Fire linkage control system and method

Abstract

The application is suitable for the field of fire fighting technology, and provides a fire fighting linkage control system and method, comprising the following steps: collecting sensing data of carbon monoxide concentration, temperature, flame flicker frequency and smoke particle size, and calculating fire confidence; confirming fire and performing three-dimensional positioning of the fire source when the confidence exceeds a first threshold value, predicting fire spreading trend according to the temperature rising rate and carbon monoxide diffusion radius; establishing a fire source risk field and a smoke diffusion field, correcting the risk value according to the real-time personnel density multiplied by the crowding coefficient, and generating a comprehensive risk space distribution map; determining the fire extinguishing coverage range based on the fire source position and the predicted fire area, and prestarting the fire extinguishing valve, taking the spatial integral of the comprehensive risk as a cost function to calculate the minimum risk path and dynamically change the evacuation indication direction. The application fuses the fire source risk field, the smoke diffusion field and the personnel density correction, so that the fire extinguishing coverage range selection and the evacuation path planning are both based on the current real risk, and the effect is better.

Description

Technical Field

[0001] This invention relates to the field of fire protection technology, specifically to a fire linkage control system and method. Background Technology

[0002] Existing fire alarm control systems generally employ fixed threshold alarms and preset linkage logic, which have the following technical drawbacks: First, detectors are easily affected by environmental interference such as kitchen fumes and construction dust, leading to a high false alarm rate; simultaneously, smoldering fires exhibit weak characteristics, resulting in delayed identification and missed opportunities for optimal firefighting. Second, risk assessment relies on a single dimension, depending solely on the fire alarm signal without incorporating information on smoke diffusion and personnel density, resulting in fixed evacuation directions that may lead people to the fire. Furthermore, the accuracy of fire source location is low, making it impossible to ascertain the fire's spread trend, and the activation range of firefighting equipment depends on manual experience. Therefore, a fire alarm control system and method are needed to address these problems. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a fire-fighting linkage control system and method to solve the problems existing in the above-mentioned background technology.

[0004] This invention is implemented as follows: a fire alarm linkage control method, the method comprising the following steps: Sensor data on carbon monoxide concentration, temperature, flame flashing frequency, and smoke particle size are collected. Environmental noise is filtered out by sliding window mid-range filtering. Fire confidence is calculated by weighting carbon monoxide concentration, temperature rise, and flame flashing frequency. When the confidence level exceeds the first threshold, the fire is confirmed and the fire source is located in three dimensions. The fire spread trend is predicted based on the temperature rise rate and carbon monoxide diffusion radius. Establish fire source risk field and smoke diffusion field, correct risk value based on real-time personnel density multiplied by crowding coefficient, and generate comprehensive risk spatial distribution map; The fire extinguishing coverage area is determined based on the location of the fire source and the predicted burned area, and the fire extinguishing valves are pre-activated. The minimum risk path is calculated using the spatial integral of the comprehensive risk as the cost function, and the evacuation direction is dynamically changed. The decision instructions are encoded into bus protocol frames and broadcast to the field equipment.

[0005] As a further aspect of the present invention, the steps of three-dimensionally locating the fire source and predicting the fire spread trend specifically include: Centroid-weighted localization is performed using the fire confidence scores of multiple adjacent sensing nodes: the node coordinates are weighted and averaged to obtain the three-dimensional coordinates of the fire source using the confidence scores of each node as weights. From the moment the fire is confirmed, the rate of temperature rise and the radius of carbon monoxide diffusion in the fire source area are recorded at preset time intervals. The first-order linear extrapolation method is used to estimate the burned area after a set period of time based on the temperature rise rate of the three most recent records, and to estimate the flue gas layer height descent rate after a set period of time based on the carbon monoxide diffusion radius of the three most recent records.

[0006] As a further aspect of the present invention, the step of establishing the fire source risk field and the smoke diffusion field specifically includes: The fire risk value at any point in space is defined as the fire source intensity divided by the square of the distance from that point to the fire source, multiplied by the negative attenuation coefficient of the natural constant e, and the product of the distance from that point to the nearest obstacle raised to the power of the product. The fire source intensity is proportional to the product of the current burned area and the rate of temperature rise. Based on the carbon monoxide sensor network within the building and smoke detectors arranged vertically, a three-layer Gaussian plume model is constructed. The smoke risk value is defined as the first coefficient multiplied by the ratio of the measured carbon monoxide concentration to the lethal concentration, plus the second coefficient multiplied by the ratio of the visibility benchmark value to the measured visibility.

[0007] As a further aspect of the present invention, the step of generating a comprehensive risk spatial distribution map specifically includes: The personnel density in each fire compartment is estimated in real time using Wi-Fi probes, infrared counting devices, and gate data. When the population density exceeds the preset density, the risk value of the area is corrected by multiplying the superposition value of the fire source risk field and the smoke diffusion field in the area by the crowding coefficient. The corrected risk values ​​of all regions are overlaid and rendered on a 1m×1m planar grid to generate a comprehensive risk spatial distribution map. The distribution map uses color or grayscale levels to represent the level of risk.

[0008] As a further aspect of the present invention, the step of determining the fire extinguishing coverage area and pre-activating the fire extinguishing valves based on the fire source location and predicted burned area specifically includes: Calculate the required fire suppression coverage radius, which is equal to the square root of the quotient obtained by multiplying the predicted fire area by pi. In an automatic sprinkler system, select a set of sprinklers whose center of coverage is closest to the fire source and whose coverage radius requirement is the highest, and whose coverage efficiency is the highest. Pre-start their electric valves to the ready-to-spray state, and maintain the pressure of the water supply pipeline but do not spray temporarily. When the rate of temperature rise exceeds 3 degrees Celsius per second, immediately instruct the pre-activated fire extinguishing equipment to spray at full flow.

[0009] As a further aspect of the present invention, the step of calculating the minimum risk path using the spatial integral of the comprehensive risk as the cost function specifically includes: Based on the current comprehensive risk spatial distribution map, the path cost from any location to any evacuation exit is defined as the integral of the comprehensive risk value of each grid point on that path; The heat-inspired Dijkstra algorithm is used to calculate the minimum cost path to all available evacuation exits starting from the center of each fire compartment, and the exit with the minimum cost and its corresponding path are selected as the recommended evacuation route.

[0010] Another object of the present invention is to provide a fire alarm linkage control system, the system comprising: The fire confidence module is used to collect sensor data on carbon monoxide concentration, temperature, flame flicker frequency, and smoke particle size. It filters out environmental noise through a sliding window mid-range filter and calculates the fire confidence based on a weighted average of carbon monoxide concentration, temperature rise, and flame flicker frequency. The fire source three-dimensional location module is used to confirm the fire and perform three-dimensional location of the fire source when the confidence level exceeds the first threshold, and predict the fire spread trend based on the temperature rise rate and carbon monoxide diffusion radius. The risk spatial distribution module is used to establish the fire source risk field and smoke diffusion field, correct the risk value based on the real-time personnel density multiplied by the crowding coefficient, and generate a comprehensive risk spatial distribution map; The fire-fighting linkage control module is used to determine the fire extinguishing coverage area based on the fire source location and the predicted fire area, and to pre-activate the fire extinguishing valves. It calculates the minimum risk path using the spatial integral of comprehensive risk as the cost function and dynamically changes the evacuation indication direction. The decision instructions are encoded into bus protocol frames and broadcast to the field equipment.

[0011] Compared with the prior art, the beneficial effects of the present invention are: Noise is filtered out using a sliding window mid-range filter, and confidence levels are calculated by weighting carbon monoxide, temperature, and flame flicker frequency. This effectively distinguishes real fires from interference such as kitchen fumes, reducing false alarms. It also allows for early identification of smoldering fires, shortening alarm delays. Multi-node confidence-weighted centroid positioning provides high accuracy. Extrapolating the burned area and smoke descent rate by combining temperature rise rate and carbon monoxide diffusion radius provides a forward-looking basis for firefighting and evacuation. Integrating the fire source risk field, smoke diffusion field, and personnel density correction generates a real-time comprehensive risk distribution map, ensuring that firefighting coverage selection and evacuation route planning are based on current real-time risk, rather than fixed logic. The minimum risk path is calculated using the risk integral as a cost function, dynamically changing evacuation direction indications. Simultaneously, pre-activating fire extinguishing valves and spraying full flow as the fire intensifies further enhances effectiveness. Attached Figure Description

[0012] Figure 1 This is a flowchart of a fire alarm linkage control method.

[0013] Figure 2 This is a flowchart for three-dimensional fire source location in a fire-fighting linkage control method.

[0014] Figure 3This is a flowchart for establishing a fire source risk field and a smoke diffusion field in a fire-fighting linkage control method.

[0015] Figure 4 This is a flowchart for generating a comprehensive risk spatial distribution map in a fire-fighting linkage control method.

[0016] Figure 5 This is a flowchart for determining the fire extinguishing coverage area and pre-activating fire extinguishing valves in a fire-fighting linkage control method.

[0017] Figure 6 This is a flowchart for calculating the minimum risk path in a fire-fighting linkage control method.

[0018] Figure 7 This is a schematic diagram of a fire alarm linkage control system. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0021] like Figure 1 As shown in the figure, an embodiment of the present invention provides a fire alarm linkage control method, the method comprising the following steps: S100 collects sensor data on carbon monoxide concentration, temperature, flame flashing frequency, and smoke particle size. It filters out environmental noise through a sliding window mid-range filter and calculates the fire confidence level based on a weighted average of carbon monoxide concentration, temperature rise, and flame flashing frequency.

[0022] In this embodiment of the invention, sensing nodes deployed within the building collect data on carbon monoxide concentration, temperature, flame flicker frequency, and smoke particle size. A sliding window mid-range filter (window width 1 second) is used to remove instantaneous spike noise and eliminate environmental interference such as kitchen fumes. Then, according to the weighted formula: Confidence = w1 × (CO concentration / CO threshold) + w2 × (temperature rise / temperature rise threshold) + w3 × (flame frequency / frequency threshold), where the weights w1, w2, and w3 can be adaptively adjusted according to the scenario (factory, garage, etc.).

[0023] S200 confirms the fire and performs three-dimensional location of the fire source when the confidence level exceeds the first threshold, and predicts the fire spread trend based on the temperature rise rate and carbon monoxide diffusion radius.

[0024] In this embodiment of the invention, a fire is confirmed when the confidence level exceeds a first threshold (e.g., 0.85). Centroid-weighted localization is performed using the confidence values ​​of multiple adjacent nodes to obtain the three-dimensional coordinates of the fire source. Simultaneously, the temperature rise rate and carbon monoxide diffusion radius are recorded every 2 seconds, and a first-order linear extrapolation method is used to predict the burned area and the rate of decrease in smoke layer height after 30 seconds.

[0025] S300 establishes a fire source risk field and a smoke diffusion field, corrects the risk value by multiplying the real-time personnel density by the congestion coefficient, and generates a comprehensive risk spatial distribution map.

[0026] In this embodiment of the invention, a fire source risk field and a smoke diffusion field are established. The risk value is directly proportional to the fire source intensity and inversely proportional to the square of the distance. Personnel density is then obtained using methods such as Wi-Fi probes. When the density exceeds 4 people / m², it is multiplied by a crowding coefficient for correction. The corrected risk values ​​are then overlaid and rendered on a 1m×1m grid to generate a color-coded risk distribution map.

[0027] The S400 determines the fire extinguishing coverage area based on the fire source location and predicted fire area, and pre-activates the fire extinguishing valves. It calculates the minimum risk path using the spatial integral of comprehensive risk as the cost function and dynamically changes the evacuation indication direction. The decision instructions are encoded into bus protocol frames and broadcast to field devices.

[0028] In this embodiment of the invention, the required coverage radius is calculated based on the fire source location and the predicted burned area, and the optimal sprinkler head is pre-activated. Simultaneously, using the risk space integral as the cost function, an improved Dijkstra algorithm is employed to calculate the risk integral value for each path, selecting the path with the lowest risk and dynamically changing the evacuation indication direction. Adjustable emergency lights are used here, each with a programmable arrow direction. Every 2 seconds, the optimal forward direction for each node is recalculated based on the latest risk distribution map, and the arrow direction is refreshed. Finally, the decision command is encoded into a Modbus / CAN bus protocol frame, broadcast to field devices, and awaits confirmation.

[0029] like Figure 2 As shown, in a preferred embodiment of the present invention, the steps of performing three-dimensional location of the fire source and predicting the fire spread trend specifically include: S201, Centroid weighted localization is performed using the fire confidence scores of multiple adjacent sensing nodes: The node coordinates are weighted and averaged using the confidence scores of each node as weights to obtain the three-dimensional coordinates of the fire source. S202, starting from the moment the fire is confirmed, record the rate of temperature rise and carbon monoxide diffusion radius of the fire source area at preset time intervals. S203 uses a first-order linear extrapolation method to estimate the burned area after a set time period based on the temperature rise rate of the three most recent records, and to estimate the flue gas layer height descent rate after a set time period based on the carbon monoxide diffusion radius of the three most recent records.

[0030] Specifically, a first-order linear extrapolation method is used to estimate the burned area after a set time period based on the temperature rise rate of the three most recent records, and to estimate the flue gas layer height descent rate after a set time period based on the carbon monoxide diffusion radius of the three most recent records. The specific steps are as follows: Step a: Arrange the three consecutive recorded temperature rise rates in chronological order to obtain the first rate, the second rate, and the third rate; calculate the increment of the second rate relative to the first rate to obtain the first rate increment; calculate the increment of the third rate relative to the second rate to obtain the second rate increment. Step b: When both the first rate increment and the second rate increment are greater than zero, the fire is determined to be in the accelerated spread stage, and the ratio of the second rate increment to the first rate increment is used as the acceleration factor; the third rate is multiplied by the acceleration factor to obtain the predicted temperature rise rate. Step c: When the second rate increment is less than or equal to the first rate increment and greater than zero, it is determined that the fire is in the uniformly accelerated spread stage, and the third rate is added to half of the second rate increment to obtain the predicted temperature rise rate. Step d: When the second rate increment is less than or equal to zero, the fire is determined to be in the deceleration and spread stage. The third rate is subtracted from half of the absolute value of the first rate increment to obtain the predicted temperature rise rate. Step e: Divide the predicted temperature rise rate by the temperature rise rate in the most recent historical record to obtain the rate growth ratio; obtain the calorific value of combustible material per unit area, and multiply the calorific value of combustible material per unit area by the rate growth ratio to obtain the fire area expansion coefficient; then multiply the current fire area by the fire area expansion coefficient to obtain the fire area after a set time period in the future. Step f: Arrange the carbon monoxide diffusion radii recorded three times in chronological order to obtain the first radius, second radius, and third radius; calculate the increment of the second radius relative to the first radius, and simultaneously calculate the increment of the third radius relative to the second radius to obtain the first radius increment and the second radius increment; use the first radius increment as the first rate increment and the second radius increment as the second rate increment, and repeat steps b to d to determine the fire spread stage and calculate the predicted rate increment to obtain the predicted change trend factor of the carbon monoxide diffusion radius; multiply the current carbon monoxide diffusion radius by the predicted change trend factor to obtain the carbon monoxide diffusion radius after a set future time period. Step g: Determine the target building floor height based on the three-dimensional coordinates of the fire source; divide the predicted carbon monoxide diffusion radius by the target building floor height to obtain the smoke settling ratio; subtract the current smoke layer height from the product of the smoke settling ratio and a preset height coefficient, and take the absolute value to obtain the smoke layer height descent rate after a set time period in the future.

[0031] Furthermore, this invention makes fire extinguishing and evacuation decisions more forward-looking and adaptable by segmenting the fire's acceleration, uniform acceleration, or deceleration trends and dynamically estimating the burned area and smoke settling velocity.

[0032] In this embodiment of the invention, centroid-weighted positioning is first performed using the fire confidence scores of multiple adjacent sensing nodes. Let the coordinates of the i-th sensing node be (xi, yi, zi), and its fire confidence score be Ci. The fire source coordinates (xf, yf, zf) are calculated using the following formulas: xf = ∑(Ci × xi) / ∑Ci, yf = ∑(Ci × yi) / ∑Ci, zf = ∑(Ci × zi) / ∑Ci, where the summation iterates through all nodes whose confidence scores exceed the warning threshold (e.g., 0.6). This method utilizes the high confidence scores of nodes near the fire source and suppresses single-node errors through weighted averaging, resulting in high positioning accuracy. The temperature rise rate and carbon monoxide diffusion radius are recorded every preset time interval. Data is collected every 2 seconds starting from the fire confirmation time t0. The temperature rise rate is calculated by dividing the difference between the two most recent temperatures by the time interval. The carbon monoxide diffusion radius RCO is defined as the Euclidean distance from the fire source center to the location where the CO concentration drops to 50% of the peak concentration, obtained through actual measurements using CO sensors deployed at grid points. Finally, a first-order linear extrapolation method is used to estimate the future fire area and the rate of descent of the smoke layer height: Let the times of the three most recent records be ta, tb, and tc, and the corresponding temperature rise rates be Va, Vb, and Vc, respectively. The first-order linear extrapolation assumes a linear rate change, and the rate 30 seconds from now = Vc + 30 × (Vc - Vb) / Δt, where Δt = 2 seconds. The fire area is approximately proportional to the rate of temperature rise, and the proportionality coefficient is determined by the building's combustible load. Similarly, the rate of descent of the smoke layer height is estimated using the rate of change of the carbon monoxide diffusion radius. This method has low computational complexity and is suitable for real-time execution by edge gateways.

[0033] like Figure 3 As shown in the preferred embodiment of the present invention, the step of establishing the fire source risk field and the smoke diffusion field specifically includes: S301, the fire source risk value at any point in space is defined as the fire source intensity divided by the square of the distance from that point to the fire source, multiplied by the negative attenuation coefficient of the natural constant e multiplied by the product of the distance from that point to the nearest obstacle, where the fire source intensity is proportional to the product of the current burned area and the rate of temperature rise. S302, based on the carbon monoxide sensor network in the building and the smoke detectors arranged in the vertical direction, constructs a three-layer Gaussian plume model, defining the smoke risk value as the first coefficient multiplied by the ratio of the measured carbon monoxide concentration to the lethal concentration plus the second coefficient multiplied by the ratio of the visibility benchmark value to the measured visibility.

[0034] In this embodiment of the invention, a fire source risk field is first established: for any point r=(x,y,z) in space, the fire source risk value is defined as: Rfire(r)= Where: I0 is the ignition source intensity, which is proportional to the product of the current burned area A and the rate of temperature rise. Let r be the Euclidean distance from point r to the center of the fire source, κ be the attenuation coefficient ranging from 0.15 to 0.3 (a larger value indicates a stronger blocking effect of the wall on the risk), and d be the shortest distance from point r to the nearest wall or fireproof partition. This formula reflects the physical characteristic that the fire source risk decreases with the square of the distance and decreases exponentially when passing through a wall. Then, a smoke diffusion field is established: a three-layer Gaussian smoke plume model is constructed based on the carbon monoxide sensor network within the floor and smoke detectors arranged vertically. The smoke risk value is defined as: Rsmoke(r) = α × [CO] / [CO]lethal + β × Vre / Vvis, where: [CO] is the measured carbon monoxide concentration (ppm), [CO]lethal is taken as 800 ppm (short-term exposure hazard threshold), Vvis is the measured visibility (meters), and Vre is the visibility baseline value (usually taken as 30 meters, i.e., normal visibility). α and β are weighting coefficients, satisfying α + β = 1. In key passageways such as corridors and stairwells, the sampling density is increased by densely deploying sensors. The entire smoke diffusion field is updated every 5 seconds.

[0035] like Figure 4 As shown, in a preferred embodiment of the present invention, the step of generating a comprehensive risk spatial distribution map specifically includes: S303 uses Wi-Fi probes, infrared counting devices, and turnstile data to estimate the personnel density in each fire compartment in real time. S304, When the population density exceeds the preset density, the risk value of the area is corrected: the superposition value of the fire source risk field and the smoke diffusion field in the area is multiplied by the crowding coefficient. S305 overlays and renders all the corrected regional risk values ​​on a 1m×1m planar grid to generate a comprehensive risk spatial distribution map. The distribution map uses color or grayscale levels to represent the level of risk.

[0036] Specifically, the personnel density in each fire compartment is estimated in real time using Wi-Fi probes, infrared counting devices, and turnstile data. The specific steps are as follows: The number of mobile device signals obtained by Wi-Fi probes in each fire compartment, the number of people passing through the infrared counting device, and the number of people swiping cards recorded in the gate data are timestamped to obtain the original personnel count data from the three sources within the time window. Sliding window mid-range filtering was applied to the raw personnel count data from the three sources within the time window to obtain smoothed counts from Wi-Fi probes, infrared probes, and turnstiles. The difference between the smoothed count from infrared probes and the smoothed count from turnstiles was used as an estimated number of personnel without access cards. The first fused personnel base is obtained by adding the smoothed count of Wi-Fi probes to the estimated number of personnel without access cards; the second fused personnel base is obtained by weighting the first fused personnel base with the smoothed infrared count; wherein, the weight of the smoothed Wi-Fi probe count is 0.6 and the weight of the smoothed infrared count is 0.4. Obtain the known area and the area occupied by fixed obstacles in each fire compartment that are pre-stored; subtract the known area from the area occupied by fixed obstacles to obtain the effective activity area; divide the second fused personnel base by the effective activity area to obtain the personnel density in the fire compartment.

[0037] Furthermore, by integrating Wi-Fi probe, infrared counting, and gate data, and eliminating fixed obstacle areas, this invention enables more accurate and real-time estimation of personnel density, providing a reliable basis for risk correction and evacuation route planning.

[0038] In this embodiment of the invention, when estimating the personnel density of each fire compartment in real time, the personnel density ρ is calculated by combining the data from the Wi-Fi probes (counting MAC addresses), infrared counting devices (counting the flow of people entering and exiting), and turnstiles originally deployed in the space with the known area of ​​each fire compartment. The calculation is: ρ = current number of people in the area / area of ​​the area. When the personnel density exceeds a preset density, a risk correction is performed. The preset density threshold is 4 people / m² (crowding significantly increases the risk of stampedes). The crowding coefficient is calculated as follows: when ρ > 4, crowding coefficient = 1 + 0.2 × (ρ - 4); when ρ < 4, crowding coefficient = 1. The corrected area risk value Rtotal(r) = (Rfire(r) + Rsmoke(r)) × crowding coefficient. This correction amplifies the risk of crowded areas, thereby guiding evacuation routes to automatically avoid high-density areas. Next, each fire compartment is divided into 1-meter square grid units, and each unit is assigned the corrected risk value. Color mapping is used: green for risks of 0-0.3 (low risk), yellow for risks of 0.3-0.7 (medium risk), and red for risks >0.7 (high risk), with linear interpolation used for smooth transitions in between. The distribution map is pushed in real time to the fire control room's large screen and mobile command terminal via the local area network.

[0039] like Figure 5 As shown, in a preferred embodiment of the present invention, the steps of determining the fire extinguishing coverage area and pre-activating the fire extinguishing valves based on the fire source location and predicted burned area specifically include: S401, calculate the required fire extinguishing coverage radius. The required coverage radius is equal to the square root of the quotient obtained by multiplying the predicted fire area by pi. S402 In an automatic sprinkler system, select a set of sprinklers whose center of coverage is closest to the fire source and whose coverage radius requirement is the highest and whose coverage efficiency is the highest. Pre-start their electric valves to the standby state, maintain pressure in the water supply pipeline but do not spray temporarily. S403: When the rate of temperature rise exceeds 3 degrees Celsius per second, immediately instruct the pre-activated fire extinguishing equipment to spray at full flow.

[0040] Specifically, in an automatic sprinkler system, select a set of sprinklers whose center of coverage is closest to the required coverage radius and whose coverage efficiency is the highest. The specific steps are as follows: Obtain the preset coordinates of all automatic sprinkler heads in the building; based on the preset coordinates of the automatic sprinkler heads, and using the coverage radius requirement as the distance benchmark, calculate the spatial straight-line distance between the three-dimensional coordinates of the coverage center of each sprinkler head and the three-dimensional coordinates of the fire source; select the three sprinkler heads with the smallest absolute value of the difference between the spatial straight-line distance and the distance benchmark as the first candidate sprinkler head group. The comprehensive risk value at the center of the sprinkler coverage is extracted from the comprehensive risk spatial distribution map. At the same time, the average value of the comprehensive risk values ​​of all grid points on the boundary of the circular area with the fire source as the center and the equivalent radius corresponding to the burned area after a set future time period as the radius is extracted. The comprehensive risk value at the center of the sprinkler coverage is divided by the average value of the comprehensive risk values ​​and the reciprocal is taken to obtain the fire location fitness. The three sprinklers in the first candidate sprinkler group are sorted in order of fire location fitness from high to low, and the sprinkler with the highest fire location fitness is selected as the main fire extinguishing sprinkler. The coordinates of all nozzles adjacent to the main fire extinguishing nozzle are obtained from the preset database, and the ratio of the comprehensive risk value at the coverage center of the current adjacent nozzle to the comprehensive risk value at the coverage center of the main fire extinguishing nozzle is calculated to obtain the risk complementarity coefficient. The risk complementarity coefficient is filtered according to the preset filtering rules, and the nozzles corresponding to the filtered risk complementarity coefficients are used as auxiliary fire extinguishing nozzles. The filtering rules are: the risk complementarity coefficient is greater than 0.6 and less than 1.4. The main fire extinguishing nozzle is the core nozzle that meets the requirement of having the closest coverage radius to the fire source and the highest coverage efficiency. The auxiliary fire extinguishing nozzle is a supplementary nozzle that works in conjunction with the core nozzle. The main fire extinguishing nozzle and the auxiliary fire extinguishing nozzle together constitute a set of coordinated fire extinguishing nozzles.

[0041] Furthermore, this invention selects primary and secondary coordinated nozzles through distance screening, risk adaptability assessment, and complementary coefficient matching, so that the fire extinguishing coverage area is highly consistent with the actual risk distribution of the fire site, thereby improving the efficiency of linkage.

[0042] In this embodiment of the invention, when calculating the required fire extinguishing coverage radius, the predicted burned area is denoted as Apred, and the required coverage radius is Rneed, where Rneed = 1.2 × A coefficient of 1.2 indicates selecting a coverage area slightly larger than the actual equivalent radius of the fire source to ensure that the extinguishing medium can completely cover the edge of the fire. In an automatic sprinkler system, the installation coordinates and protection radius (or adjustable swing angle range) of each sprinkler head are known. Calculate the distance between the coverage center of each sprinkler head and the center of the fire source, and select the sprinkler head with the closest distance to Rneed and the highest coverage efficiency (the proportion of the fire source area covered within the protection range). Open the electric valve of the sprinkler head to the ready-to-spray position, maintain the water supply line pressure (e.g., maintain 0.4 MPa water pressure), but keep the solenoid valve closed and do not spray. This way, once the fire is confirmed to be expanding, spraying can be started immediately, saving water filling time. Continuously monitor the temperature rise rate of the fire source area. When the temperature rise rate exceeds 3 degrees Celsius per second, it indicates that the fire has entered a rapid growth stage, and fire extinguishing should be implemented immediately. The system issues an activation command to the pre-activated sprinklers, causing the fire extinguishing equipment to spray at full flow.

[0043] like Figure 6 As shown, in a preferred embodiment of the present invention, the step of calculating the minimum risk path using the spatial integral of the comprehensive risk as the cost function specifically includes: S404, based on the current comprehensive risk spatial distribution map, defines the path cost from any location to any evacuation exit as the integral of the comprehensive risk value of each grid point on that path; S405 employs the heat-inspired Dijkstra algorithm to calculate the minimum cost path to all available evacuation exits starting from the center of each fire compartment, and selects the exit with the minimum cost and its corresponding path as the recommended evacuation route.

[0044] Specifically, the heat-based heuristic Dijkstra algorithm is used to calculate the minimum-cost path to all available evacuation exits starting from the center of each fire compartment. The exit with the minimum cost and its corresponding path are selected as the recommended evacuation route. The specific steps are as follows: Extract the geometric center coordinates of each fire compartment from the current comprehensive risk spatial distribution map to serve as the initial reference point for personnel gathering in each compartment; For each fire compartment, starting from the initial reference point where people gather within the compartment and ending at each available evacuation exit, the comprehensive risk value of each grid point is calculated using the heat-inducing Dijkstra algorithm on a 1m×1m grid divided by the comprehensive risk spatial distribution map. The comprehensive risk values ​​of each grid point are then summed to obtain the total risk cost of the path. Specifically, the heat-inducing Dijkstra algorithm works as follows: the edge weight between two adjacent grid points is set to the arithmetic mean of their respective comprehensive risk values; the priority of nodes is ranked based on the cumulative sum of the comprehensive risk values ​​of all grid points along the straight-line distance from the current node to the current target evacuation exit; the cumulative sum is used as the heat value, and nodes with lower heat values ​​have higher priority during the search process. The total risk cost of paths corresponding to the same fire compartment is compared, and the path with the minimum total risk cost is selected to obtain the target path. The evacuation exit corresponding to the target path is determined as the recommended evacuation exit of the current fire compartment, and the target path is also used as the recommended evacuation route of the current fire compartment.

[0045] Furthermore, this invention uses the accumulated risk values ​​of grid points as path costs and the accumulation of risk in the straight line direction to guide node priority, thereby quickly finding the evacuation route with the lowest risk in each zone and effectively guiding personnel to avoid high-risk areas.

[0046] In this embodiment of the invention, when using path cost as the spatial integral of the comprehensive risk value, let the current comprehensive risk spatial distribution map be R(x,y), where (x,y) are planar grid coordinates. The cost J(Γ) of any path Γ from the starting point A to the ending point B (evacuation exit) is defined as: J(Γ) = , where ds is the path element length. This definition is independent of the geometric path length and depends only on the accumulated risk exposure along the way. Therefore, taking a longer route but passing through a low-risk area may be safer than taking a shorter route but crossing a fire. The minimum-cost path is then calculated using a heat-heuristic Dijkstra algorithm. The standard Dijkstra algorithm uses geometric distance as the cost; here, the edge weights between grid nodes are set to the average risk value of two adjacent grid points multiplied by the grid spacing (1 meter). Heat heuristic: When prioritizing, a heuristic term is introduced: an estimate of the average risk value over the straight-line distance from the current node to the target exit, accelerating convergence. The algorithm starts from the center of each fire compartment (personnel gathering point) and calculates the minimum cumulative cost path to all available evacuation exits (stairwells, outdoor gates, etc.). Finally, the exit with the minimum cost and its corresponding path are selected as the recommended evacuation route for that compartment.

[0047] like Figure 7 As shown in the figure, this embodiment of the invention also provides a fire alarm linkage control system, the system comprising: The fire confidence module 100 is used to collect sensor data on carbon monoxide concentration, temperature, flame flashing frequency and smoke particle size. It filters out environmental noise through sliding window mid-range filtering and calculates the fire confidence based on the weighted average of carbon monoxide concentration, temperature rise and flame flashing frequency. The three-dimensional fire source location module 200 is used to confirm the fire and perform three-dimensional fire source location when the confidence level exceeds the first threshold, and to predict the fire spread trend based on the temperature rise rate and carbon monoxide diffusion radius. The risk spatial distribution module 300 is used to establish the fire source risk field and the smoke diffusion field, correct the risk value based on the real-time personnel density multiplied by the crowding coefficient, and generate a comprehensive risk spatial distribution map. The fire linkage control module 400 is used to determine the fire extinguishing coverage area based on the fire source location and the predicted fire area, and to pre-activate the fire extinguishing valves. It calculates the minimum risk path using the spatial integral of comprehensive risk as the cost function and dynamically changes the evacuation indication direction. The decision instructions are encoded into bus protocol frames and broadcast to the field equipment.

[0048] In a preferred embodiment of the present invention, the fire source three-dimensional positioning module 200 includes: The three-dimensional coordinate determination unit is used to perform centroid-weighted positioning by utilizing the fire confidence scores of multiple adjacent sensing nodes: the node coordinates are weighted and averaged using the confidence scores of each node as weights to obtain the three-dimensional coordinates of the fire source. The information recording unit is used to record the rate of temperature rise and carbon monoxide diffusion radius of the fire source area at preset time intervals, starting from the moment the fire is confirmed. The information prediction unit is used to estimate the burned area after a set period of time based on the temperature rise rate of the three most recent records using a first-order linear extrapolation method, and to estimate the flue gas layer height descent rate after a set period of time based on the carbon monoxide diffusion radius of the three most recent records.

[0049] In a preferred embodiment of the present invention, the risk spatial distribution module 300 includes: The fire source risk field unit is used to define the fire source risk value at any point in space as the fire source intensity divided by the square of the distance from that point to the fire source, multiplied by the negative attenuation coefficient of the natural constant e, and the product of the distance from that point to the nearest obstacle raised to the power of the product. The fire source intensity is proportional to the product of the current burned area and the rate of temperature rise. The smoke diffusion field unit is used to construct a three-layer Gaussian plume model based on the carbon monoxide sensor network within the building and smoke detectors arranged in the vertical direction. The smoke risk value is defined as the first coefficient multiplied by the ratio of the measured carbon monoxide concentration to the lethal concentration, plus the second coefficient multiplied by the ratio of the visibility benchmark value to the measured visibility.

[0050] In a preferred embodiment of the present invention, the risk spatial distribution module 300 further includes: The personnel density determination unit is used to estimate the personnel density in each fire compartment in real time by using Wi-Fi probes, infrared counting devices and gate data; The risk value correction unit is used to correct the risk value of the area when the personnel density exceeds the preset density: multiply the superposition value of the fire source risk field and the smoke diffusion field in the area by the crowding coefficient. The risk distribution map unit is used to overlay and render the corrected risk values ​​of all regions on a 1m×1m planar grid to generate a comprehensive risk spatial distribution map. The distribution map uses color or grayscale levels to represent the level of risk.

[0051] The above description only details the preferred embodiments of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0052] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0053] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the disclosure in the specification and embodiments. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

Claims

1. A fire alarm linkage control method, characterized in that, The method includes the following steps: Sensor data on carbon monoxide concentration, temperature, flame flashing frequency, and smoke particle size are collected. Environmental noise is filtered out by sliding window mid-range filtering. Fire confidence is calculated by weighting carbon monoxide concentration, temperature rise, and flame flashing frequency. When the confidence level exceeds the first threshold, the fire is confirmed and the fire source is located in three dimensions. The fire spread trend is predicted based on the temperature rise rate and carbon monoxide diffusion radius. Establish fire source risk field and smoke diffusion field, correct risk value based on real-time personnel density multiplied by crowding coefficient, and generate comprehensive risk spatial distribution map; The fire extinguishing coverage area is determined based on the location of the fire source and the predicted burned area, and the fire extinguishing valves are pre-activated. The minimum risk path is calculated using the spatial integral of the comprehensive risk as the cost function, and the evacuation direction is dynamically changed. The decision instructions are encoded into bus protocol frames and broadcast to the field equipment.

2. The fire-fighting linkage control method according to claim 1, characterized in that, The steps for performing three-dimensional fire source localization and predicting the fire spread trend include: Centroid-weighted localization is performed using the fire confidence scores of multiple adjacent sensing nodes: the node coordinates are weighted and averaged to obtain the three-dimensional coordinates of the fire source using the confidence scores of each node as weights. From the moment the fire is confirmed, the rate of temperature rise and the radius of carbon monoxide diffusion in the fire source area are recorded at preset time intervals. The first-order linear extrapolation method is used to estimate the burned area after a set period of time based on the temperature rise rate of the three most recent records, and to estimate the flue gas layer height descent rate after a set period of time based on the carbon monoxide diffusion radius of the three most recent records.

3. The fire-fighting linkage control method according to claim 2, characterized in that, A first-order linear extrapolation method was used to estimate the burned area after a set future time period based on the temperature rise rate of the three most recent records, and to estimate the flue gas layer height descent rate after a set future time period based on the carbon monoxide diffusion radius of the three most recent records. The specific steps are as follows: Step a: Arrange the three consecutive recorded temperature rise rates in chronological order to obtain the first rate, the second rate, and the third rate; calculate the increment of the second rate over the first rate to obtain the increment of the first rate. Calculate the increment of the third rate compared to the second rate to obtain the increment of the second rate; Step b: When both the first rate increment and the second rate increment are greater than zero, the fire is determined to be in the accelerated spread stage, and the ratio of the second rate increment to the first rate increment is used as the acceleration factor; the third rate is multiplied by the acceleration factor to obtain the predicted temperature rise rate. Step c: When the second rate increment is less than or equal to the first rate increment and greater than zero, it is determined that the fire is in the uniformly accelerated spread stage, and the third rate is added to half of the second rate increment to obtain the predicted temperature rise rate. Step d: When the second rate increment is less than or equal to zero, the fire is determined to be in the deceleration and spread stage. The third rate is subtracted from half of the absolute value of the first rate increment to obtain the predicted temperature rise rate. Step e: Divide the predicted temperature rise rate by the temperature rise rate in the most recent historical record to obtain the rate growth ratio; obtain the calorific value of combustible material per unit area, and multiply the calorific value of combustible material per unit area by the rate growth ratio to obtain the fire area expansion coefficient. Then multiply the current burned area by the burned area expansion coefficient to obtain the burned area after a set future time period; Step f: Arrange the carbon monoxide diffusion radii recorded three times in chronological order to obtain the first radius, the second radius, and the third radius; calculate the increment of the second radius relative to the first radius, and simultaneously calculate the increment of the third radius relative to the second radius to obtain the first radius increment and the second radius increment; use the first radius increment as the first rate increment and the second radius increment as the second rate increment, and repeat steps b to d to determine the fire spread stage and calculate the predicted rate increment to obtain the predicted change trend factor of the carbon monoxide diffusion radius; Multiply the current carbon monoxide diffusion radius by the predicted trend factor to obtain the carbon monoxide diffusion radius after a set future time period. Step g: Determine the floor height of the target building based on the three-dimensional coordinates of the fire source; Divide the predicted carbon monoxide diffusion radius by the target building height to obtain the flue gas settling ratio; multiply the flue gas settling ratio by a preset height coefficient, subtract the current flue gas layer height from the product, and take the absolute value to obtain the flue gas layer height descent rate after a set time period in the future.

4. The fire-fighting linkage control method according to claim 3, characterized in that, The steps for establishing the fire source risk field and the smoke diffusion field specifically include: The fire risk value at any point in space is defined as the fire source intensity divided by the square of the distance from that point to the fire source, multiplied by the negative attenuation coefficient of the natural constant e, and the product of the distance from that point to the nearest obstacle raised to the power of the product. The fire source intensity is proportional to the product of the current burned area and the rate of temperature rise. Based on the carbon monoxide sensor network within the building and smoke detectors arranged vertically, a three-layer Gaussian plume model is constructed. The smoke risk value is defined as the first coefficient multiplied by the ratio of the measured carbon monoxide concentration to the lethal concentration, plus the second coefficient multiplied by the ratio of the visibility benchmark value to the measured visibility.

5. The fire-fighting linkage control method according to claim 4, characterized in that, The steps for generating a comprehensive risk spatial distribution map specifically include: The personnel density in each fire compartment is estimated in real time using Wi-Fi probes, infrared counting devices, and gate data. When the population density exceeds the preset density, the risk value of the area is corrected by multiplying the superposition value of the fire source risk field and the smoke diffusion field in the area by the crowding coefficient. The corrected risk values ​​of all regions are overlaid and rendered on a 1m×1m planar grid to generate a comprehensive risk spatial distribution map. The distribution map uses color or grayscale levels to represent the level of risk.

6. The fire-fighting linkage control method according to claim 4, characterized in that, The personnel density in each fire compartment is estimated in real time using Wi-Fi probes, infrared counting devices, and turnstile data. The specific steps are as follows: The number of mobile device signals obtained by Wi-Fi probes in each fire compartment, the number of people passing through the infrared counting device, and the number of people swiping cards recorded in the gate data are timestamped to obtain the original personnel count data from the three sources within the time window. Sliding window mid-range filtering was applied to the raw personnel count data from the three sources within the time window to obtain smoothed counts from Wi-Fi probes, infrared probes, and turnstiles. The difference between the smoothed count from infrared probes and the smoothed count from turnstiles was used as an estimated number of personnel without access cards. The smoothed count of Wi-Fi probes is added to the estimated number of people without access cards to obtain the first fused personnel base; the first fused personnel base is then weighted and averaged with the infrared smoothed count to obtain the second fused personnel base. Obtain the known area and the area occupied by fixed obstacles in each fire compartment that are pre-stored; subtract the known area from the area occupied by fixed obstacles to obtain the effective activity area; divide the second fused personnel base by the effective activity area to obtain the personnel density in the fire compartment.

7. The fire-fighting linkage control method according to claim 6, characterized in that, The steps for determining the fire suppression coverage area and pre-activating fire suppression valves based on the location of the fire source and the predicted burned area specifically include: Calculate the required fire suppression coverage radius, which is equal to the square root of the quotient obtained by multiplying the predicted fire area by pi. In an automatic sprinkler system, select a set of sprinklers whose center of coverage is closest to the fire source and whose coverage radius requirement is the highest, and whose coverage efficiency is the highest. Pre-start their electric valves to the ready-to-spray state, and maintain the pressure of the water supply pipeline but do not spray temporarily. When the rate of temperature rise exceeds 3 degrees Celsius per second, immediately instruct the pre-activated fire extinguishing equipment to spray at full flow.

8. The fire-fighting linkage control method according to claim 7, characterized in that, In an automatic sprinkler system, select a set of sprinklers whose center of coverage is closest to the required coverage radius and whose coverage efficiency is the highest. The specific steps are as follows: Obtain the preset coordinates of all automatic sprinkler heads in the building; based on the preset coordinates of the automatic sprinkler heads, and using the coverage radius requirement as the distance benchmark, calculate the spatial straight-line distance between the three-dimensional coordinates of the coverage center of each sprinkler head and the three-dimensional coordinates of the fire source; select the three sprinkler heads with the smallest absolute value of the difference between the spatial straight-line distance and the distance benchmark as the first candidate sprinkler head group. The comprehensive risk value at the center of the sprinkler coverage area is extracted from the comprehensive risk spatial distribution map. At the same time, the average comprehensive risk value of all grid points on the boundary of the circular area with the fire source as the center and the equivalent radius corresponding to the burned area after a set future time period is extracted. The comprehensive risk value at the center of the sprinkler coverage area is divided by the average comprehensive risk value and the reciprocal is taken to obtain the fire location fitness. The three sprinklers in the first candidate sprinkler group are sorted in order of fire location fitness from high to low, and the sprinkler with the highest fire location fitness is selected as the main extinguishing sprinkler. The coordinates of all nozzles adjacent to the main fire extinguishing nozzle are obtained from the preset database, and the ratio of the comprehensive risk value at the coverage center of the current adjacent nozzle to the comprehensive risk value at the coverage center of the main fire extinguishing nozzle is calculated to obtain the risk complementarity coefficient. The risk complementarity coefficient is filtered through preset filtering rules to obtain the selected risk complementarity coefficient. The nozzles corresponding to the selected risk complementarity coefficients are used as auxiliary fire extinguishing nozzles. The main fire extinguishing nozzle is the core nozzle that meets the requirement of having the closest coverage radius to the fire source and the highest coverage efficiency. The auxiliary fire extinguishing nozzle is a supplementary nozzle that works in conjunction with the core nozzle. The main fire extinguishing nozzle and the auxiliary fire extinguishing nozzle together constitute a set of coordinated fire extinguishing nozzles.

9. The fire-fighting linkage control method according to claim 8, characterized in that, The steps for calculating the minimum risk path using the spatial integral of the overall risk as the cost function specifically include: Based on the current comprehensive risk spatial distribution map, the path cost from any location to any evacuation exit is defined as the integral of the comprehensive risk value of each grid point on that path; The heat-inspired Dijkstra algorithm is used to calculate the minimum cost path to all available evacuation exits starting from the center of each fire compartment, and the exit with the minimum cost and its corresponding path are selected as the recommended evacuation route.

10. The fire-fighting linkage control method according to claim 9, characterized in that, The Dijkstra algorithm, based on heat heuristics, is used to calculate the minimum-cost path to all available evacuation exits starting from the center of each fire compartment. The exit with the minimum cost and its corresponding path are then selected as the recommended evacuation route. The specific steps are as follows: Extract the geometric center coordinates of each fire compartment from the current comprehensive risk spatial distribution map to serve as the initial reference point for personnel gathering in each compartment; For each fire compartment, starting from the initial reference point where people gather within the compartment and ending at each available evacuation exit, the comprehensive risk value of each grid point is calculated using the heat-heuristic Dijkstra algorithm on the planar grid divided by the comprehensive risk spatial distribution map. The comprehensive risk values ​​of each grid point are then summed to obtain the total risk cost of the path. Specifically, the heat-heuristic Dijkstra algorithm is as follows: the edge weight between two adjacent grid points is set to the arithmetic mean of the comprehensive risk values ​​of the two grid points; the priority of nodes is ranked according to the cumulative sum of the comprehensive risk values ​​of all grid points along the straight-line distance from the current node to the current target evacuation exit; the cumulative sum is used as the heat value, and nodes with lower heat values ​​have higher priority during the search process. The total risk cost of paths corresponding to the same fire compartment is compared, and the path with the minimum total risk cost is selected to obtain the target path. The evacuation exit corresponding to the target path is determined as the recommended evacuation exit of the current fire compartment, and the target path is also used as the recommended evacuation route of the current fire compartment.

11. A fire alarm linkage control system, characterized in that, The system employs the fire-fighting linkage control method as described in any one of claims 1 to 10 above, and the system comprises: The fire confidence module is used to collect sensor data on carbon monoxide concentration, temperature, flame flicker frequency, and smoke particle size. It filters out environmental noise through a sliding window mid-range filter and calculates the fire confidence based on a weighted average of carbon monoxide concentration, temperature rise, and flame flicker frequency. The fire source three-dimensional location module is used to confirm the fire and perform three-dimensional location of the fire source when the confidence level exceeds the first threshold, and predict the fire spread trend based on the temperature rise rate and carbon monoxide diffusion radius. The risk spatial distribution module is used to establish the fire source risk field and smoke diffusion field, correct the risk value based on the real-time personnel density multiplied by the crowding coefficient, and generate a comprehensive risk spatial distribution map; The fire-fighting linkage control module is used to determine the fire extinguishing coverage area based on the fire source location and the predicted fire area, and to pre-activate the fire extinguishing valves. It calculates the minimum risk path using the spatial integral of comprehensive risk as the cost function and dynamically changes the evacuation indication direction. The decision instructions are encoded into bus protocol frames and broadcast to the field equipment.

Images