Intelligent perception and risk warning method for firefighters entering and exiting fire scene

Abstract

This invention proposes an intelligent sensing and risk warning method for firefighters entering and exiting fire scenes. The method includes: collecting firefighter identification, vital sign data, and real-time location signals from firefighters wearing wearable devices via RFID sensing nodes deployed in different areas of the fire scene; automatically determining the firefighter's entry and exit status based on the mapping relationship between the timestamp sequence read by the RFID sensing nodes and the node location identifier; and constructing a fatigue-risk index model by correlating the firefighter's entry and exit status with real-time vital sign data and historical time-series data to calculate the current real-time risk index of the firefighter and provide graded warnings based on index thresholds. This invention achieves automated, non-intrusive, and accurate recording of firefighters entering and exiting fire scenes, completely solving the problems of missed or incorrect recording during manual counting; it establishes a correlation mechanism between vital sign data and the operational scenario status, providing clear contextual basis for risk assessment, improving the accuracy of warnings, and ensuring the operational safety of firefighters.

Description

Technical Field

[0001] This invention relates to the field of IoT intelligent monitoring technology, and in particular to a method for intelligent sensing and risk warning of firefighters entering and leaving a fire scene. Background Technology

[0002] In firefighting and rescue operations, the personal safety of firefighters is always the core concern of command and decision-making. With the acceleration of urbanization, complex fire scenarios such as high-rise buildings, underground spaces, large complexes, and hazardous chemical storage sites are becoming increasingly common. The firefighting environment is characterized by high temperatures, dense smoke, low visibility, complex spatial structures, and severe communication obstruction, which places extremely high demands on firefighters' on-site situational awareness, vital sign monitoring capabilities, and the emergency decision-making capabilities of the command center.

[0003] In recent years, the Internet of Things (IoT), wearable devices, wireless communication, and big data analytics have developed rapidly, and various intelligent sensing devices have been gradually applied in the firefighting field. For example, firefighters can wear smart bracelets or vests integrating heart rate sensors, temperature sensors, and motion sensors to transmit data back to the command platform via mobile communication networks. However, in real fire scene operations, the application of these technologies still faces the following systemic shortcomings: (1) The entry and exit management is crude and lacks automated sensing methods. Currently, personnel entry and exit management at fire scenes mainly relies on manual headcounts by safety officers or verbal reports via walkie-talkie. In the noisy, chaotic, and obstructed environment of a fire scene, information such as the time of firefighters' entry and exit, and their identities, is easily missed, misrecorded, or delayed. Especially in scenarios involving multiple fire brigades operating jointly and frequent personnel rotations, the command center struggles to promptly and accurately ascertain the actual number of people and their specific identities inside the fire scene, posing a significant safety hazard of personnel being "out of contact" or "trapped undetected."

[0004] (2) Vital signs monitoring is disconnected from the work scenario, and risk assessment lacks contextual relevance.

[0005] While existing wearable devices can collect physiological parameters such as heart rate, blood oxygen, and body surface temperature in real time, this data is usually presented independently as numerical values ​​or curves, lacking correlation analysis with the crucial scenario of whether a firefighter is "inside a fire." In fact, the increase in heart rate caused by physical recovery in a firefighter's rest area or safe zone is entirely different in risk level and emergency response strategy from an increase in heart rate caused by high temperature, dense smoke, toxic gases, or psychological stress inside a fire. Current technology cannot distinguish between these scenario differences, making it difficult for the command center to accurately determine whether abnormal vital sign data constitutes a genuine operational risk when received. This easily leads to either frequent false alarms or missed alarms, reducing the commander's trust in the system.

[0006] (3) The communication link is single and fragile, and the data transmission reliability is insufficient in extreme environments.

[0007] Fire environments are highly uncertain and complex. Reinforced concrete walls, metal structures, and equipment pipelines inside buildings exert strong attenuation and obstruction effects on wireless signals; high temperatures can damage communication base station equipment; and dense smoke and dust further exacerbate signal scattering and absorption. In such environments, any single communication method—whether relying on public 4G / 5G links or dedicated wireless network links on specific frequency bands—is highly susceptible to signal interruption or base station damage, resulting in data loss and creating monitoring blind spots.

[0008] (4) Insufficient data value mining makes it difficult to achieve early warning.

[0009] Most existing fire monitoring systems primarily function as real-time data displays, triggering alarms only when heart rate exceeds a preset safety threshold. They lack the ability to store, analyze, and mine historical time-series data. However, firefighters' physical exhaustion and physiological instability are often a gradual process—trend characteristics such as a sustained increase in heart rate, prolonged recovery time, and increased volatility are more valuable for early warning than instantaneous absolute values. Current technology cannot predict the critical point of firefighters' physical exhaustion by analyzing historical data, leaving early warning mechanisms at the "post-event statistics" level, failing to achieve "pre-event warnings" and proactive intervention, and resulting in missed optimal evacuation opportunities.

[0010] In summary, existing technologies have significant shortcomings in areas such as firefighter entry and exit perception, vital sign and scene correlation analysis, reliable multi-link transmission, and intelligent early warning based on time-series data. There is an urgent need for a systematic solution that can achieve intelligent perception of firefighters entering and exiting fire scenes, deep correlation between vital sign data and operational status, reliable transmission through multiple communication links, and advanced early warning based on time-series data models. Summary of the Invention

[0011] This invention proposes an intelligent sensing and risk warning method for firefighters entering and exiting fire scenes, which solves the problems of existing technologies such as extensive management of entering and exiting fire scenes, disconnect between vital sign monitoring and operational scenarios, and difficulty in achieving early warning.

[0012] The technical solution of this invention is implemented as follows: The first aspect of this invention provides a method for intelligent sensing and risk warning of firefighters entering and exiting a fire scene, comprising the following steps: By deploying RFID sensing nodes in different areas of the fire scene, the identification, vital signs data and real-time location signals of firefighters wearing wearable devices are collected. Based on the mapping relationship between the timestamp sequence read by RFID sensing nodes and the node location identifier, the system automatically determines the status of firefighters entering and leaving the fire scene. By linking firefighters' entry and exit status from the fire scene with real-time vital signs data and historical time series data, a fatigue-risk index model is constructed to calculate the current real-time risk index of firefighters and to issue graded warnings based on the index threshold.

[0013] Specifically, the method for automatically determining the status of firefighters entering and leaving a fire scene includes the following steps: The system node topology is preset, and a mapping table between sensing node identifiers and region types is established, wherein the region types include safe zones, buffer zones, and danger zones; Wearable devices periodically broadcast RFID signals. After each sensing node captures the signal, it uploads a record to the server. The record includes the firefighter's identification, the sensing node's identification, and a timestamp. The server retrieves the current record and the previous valid record, and updates the firefighter's entry and exit status according to the following logic: If the area to which the sensing node in the current record belongs is a danger zone, and the area to which the sensing node in the previous valid record belongs is a safe zone or buffer zone, then the status is determined to be entering the fire scene. If the area to which the sensing node in the current record belongs is a safe zone or buffer zone, and the area to which the sensing node in the previous valid record belongs is a danger zone, then the status is determined to be leaving the fire scene. If the area in the current record has not changed from that in the previous valid record, the status is determined to be "in operation".

[0014] Furthermore, when a region jump occurs—that is, a direct jump from the safe zone to the danger zone or from the danger zone to the safe zone without a missing buffer record—the time difference between the current record and the previous valid record is calculated. ; If time difference Greater than or equal to the preset minimum travel time threshold If the move is deemed legitimate, data completion is performed, and a virtual buffer record is inserted, containing the timestamp of the virtual buffer record. Calculate using the following formula: ; in, The timestamp of the previous valid record. The preset buffer time ratio, ; If time difference Less than the minimum travel time threshold If the current record is misread, it will be discarded.

[0015] Specifically, the method for calculating the real-time risk index of firefighters and issuing graded warnings includes the following steps: The collected heart rate data is filtered and denoised to calculate the average heart rate per minute and obtain the current continuous fire operation duration. ; Extract the difference between the current heart rate and the resting heart rate. The slope of the linear regression of the heart rate within the most recent preset time window Safe heart rate limits and personalized warning thresholds configured according to age groups or individual health records. Calculate the risk index : ; in, This is the standard duration for continuous fire scene operations; This is a growth coefficient used to control the steepness of the risk increase; The threshold for the linear regression slope of the heart rate within the most recent preset time window; , , These are the weighting coefficients; The risk index is used to issue tiered warnings: if the risk index is less than the first threshold, it is considered a normal state; if the risk index is between the first and second thresholds, it is considered a state requiring attention; if the risk index is greater than the second threshold, it is considered a dangerous state, an evacuation alarm is sent, and the wearable device is vibrated; if the average heart rate is greater than the upper limit of the safe heart rate, it is directly considered a dangerous state.

[0016] Preferably, the method further includes a multi-link communication step: Periodically check the signal strength of the current link and the backup link. Packet loss rate and delay ; Calculate the link quality score for each communication link separately. The calculation formula is: ; in, , , These are the weighting coefficients; Preset switching threshold and recovery threshold, with the switching threshold being greater than the recovery threshold; If the link quality score of the current link is less than the recovery threshold and the link quality score of the backup link is greater than the handover threshold, then switch to the backup link; otherwise, maintain the current link.

[0017] Furthermore, if the quality score of all links is below the recovery threshold, local caching mode is enabled; When any communication link is detected to be restored, the data in the local cache queue is read and retransmitted to the server in the original timestamp order. After successful retransmission, the corresponding cache is cleared.

[0018] Preferably, the method further includes a time-series data storage step: Create a super table, which includes dynamic time-series data columns and static metadata tag columns. The dynamic time-series data columns include at least several of the following: timestamp, heart rate, body surface temperature, exercise status, and battery level. The static metadata tag columns include at least several of the following: firefighter identification, affiliated squadron, sensing node location, device ID, and indicator type. When writing data, the following sub-steps are executed: Wearable devices report data, and the data access gateway parses the reported data, extracting the tag field and time series field. The routing decision is made based on the extracted tag fields to determine whether there is an independent sub-table corresponding to the current firefighter's identifier and indicator type; If no corresponding independent sub-table exists, the independent sub-table will be created automatically. The independent sub-table inherits all column definitions of the super table and stores the value of the tag column as metadata, synchronously storing time-series data. At the same time, an in-memory inverted index is built based on the tag field to establish a mapping relationship between tag values ​​and sub-table identifiers. If a corresponding independent sub-table already exists, the parsed time-series data is directly written into that independent sub-table, and the in-memory inverted index is updated after the writing is completed. When performing a query, the target sub-table set is quickly located using an in-memory inverted index, and only sub-tables related to the query conditions are scanned.

[0019] A second aspect of the present invention provides an intelligent sensing and risk warning system for firefighters entering and exiting a fire scene, comprising: The perception layer includes wearable devices for firefighters and fire scene perception nodes. The wearable devices for firefighters integrate RFID tags, vital sign sensors, multi-mode communication modules, and local cache storage. The fire scene perception nodes include RFID readers and data aggregation gateways deployed at the boundaries of safe zones, buffer zones, and danger zones. Network layer: includes mobile communication network links and ad hoc network links, supporting dynamic handover and breakpoint resumption based on link quality scores; Application layer: includes data access gateway, time series database, data analysis engine and command visualization terminal; the data analysis engine is used to perform entry and exit status determination, multi-link communication switching decision and safety risk assessment, and push the results to command visualization terminal.

[0020] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method.

[0021] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention automatically collects firefighters’ identity information, vital signs data and real-time location signals by deploying RFID sensing nodes in different areas of the fire scene. Based on the mapping relationship between RFID timestamp sequence and node location identifier, it automatically determines the status of entering and leaving the fire scene and associates the status with real-time vital signs data and historical time series data to construct a fatigue-risk index model, calculate the real-time risk index and issue graded warnings. It realizes the automated, non-intrusive and accurate recording of firefighters entering and leaving the fire scene, and completely solves the problem of omissions and errors in manual counting. It establishes a correlation mechanism between vital signs data and the status of the work scene, so that risk judgment has a clear contextual basis, significantly improves the accuracy of warnings, and effectively protects the work safety of firefighters.

[0023] (2) This invention automatically determines the status of entering the fire scene, leaving the fire scene, or working based on the regional change logic of the current record and the previous valid record by pre-setting the node topology relationship of the safe zone, buffer zone, and dangerous zone. When a regional jump occurs, it determines the legal movement or misread by calculating the time difference and comparing it with the minimum crossing time threshold. For legal movements, it inserts the virtual buffer record to complete the data. This mechanism effectively solves the problem of record loss or jump caused by unstable fire scene signals, ensures the continuity and integrity of the entry and exit trajectory, and filters out misread signals, thereby improving the robustness and accuracy of status determination.

[0024] (3) This invention constructs a comprehensive risk index model by extracting the difference between the current heart rate and the resting heart rate, the linear regression slope of the heart rate in the most recent time window, and combining it with the duration of continuous fire operation. This model not only focuses on real-time vital signs values, but also pays more attention to the trend of vital signs changes. It can capture the accelerated process of firefighters' physical exertion and can identify the risk increase trend in advance before the heart rate reaches the absolute safety limit, thus achieving early warning and significantly reducing the probability of false alarms and missed alarms.

[0025] (4) This invention periodically detects the signal strength, packet loss rate, and delay of the current link and the backup link, calculates the link quality score by weighting, and makes dynamic switching decisions by setting switching thresholds and recovery thresholds. When the quality of all links is lower than the recovery threshold, the local caching mode is automatically activated to temporarily store the data in the non-volatile memory of the wearable device. After any communication link is restored, the data is retransmitted in the order of the original timestamp. This mechanism effectively solves the problem of easy interruption of a single communication link in the complex environment of the fire scene, realizes smooth switching between multiple links, avoids frequent link oscillations, and ensures zero data loss in extreme signal loss scenarios, thus ensuring the command center's ability to continuously monitor the status of firefighters.

[0026] (5) This invention uses a super table to define dynamic time-series data columns and static metadata tag columns, and automatically routes and creates independent sub-tables through tag fields to achieve physical data isolation and parallel writing; at the same time, it constructs an in-memory inverted index based on the tag fields to establish a mapping relationship between tag values ​​and sub-table identifiers, and quickly locates the target sub-table set through the index during querying, scanning only the relevant sub-tables; it realizes the separate storage of dynamic data and static metadata, avoiding metadata redundancy; it supports high-concurrency writing and solves the problem of single-table write lock contention; it achieves millisecond-level query response through the in-memory inverted index, which can meet the real-time retrieval and analysis needs of massive historical time-series data in emergency command scenarios, and provides efficient data support for risk assessment models. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art 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.

[0028] Figure 1 This is a flowchart illustrating the overall process framework of the intelligent sensing and risk warning method for firefighters entering and exiting a fire scene according to the present invention.

[0029] Figure 2 This is a schematic diagram of the process for automatically determining the entry and exit status of a fire scene in an embodiment of the present invention.

[0030] Figure 3 This is a schematic diagram of the firefighter safety risk assessment method in an embodiment of the present invention.

[0031] Figure 4 This is a flowchart illustrating the multi-link communication fallback transmission algorithm in an embodiment of the present invention.

[0032] Figure 5 This is a schematic diagram of the data writing process in an embodiment of the present invention. Detailed Implementation

[0033] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0034] Reference Figure 1 The first aspect of this invention provides a method for intelligent sensing and risk warning of firefighters entering and exiting a fire scene, comprising the following steps: Step S1, System Deployment and Perception Layer Configuration By deploying RFID sensing nodes in different areas of the fire scene, the identification, vital signs data, and real-time location signals of firefighters wearing wearable devices are collected; the specific methods are as follows: Deploy sensing layer equipment in the fire operation area. The sensing layer includes wearable devices for firefighters and fire sensing nodes.

[0035] The firefighter wearable device integrates the following units: a low-power MCU main control unit, responsible for data fusion and logic control; passive or active RFID tags, storing unique identification identifiers, operating at a frequency of 13.56MHz or 900MHz; a vital signs sensing unit, including a photoelectric heart rate sensor (sampling rate set to 1Hz), a body surface temperature sensor, and a nine-axis motion sensor; a multi-mode communication unit, simultaneously supporting cellular networks (4G / 5G) and short-range self-organizing networks (Mesh / LoRa); and non-volatile flash memory with a cache capacity of not less than 128MB for resuming interrupted data transmission.

[0036] Fire scene sensing nodes are deployed at fire scene entrances and exits, safe zone boundaries, buffer zone boundaries, and hazardous zone boundaries. Each sensing node integrates an RFID reader and a data aggregation gateway. The RFID reader's reading distance is adjustable from 3 to 10 meters, with a concurrent reading capacity of no less than 50 tags per second. The equipment has an IP67 protection rating, making it suitable for harsh fire scene environments.

[0037] Step S2: Automatic determination of entry and exit status from the fire scene based on RFID sequence.

[0038] Based on the mapping relationship between the timestamp sequence read by RFID sensing nodes and the node location identifier, the system automatically determines the entry and exit status of firefighters from the fire scene; for example... Figure 2 As shown, the specific method is as follows: During the system initialization phase, the node topology is preset, and a mapping table between sensing node identifiers and area types is established. Area types are divided into three categories: safe zone, buffer zone, and danger zone; among them, the buffer zone refers to the transition area set at the entrance or exit of the fire scene, which also belongs to the broad sense of safe zone.

[0039] After firefighters enter the fire scene, their wearable devices periodically broadcast RFID signals. After the sensing nodes deployed in various areas capture the signals, they upload a record to the server. This record includes the firefighter's identification, the sensing node's identification, and a timestamp. After receiving the record, the server retrieves the current record and the previous valid record, and updates the firefighter's entry and exit status according to the following logic: If the area to which the sensing node in the current record belongs is a danger zone, and the area to which the sensing node in the previous valid record belongs is a safe zone or buffer zone, then the status is determined as "entering the fire scene". If the area to which the sensing node in the current record belongs is a safe zone or buffer zone, and the area to which the sensing node in the previous valid record belongs is a danger zone, then the status is determined as "leaving the fire scene". If the area in the current record has not changed from that in the previous valid record, the status is determined to be "In Operation".

[0040] In a real fire scene, signal obstruction or momentary equipment outages may cause jumps in zone records, such as a direct transition from a safe zone to a danger zone, with missing records in the buffer zone. In this case, the system performs the following exception handling: Calculate the time difference between the current record and the previous valid record. ; If time difference Greater than or equal to the preset minimum travel time threshold (For example, a value of 5 seconds can be determined based on the actual walking time at the fire scene entrance), then it is determined to be a legitimate movement and data completion is performed. A virtual buffer record is inserted into the time series database to complete the complete entry and exit trajectory. The timestamp recorded in this virtual buffer is... Calculate using the following formula: ; in, The timestamp of the current record. The timestamp of the previous valid record. The preset buffer time ratio, ; If time difference Less than the minimum travel time threshold If the current record is misread, it will be discarded.

[0041] Step S3: Firefighter safety risk assessment based on historical time-series data

[0042] By correlating firefighters' entry and exit status from fire scenes with real-time vital sign data and historical time-series data, a fatigue-risk index model is constructed to calculate the current real-time risk index of firefighters and to issue graded warnings based on index thresholds; for example... Figure 3 As shown, the specific method is as follows: The collected heart rate data is filtered and denoised to remove motion artifacts and high-frequency noise. The average heart rate per minute is calculated, and the current continuous fire operation duration is obtained. ; Extract the following key feature values: The difference between current heart rate and resting heart rate (resting heart rate is the average of the average heart rate under normal resting conditions). ; The slope of the linear regression of heart rate within the most recent preset time window (e.g., the last 5 minutes). This reflects the accelerated rate of physical exertion; The upper limit of safe heart rate is dynamically configured according to age group, and can also be personalized by combining personal health records; Personalized warning thresholds This refers to the maximum permissible safe heart rate difference. Calculate the risk index : ; in, This is the standard duration for continuous fire scene operations; This is a growth coefficient used to control the steepness of the risk increase; The threshold for the linear regression slope of the heart rate within the most recent preset time window; , , These are the weighting coefficients; Based on the calculated risk index, a tiered early warning system is implemented. If the risk index is less than the first threshold (0.6 in this example), it is determined to be in a normal state; If the risk index is between the first and second thresholds, it is determined to be a state that requires attention. The system sends a attention prompt to the command terminal, suggesting that the commander pay attention to the situation. If the risk index is greater than or equal to the second threshold (0.8 in this embodiment), it is determined to be a dangerous situation, an evacuation alarm is sent and the wearable device is driven to vibrate to remind firefighters to evacuate quickly; In addition, a strict rule is set: if the average heart rate per minute is greater than the upper limit of the safe heart rate, it is directly judged as a dangerous state, no longer relying on the risk index.

[0043] To ensure the reliability of data transmission in a fire scene environment, this invention employs a multi-link communication fallback transmission algorithm. For example... Figure 4 As shown, the specific implementation steps are as follows: Periodically test the normalized metrics of the current link (e.g., cellular network) and backup links (e.g., Mesh / LoRa ad hoc networks): signal strength. Packet loss rate and delay The values ​​for each indicator range from 0 to 1. Calculate the link quality score for each communication link separately. The calculation formula is: ; in, , , These are the weighting coefficients; Preset switching threshold and recovery threshold, with the switching threshold being greater than the recovery threshold; If the link quality score of the current link is less than the recovery threshold and the link quality score of the backup link is greater than the handover threshold, then switch to the backup link; otherwise, maintain the current link.

[0044] Furthermore, when the quality scores of all links fall below the recovery threshold, the system activates local caching mode: Wearable devices write the time-series data to be reported (including heart rate, temperature, location, etc.) into local non-volatile flash memory, and each data point is stored according to the original collection timestamp.

[0045] The system continuously monitors the link status. Once any link is detected to have recovered (quality score higher than recovery threshold), it immediately reads the data from the local cache queue and retransmits it to the server in the order of timestamps. After successful retransmission, the uploaded cached data is cleared and storage space is released.

[0046] To achieve efficient storage and retrieval of massive amounts of high-frequency time-series data, this invention employs a storage method based on the TDengine database; the specific implementation is as follows: Create a super table containing two types of columns: Dynamic time-series data columns (ordinary columns) include: timestamps, heart rate, body surface temperature, exercise status, battery level and other indicators that change frequently over time. These are stored as ordinary columns and support efficient time series compression. The static metadata tag column (TAG column) includes infrequently changing attribute data such as firefighter identification, affiliated squadron, sensing node location, equipment ID, and indicator type. It is stored as a tag and only one copy is stored globally to avoid redundancy. like Figure 5 As shown, the data writing process is executed according to the following sub-steps: Wearable devices report data, and the data access gateway parses the reported data, extracting the tag field and time series field. The routing decision is made based on the extracted tag fields to determine whether there is an independent sub-table corresponding to the current firefighter's identifier and indicator type; If no corresponding independent sub-table exists, the independent sub-table will be created automatically. The sub-table inherits all column definitions of the super table and stores the value of the tag column as metadata. It will not be written repeatedly to each time series record. The time series data will be synchronously stored in the independent sub-table. At the same time, an in-memory inverted index will be built based on the tag field to establish a mapping relationship between tag values ​​and sub-table identifiers. If a corresponding independent sub-table already exists, the parsed time-series data is directly written into that independent sub-table, and the in-memory inverted index is updated after the writing is completed. When executing a query, the target sub-table set is quickly located using an in-memory inverted index, and only the sub-tables related to the query conditions are scanned, avoiding a full table scan and achieving millisecond-level response.

[0047] A second aspect of the present invention provides an intelligent sensing and risk warning system for firefighters entering and exiting a fire scene, comprising: The perception layer includes wearable devices for firefighters and fire scene perception nodes. The wearable devices for firefighters integrate RFID tags, vital sign sensors, multi-mode communication modules, and local cache storage. The fire scene perception nodes include RFID readers and data aggregation gateways deployed at the boundaries of safe zones, buffer zones, and danger zones. Network layer: includes mobile communication network links (4G / 5G) and self-organizing network (Mesh / LoRa) links, supporting dynamic handover and breakpoint resumption based on link quality scores; Application layer: includes data access gateway, time series database, data analysis engine and command visualization terminal; the data analysis engine is used to perform entry and exit status determination, multi-link communication switching decision and safety risk assessment, and push the results to command visualization terminal.

[0048] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method.

[0049] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent sensing and risk warning of firefighters entering and exiting a fire scene, characterized in that, Includes the following steps: By deploying RFID sensing nodes in different areas of the fire scene, the identification, vital signs data and real-time location signals of firefighters wearing wearable devices are collected. Based on the mapping relationship between the timestamp sequence read by RFID sensing nodes and the node location identifier, the system automatically determines the status of firefighters entering and leaving the fire scene. By linking firefighters' entry and exit status from the fire scene with real-time vital signs data and historical time series data, a fatigue-risk index model is constructed to calculate the current real-time risk index of firefighters and to issue graded warnings based on index thresholds.

2. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 1, characterized in that, The method for automatically determining the status of firefighters entering and leaving a fire scene includes the following steps: The system node topology is preset, and a mapping table between sensing node identifiers and region types is established, wherein the region types include safe zones, buffer zones, and danger zones; Wearable devices periodically broadcast RFID signals. After each sensing node captures the signal, it uploads a record to the server. The record includes the firefighter's identification, the sensing node's identification, and a timestamp. The server retrieves the current record and the previous valid record, and updates the firefighter's entry and exit status according to the following logic: If the area to which the sensing node in the current record belongs is a danger zone, and the area to which the sensing node in the previous valid record belongs is a safe zone or buffer zone, then the status is determined to be entering the fire scene. If the area to which the sensing node in the current record belongs is a safe zone or buffer zone, and the area to which the sensing node in the previous valid record belongs is a danger zone, then the status is determined to be leaving the fire scene. If the area in the current record has not changed from that in the previous valid record, the status is determined to be "in operation".

3. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 2, characterized in that, When a region jump occurs—that is, a direct jump from a safe zone to a danger zone or from a danger zone to a safe zone without a buffer record—the time difference between the current record and the previous valid record is calculated. ; If time difference Greater than or equal to the preset minimum travel time threshold If the move is deemed legitimate, data completion is performed, and a virtual buffer record is inserted, containing the timestamp of the virtual buffer record. Calculate using the following formula: ； in, The timestamp of the previous valid record. The preset buffer time ratio, ; If time difference Less than the minimum travel time threshold If the current record is misread, it will be discarded.

4. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 1, characterized in that, The method for calculating the real-time risk index of firefighters and issuing graded warnings includes the following steps: The collected heart rate data is filtered and denoised to calculate the average heart rate per minute and obtain the current continuous fire operation duration. ; Extract the difference between the current heart rate and the resting heart rate. The slope of the linear regression of the heart rate within the most recent preset time window Safe heart rate limits and personalized warning thresholds configured according to age groups or individual health records. Calculate the risk index : ； in, This is the standard duration for continuous fire scene operations; This is a growth coefficient used to control the steepness of the risk increase; The threshold for the linear regression slope of the heart rate within the most recent preset time window; , , These are the weighting coefficients; The risk index is used to issue tiered warnings: if the risk index is less than the first threshold, it is considered a normal state; if the risk index is between the first and second thresholds, it is considered a state requiring attention; if the risk index is greater than the second threshold, it is considered a dangerous state, an evacuation alarm is sent, and the wearable device is vibrated; if the average heart rate is greater than the upper limit of the safe heart rate, it is directly considered a dangerous state.

5. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 1, characterized in that, The method also includes a multi-link communication step: Periodically check the signal strength of the current link and the backup link. Packet loss rate and delay ; Calculate the link quality score for each communication link separately. The calculation formula is: ； in, , , These are the weighting coefficients; Preset switching threshold and recovery threshold, with the switching threshold being greater than the recovery threshold; If the link quality score of the current link is less than the recovery threshold and the link quality score of the backup link is greater than the handover threshold, then switch to the backup link; otherwise, maintain the current link.

6. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 5, characterized in that, If the quality score of all links is below the recovery threshold, then local caching mode is enabled; When any communication link is detected to be restored, the data in the local cache queue is read and retransmitted to the server in the original timestamp order. After successful retransmission, the corresponding cache is cleared.

7. The intelligent sensing and risk warning method for firefighters entering and exiting a fire scene as described in claim 1, characterized in that, The method further includes a time-series data storage step: Create a super table, which includes dynamic time-series data columns and static metadata tag columns. The dynamic time-series data columns include at least several of the following: timestamp, heart rate, body surface temperature, exercise status, and battery level. The static metadata tag columns include at least several of the following: firefighter identification, affiliated squadron, sensing node location, device ID, and indicator type. When writing data, the following sub-steps are executed: Wearable devices report data, and the data access gateway parses the reported data, extracting the tag field and time series field. The routing decision is made based on the extracted tag fields to determine whether there is an independent sub-table corresponding to the current firefighter's identifier and indicator type; If no corresponding independent sub-table exists, the independent sub-table will be created automatically. The independent sub-table inherits all column definitions of the super table and stores the value of the tag column as metadata, synchronously storing time-series data. At the same time, an in-memory inverted index is built based on the tag field to establish a mapping relationship between tag values ​​and sub-table identifiers. If a corresponding independent sub-table already exists, the parsed time-series data is directly written into that independent sub-table, and the in-memory inverted index is updated after the writing is completed. When performing a query, the target sub-table set is quickly located using an in-memory inverted index, and only sub-tables related to the query conditions are scanned.

8. A smart sensing and risk warning system for firefighters entering and exiting a fire scene, used to execute the method according to any one of claims 1-7, characterized in that, The system includes: The perception layer includes wearable devices for firefighters and fire scene perception nodes. The wearable devices for firefighters integrate RFID tags, vital sign sensors, multi-mode communication modules, and local cache storage. The fire scene perception nodes include RFID readers and data aggregation gateways deployed at the boundaries of safe zones, buffer zones, and danger zones. Network layer: includes mobile communication network links and ad hoc network links, supporting dynamic handover and breakpoint resumption based on link quality scores; Application layer: includes data access gateway, time series database, data analysis engine and command visualization terminal; the data analysis engine is used to perform entry and exit status determination, multi-link communication switching decision and safety risk assessment, and push the results to command visualization terminal.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images