, activating intercom, calling the command center, and adjusting equipment parameters. The results of command execution are simultaneously fed back in both screen display and voice modes to ensure that the operation is perceptible and verifiable.

[0058] Furthermore, a speech command semantic confidence algorithm is employed to achieve highly robust speech command recognition. The confidence algorithm formula is as follows: In the formula: For instruction confidence; For speech signal-to-noise ratio; Noise suppression degree; For instruction template matching degree; These are weighting coefficients, all of which are positive numbers between 0 and 1, and the sum of the coefficients is 1, ensuring the accuracy of identification in high-noise environments at fire scenes;

[0059] Step 8: Wireless network transmission and command reception.

[0060] Specifically, data and command transmission is achieved through a dedicated fire-fighting mesh self-organizing network: Uplink: Real-time transmission of environmental data, audio and video, location information, equipment status, and search and rescue progress; Downlink: Receiving tasks, route adjustments, dispatch instructions, and global early warning information from the command center; Supports multi-hop relay, automatic link switching, breakpoint resumption, and high-strength encrypted transmission. When communication is interrupted, the system automatically caches the data locally and resumes transmission automatically after communication is restored, ensuring that data is not lost and operation records are complete.

[0061] Step 9: Coordinated scheduling of multiple rescue nodes.

[0062] Specifically, based on real-time data from all Mesh terminals, efficient collaboration among multiple rescue nodes is achieved: search and rescue areas are automatically allocated to avoid duplication and blind spots; teammates' locations, statuses, and progress are shared in real time; action sequences between teams are coordinated; and instructions are issued and execution statuses are transmitted in a unified manner.

[0063] Step 10: Prioritize emergency evacuation and control.

[0064] Specifically, when the system detects any of the following trigger conditions: high temperature, approaching open flame, risk of collapse, excessive toxic gas, or abnormal vital signs of personnel, it immediately enters the highest priority emergency evacuation mode. This mode is the highest execution logic of the system. Once triggered, it immediately interrupts all non-evacuation processes to prioritize personnel safety: the screen display is forcibly switched and the optimal evacuation route is locked; high-frequency audible and visual alarms and emergency evacuation announcements are activated; dangerous locations and evacuation tips are broadcast to the entire network; the real-time location of firefighters is locked and continuously transmitted back to facilitate external rescue. When sensors fail, the system automatically enters a redundant sensing and degraded operation mode and provides voice prompts to firefighters to operate with caution. When the positioning signal is lost, it automatically switches to pure inertial navigation to continue positioning, combined with thermal imaging to achieve relative positioning, ensuring that evacuation capabilities are not interrupted.

[0065] A fire rescue mask with a built-in intelligent AI fire command and search and rescue collaborative control system includes a fire mask body, a main control unit, an edge AI module, a multi-source sensing module, an intelligent display module, a voice interaction module, a wireless mesh communication module, a positioning module, a storage module, a power supply module, and an audible and visual warning module. All modules are embedded and sealed inside the mask without damaging the original protective structure.

[0066] Specifically, the fire-fighting mask body is made of flame-retardant, high-temperature resistant, and impact-resistant engineering plastics and medical-grade silicone sealing materials. It has all the protective functions of a standard fire rescue mask and meets the relevant national standards for fire-fighting air respirator masks. The mask has an integrated modular installation compartment inside, which can seal and embed various electronic modules to achieve waterproof, dustproof, and smoke penetration prevention effects. The mask window is treated with an anti-fog, wear-resistant, and high-temperature resistant coating, which does not affect the visibility of the smart screen. At the same time, it is reserved for sensor acquisition holes, sound pickup holes, speaker holes, and antenna windows, and is sealed with a breathable but smoke-proof structure to maintain the overall protection level of the mask while ensuring the perception and interaction functions.

[0067] Main control unit: As the core control and scheduling hub of the entire system, it adopts an industrial-grade high-reliability main control chip, has strong adaptability to wide temperature working environment, and can stably cope with extreme temperature changes in the fire scene. It mainly completes system power-on initialization, self-test of each functional module, equipment status monitoring, fault diagnosis, task priority scheduling and full-process logic control. It uniformly schedules edge AI module, multi-source sensing module, wireless Mesh communication module, intelligent display module, voice interaction module, positioning module and other modules to complete data interaction and command execution. It has a multi-level hardware interrupt and emergency preemption mechanism to ensure that emergency evacuation commands are responded to with the highest priority. At the same time, it integrates hardware watchdog, over-temperature protection, over-current protection, under-voltage protection and other safety mechanisms. It can automatically reset or enter safe operation mode under abnormal conditions. It monitors power supply voltage, current and temperature in real time to ensure long-term stable operation of the system.

[0068] Edge AI Module: This is the core of local intelligent computing on the face mask. It is equipped with an AI acceleration processor and supports real-time data inference, multi-source data fusion, multi-dimensional feature extraction, and intelligent judgment and decision-making. All computing tasks are completed locally on the face mask without relying on cloud computing or public network environment. It can still run independently and stably when the network is disconnected. This module can run lightweight AI models such as life heat source recognition, obstacle and passage recognition, fire scene environment risk assessment, dynamic path planning, and voice semantic parsing locally. It has low computing latency and multi-task parallel processing capabilities. It can quickly output decision results such as fire type determination, fire extinguishing water consumption calculation, fire fighting plan generation, attack and evacuation route planning, and search and rescue area division. At the same time, it supports local encrypted storage and offline updates and upgrades of models to adapt to the intelligent operation needs of different fire scene scenarios.

[0069] Multi-source sensing module: Composed of arrays of multiple types of sensors, it can realize the full-area, real-time, and multi-dimensional data acquisition of the fire scene environment, providing underlying data support for AI analysis and system control. This module includes an infrared thermal imaging unit, a laser ranging unit, a multi-component gas sensor, a temperature and humidity sensor, a smoke sensor, a high-definition audio and video acquisition unit, and an attitude sensor. Among them, the infrared thermal imaging unit can penetrate smoke to identify human life heat sources, high-temperature areas, and hidden fire sources; the laser ranging unit can measure the distance between obstacles and spatial dimensions in real time; the gas sensor can monitor the concentration of carbon monoxide, hydrogen sulfide, oxygen, and combustible gases in real time; the temperature, humidity, and smoke sensors are used to collect environmental parameters and determine the combustion intensity and visibility level; the audio and video acquisition unit and attitude sensor can acquire on-site images, voice signals, and the movement posture of firefighters, comprehensively sensing the fire situation and personnel status.

[0070] Intelligent display module: The mask has a built-in miniature high-brightness visualization unit with a high-brightness display panel. It has the ability to automatically enhance the display in strong light and dense smoke environments. It can push and display information such as AI-generated attack routes, evacuation routes, location coordinates, environmental parameters, trapped personnel markers, teammate positions, risk levels, and equipment status in real time. The module supports dynamic screen refresh, highlighting of key information, and hierarchical display of content. In dense smoke and low light environments, it can automatically improve brightness and contrast. With wide viewing angle, anti-glare, and anti-fog treatment design, it allows firefighters to clearly read information at eye level without affecting their forward vision. At the same time, it can force full-screen display of evacuation routes according to evacuation instructions to ensure intuitive and efficient information transmission in emergency situations.

[0071] Voice Interaction Module: Designed for high-noise environments in fire zones, this module integrates noise reduction pickup, voice recognition, voice broadcasting, and two-way intercom units to achieve stable and efficient voice interaction between people and the system. It employs a dual-microphone array to effectively suppress background noise and breathing apparatus fan noise, improving voice signal clarity. Offline voice command recognition is achieved through local semantic parsing, supporting common operation commands such as route planning, life search, water volume calculation, evacuation navigation, display switching, intercom control, and equipment adjustment. After parsing, the corresponding modules are directly driven to execute and receive synchronous feedback. The voice broadcasting unit uses a high-volume noise-canceling speaker, outputting prompts, warnings, and command information according to risk level. The two-way intercom unit can be linked with a Mesh network to achieve stable point-to-point and point-to-multipoint intercom communication within the team.

[0072] Wireless Mesh Communication Module: Utilizing fire-fighting dedicated frequency band self-organizing network communication technology, it can achieve full-area connectivity and data transmission for multiple rescue nodes without the need for public networks, base stations, or cloud support. It supports multi-level multi-hop relay, automatic topology, automatic routing, and automatic link switching, featuring long communication distance, multiple access nodes, and strong stability. This module is responsible for uplink data transmission and downlink command reception. Uplink can transmit environmental data, audio and video streams, location information, equipment status, and search and rescue progress in real time. Downlink can receive command tasks, route adjustments, dispatch instructions, and full-area early warning information. It also supports high-strength encrypted transmission, breakpoint resumption, communication interruption caching, and retransmission functions to ensure the secure, continuous, and lossless transmission of fire scene data and commands.

[0073] Positioning Module: Employing a fusion positioning solution combining UWB (Ultra-Wideband) positioning and inertial navigation, this module achieves high-precision, continuous positioning in complex indoor and outdoor environments, addressing the issues of interruption and drift inherent in traditional positioning methods in building fire scenes. The UWB unit provides centimeter-level indoor positioning accuracy, while the inertial navigation unit, composed of an accelerometer, gyroscope, and magnetometer, can continuously calculate position even without a UWB signal. The fusion positioning algorithm automatically switches between and calibrates the two positioning modes, ensuring uninterrupted and drift-free positioning. The module outputs real-time 3D coordinates, direction of travel, speed, and attitude angles, with high-frequency data updates to meet the real-time requirements of search and rescue navigation and emergency evacuation. It also supports historical path backtracking.

[0074] Storage Module: Employs industrial-grade high-capacity, high-speed storage chips for local encrypted storage of system operation data. It supports full-process storage of information such as environmental perception data, positioning trajectory, audio and video streams, AI decision logs, and equipment operating status. The storage duration can cover the entire process of a typical fire rescue, meeting the requirements for complete operation records. The module features cyclic overwrite, breakpoint resume, power failure protection, and data encryption functions to prevent data loss, tampering, and leakage. After the rescue operation is completed, data can be exported through a dedicated interface for fire rescue review, practical training, accident investigation, and command process optimization, improving the standardization and traceability of rescue operations.

[0075] Power supply module: Provides a stable, safe, and long-lasting energy supply for the system. It adopts a high-temperature resistant lithium battery pack with wide-temperature discharge characteristics to adapt to fire scene environments. The continuous working time can meet the needs of conventional fire rescue operations. The module has a complete battery protection mechanism, including overcharge protection, over-discharge protection, overcurrent protection, over-temperature protection, and short-circuit protection, effectively avoiding electrical safety hazards. It also supports low voltage graded voice prompts, activates an early warning when the battery level is lower than a set threshold, and automatically enters a power-saving and risk-avoidance mode when the battery level is low to ensure the continuous operation of core functions. The power supply unit adopts a sealed and waterproof design, and the charging interface is equipped with a dustproof and smokeproof protective cap. It supports fast charging and convenient replacement, improving on-site deployment and continuous operation capabilities.

[0076] Audible and visual warning module: Used for system risk warning and emergency evacuation alarm. It adopts a synchronous audio and visual output method. The sound intensity is high and can effectively penetrate the complex noise of the fire scene to achieve long-distance, strong warning. The light indicator uses high-brightness LEDs, with green, yellow and red corresponding to normal state, warning state and dangerous emergency evacuation state, respectively. When the emergency evacuation mode is triggered, it activates a triple reminder of high-frequency light flashing, continuous buzzing and voice evacuation broadcast. At the same time, it can be linked with the wireless Mesh network to realize synchronous audio and visual alarms of all nodes in the network when danger is broadcast, so as to ensure that rescue personnel in the entire area can quickly perceive danger and respond to evacuation in a timely manner.

[0077] Example 1: Single-person search and rescue

[0078] Firefighters wearing masks entered the underground parking garage fire scene. The system completed its power-on self-test; collected thermal imaging, gas, ranging, and positioning data; local AI identified life-threatening heat sources and wall passages; automatically calculated water volume and generated firefighting and evacuation routes; pushed displays to the screen; provided voice warnings of risks; responded to the "search for trapped personnel" voice command; transmitted data back via Mesh relay; shared location with teammates; and automatically triggered evacuation prompts upon detecting high temperatures. The system operated stably throughout the entire process, even without a public network connection.

[0079] Example 2: Multi-shift Collaborative Command

[0080] Multiple teams enter the floor, the Mesh self-organizing network automatically connects, the command center divides the search and rescue area, AI uniformly plans the route, each terminal reports the progress in real time, the system avoids repeated search and rescue, when a certain point detects that the toxic gas exceeds the standard, the whole network synchronously issues an early warning and automatically adjusts the route, realizing full-area collaboration.

[0081] When the AI ​​detects a sudden rise in temperature or blockage in the passageway, it immediately enters the risk avoidance mode, forcibly displays the optimal evacuation route, continuously broadcasts "Evacuate immediately" in voice, and broadcasts the dangerous location to the command center and teammates to ensure that everyone evacuates quickly.

[0082] 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 method for integrated intelligent AI-powered fire scene command and search and rescue collaborative control in a fire rescue mask, characterized in that, Includes the following steps: Step 1: Power-on initialization and module self-test of the device; Step 2: Real-time acquisition of multi-source sensing data; Step 3: Edge AI Data Fusion and Feature Extraction; Step 4: AI intelligently analyzes and generates decisions; Step 5: Visualize the decision results displayed on the screen; Step Six: Risk-based voice alerts and interactive broadcasts; Step 7: Voice command recognition and execution control; Step 8: Wireless network transmission and command sending / receiving; Step 9: Coordinated scheduling of multiple rescue nodes; Step 10: Prioritize emergency evacuation and control; Step four includes fire type determination, automatic calculation of fire extinguishing water volume, generation of fire extinguishing plan, planning of attack route, generation of emergency evacuation route, and division of search and rescue area. Step ten is the highest priority execution logic of the system.

2. The method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: The data collected in step two includes ambient temperature and humidity, smoke concentration, toxic gas concentration, oxygen concentration, laser ranging, infrared thermal imaging, audio and video, heat source distribution, fused positioning location, attitude, and voice commands.

3. The method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: In step three, the edge AI performs human life heat source recognition, obstacle and passage recognition, environmental risk level calculation, speech noise reduction and semantic feature extraction. All analysis is completed locally on the mask and does not rely on the cloud.

4. The method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: The voice commands in step seven include route planning, life search, water volume calculation, evacuation navigation, display switching, intercom control, and equipment adjustment. After AI parsing, they directly drive the corresponding modules to execute and provide feedback to both ends.

5. A method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: Step eight adopts a fire-fighting dedicated Mesh self-organizing network, which supports multi-hop relay, automatic link switching, encrypted transmission, and breakpoint resume transmission to realize data backhaul and command issuance.

6. The method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: Step nine, multi-rescue node collaboration, includes automatic allocation of search and rescue areas, real-time sharing of personnel locations, synchronization of task progress, and coordination of team action sequences to avoid duplicate search and rescue and omission of areas.

7. The method for integrated intelligent AI fire scene command and search and rescue collaborative control in a fire rescue mask according to claim 1, characterized in that: The emergency evacuation mode in step ten is triggered by conditions including high temperature, open flame, collapse, excessive toxic gas, and abnormal vital signs of personnel. Once triggered, the evacuation route is forcibly displayed, alarms are continuously issued, danger information is broadcast across the network, and the transmission location is locked.

8. A fire rescue mask with a built-in intelligent AI fire scene command and search and rescue collaborative control system, characterized in that, It includes the fire-fighting mask body, main control unit, edge AI module, multi-source sensing module, intelligent display module, voice interaction module, wireless Mesh communication module, positioning module, storage module, power supply module and sound and light warning module. All modules are embedded and sealed inside the mask without damaging the original protective structure.

9. The intelligent AI fire scene command and search and rescue collaborative control system built into a fire rescue mask according to claim 1, characterized in that: The multi-source sensing module includes an infrared thermal imaging unit, a laser ranging unit, a gas sensor, a temperature and humidity sensor, a smoke sensor, and an audio and video acquisition unit; the voice interaction module includes noise reduction and sound pickup, voice recognition, voice broadcasting, and two-way intercom units.

10. The fire rescue mask with built-in intelligent AI fire scene command and search and rescue collaborative control system according to claim 1, characterized in that: The positioning module adopts UWB + inertial navigation fusion positioning, the wireless mesh communication module supports multi-level relay and operation without public network, and the screen display module supports high-brightness enhanced display in dense smoke environment.

Images