Intelligent fire safety decision-making method and device based on Internet of Things

Abstract

The invention discloses an intelligent fire safety decision-making method and device based on the Internet of Things, and relates to the field of intelligent fire protection. The method comprises the following steps: dividing an internal area of a to-be-monitored industrial factory building into lambda * lambda grids, defining a fire behavior state of the grids, and defining a fire type state, a tuyere state, a distance state, an obstacle state and a fire time state to obtain historical fire statistical data; classifying the statistical data according to a fire type state, a tuyere state, a distance state, an obstacle state and a fire time state during measurement to obtain a classified statistical data sequence, and calculating the probability that the fire state of each grid in the internal area of the industrial plant to be monitored is transferred from a non-fire state to a fire state; the fire state transition probability threshold value is set, spraying operation is carried out on the grid area with the fire state being 1 and the fire state transition probability being within the threshold value range in the industrial factory building to be monitored, the spraying strategy can be dynamically adjusted according to the fire, the spraying accuracy is improved, and the cargo loss is reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent fire protection, and particularly to an intelligent fire safety decision-making method and device based on the Internet of Things. Background Art

[0002] In modern industrial production, the fire safety of industrial plants is of crucial importance. Fires not only threaten the safety of personnel's lives but also may cause huge property losses. Traditional fire sprinkler systems usually activate all or part of the sprinkler heads to extinguish fires when a fire occurs. Although this "one-size-fits-all" method can effectively control the fire, it often causes damage to the goods in the plant due to excessive sprinkling. With the development of technologies such as the Internet of Things, big data, and artificial intelligence, intelligent fire protection systems have gradually become an important development direction in the field of fire safety. Intelligent fire protection systems can achieve early warning and precise decision-making for fires through real-time monitoring and data analysis.

[0003] However, most of the existing intelligent fire protection systems in the prior art only stay at the level of equipment status monitoring and remote control, and often have the following problems: Uncontrollable sprinkling range: 1) The sprinkler systems in the prior art cannot accurately locate the fire source when a fire occurs, resulting in an overly large sprinkling area and unnecessary damage to goods; 2) Lack of intelligent decision-making support: The existing intelligent fire protection systems fail to make full use of Internet of Things data and artificial intelligence algorithms for precise decision-making and cannot dynamically adjust the sprinkling strategy according to the fire situation; 3) Insufficient response speed: The traditional systems have a slow response speed in the initial stage of a fire and cannot quickly activate the sprinkler system, resulting in the spread of the fire. Therefore, to solve the problems existing in the above prior art, the present invention proposes an intelligent fire safety decision-making method and device based on the Internet of Things. Summary of the Invention

[0004] The main purpose of the present invention is to provide an intelligent fire safety decision-making method and device based on the Internet of Things. By accurately locating the fire source, dynamically adjusting the sprinkling range, and combining artificial intelligence algorithms to optimize the sprinkling strategy, the present invention can minimize the sprinkling area and reduce the loss of goods in the plant on the premise of effectively controlling the fire situation. It can effectively solve the problems in the background art.

[0005] To achieve the above object, the technical solution adopted by the present invention is that the intelligent fire safety decision-making method based on the Internet of Things includes: Step 1: Divide the internal area of the industrial plant to be monitored into several λ×λ grids, and define the fire situation state of the th grid ={0, 1}, where when = 0, it indicates that the fire situation state of the th grid is an unburned state; when = 1, it indicates that the fire situation state of the th grid is a burned state; = 1, 2, … λ 2 ; Step Two: Define the fire type status ; Air vent status ; Distance status ; Obstacle status ; Fire time status ; Among them, , , respectively represent that the fire type status is smoldering fire, flashover fire, and non - flashover fire; , when, respectively represent that the air vent status is the upwind and downwind of the ignition point; represents that the distance status is the th kind; represents that the obstacle status is the th kind; represents that the fire time status is the th kind; Among them, the specific classification process steps of the fire type status are as follows: Step S211: Obtain the ignition point image at a fixed sampling time , where , and pre - process the ignition point image to extract the ignition point image with enhanced flame features; Step S212: Respectively count the number of pixels with a gray value equal to 255 in the adjacent ignition point images with enhanced flame features and ; Step S213: Calculate the fire spread change rate and based on the obtained number of pixels, and the calculation formula is: ; ; Step S214: Set the judgment threshold η, and classify the fire type status according to the set judgment threshold and the calculation result of the fire spread change rate . The specific classification principle is: When , the fire type status is flashover fire; When , the fire type status is non - flashover fire; When , the fire type status is smoldering fire.

[0006] The specific classification process steps of the obstacle status are as follows: Step S221: Obtain the The obstacle type in a grid, and calculate the fire exposure time of the th grid according to the obstacle type and the fire resistance time , and the calculation formulas are respectively: ; In the formula, represents the radiant heat flux, , where is the emissivity; is the Stefan-Boltzmann constant; is the flame temperature; is the surface temperature of the barrier; represents the convective heat flux, , is the convective heat transfer coefficient; ; In the formula, is the fire load density of the fire space, with the unit of MJ·m -2 ; is the thermal inertia coefficient of the boundary material; is the ventilation factor of the fire space; the ventilation factor of the fire space is calculated as follows: When the ventilation condition of the fire space is natural ventilation: ; In the formula, represents the area of the ventilation opening, with the unit of square meters; is the height of the ventilation opening, with the unit of meters; is the proportionality coefficient, and ≥1; When the number of ventilation openings in the fire space is greater than one: ; In the formula, represents the area of the th ventilation opening, with the unit of square meters; represents the height of the th ventilation opening, with the unit of meters; is an integer greater than 1; Under complex ventilation conditions, the ventilation factor of the fire space is calculated as: ; In the formula, ; represents the ventilation opening area, with the unit of square meters; represents the height of the ventilation opening from the ground, with the unit of meters; is expressed as the position coefficient. When the ignition point is located at the central position of the ignition space, take = 0.55; when the ignition point is located at a position close to the wall of the ignition space, take = 0.85; when the ignition point is located at the corner position of the ignition space, take = 1; is expressed as the air flow rate in the ignition space, with the unit of kilograms per second; is expressed as the ventilation time, with the unit of seconds; is expressed as the temperature of the ignition point, with the unit of degrees Celsius; is expressed as the ambient temperature, with the unit of degrees Celsius.

[0007] Step S222: According to the fire exposure time and the fire resistance time to determine the blocking state value of the th grid, and the calculation formula is: ; ; Step S223: Arrange the obtained blocking state values of the grids in ascending order to form a sequence of blocking state values, and divide the sequence of blocking state values into equal subsequences, and respectively obtain the 1st quantile , the 2nd quantile ,..., the -1 quantile ; Step S224: Classify the blocking state values using the obtained quantiles, specifically: When , the blocking state value of the th grid is classified as first level; When , the blocking state value of the th grid is classified as second level; And so on, when , the blocking state value of the th grid is classified as level; When , the blocking state value of the th grid is classified as level; Step S225: According to the obtained classification result of the blocking state value , for the Classify the blocking states of the grids as follows: When the blocking state value of the th grid is classified as level one, its blocking state is ; When the blocking state value of the th grid is classified as level two, its blocking state is ; And so on. When the blocking state value of the th grid is classified as level, its blocking state is .

[0008] The distance state and the fire time state are determined according to the threshold method as follows: Step S231: Set the distance state classification thresholds , ,..., according to the distance from the ignition point, and set the fire time state classification thresholds , ,..., according to the time interval from the fire occurrence time; where ; ; Step S232: Set the distance value between the th grid and the ignition point as, and classify the distance state by using the set distance state classification thresholds as follows: When , the distance state of the th grid is ; When , the distance state of the th grid is ; And so on. When , the distance state of the th grid is ; When , the distance state of the th grid is ; Step S233: Set the time interval between the th grid and the fire occurrence time as , and classify the fire time state by using the set fire time state classification thresholds as follows: When , the fire time state of the th grid is ; When it is the fire time status of the th grid is ; and so on. When it is the fire time status of the th grid is ; When it is the fire time status of the .

[0009] Step 3: Obtain historical fire statistics data, classify the statistical data according to the fire type status, air outlet status, distance status, obstacle status, and fire time status during measurement, and obtain the classified statistical data sequence ; Step 4: Calculate the probability that the fire condition status of the th grid in the internal area of the industrial plant to be monitored transfers from the unburned state to the burned state , and the calculation formula is: ; In the formula, represents the statistical frequency of the fire condition status transferring from the unburned state to the burned state in the sequence ; represents the frequency of the fire condition status transfer in the sequence ; Step 5: Set the fire condition status transfer probability threshold , and perform sprinkler operations on the = 1 and ≥ th grid areas in the internal area of the industrial plant to be monitored .

[0010] The intelligent fire safety decision-making device based on the Internet of Things includes: A fire condition recognition module, which is used to obtain images of the internal area of the industrial plant to be monitored at a fixed sampling time , and when a fire is recognized, obtain the fire condition status of the th grid and the fire type status ; A fire condition real-time data acquisition module, which is used to collect the distance value, ignition time interval, obstacle type, and location situation between the th grid area and the ignition point location when a fire occurs, and determine the The air vent status, distance status, obstacle status, and fire time status of each grid; A historical fire situation data acquisition module, configured to obtain historical fire statistical data, classify the statistical data according to the fire type status, air vent status, distance status, obstacle status, and fire time status at the time of measurement, and obtain a classified statistical data sequence ; A data processing module, configured to calculate, according to the sequence the probability that the fire situation status of the th grid in the internal area of the industrial plant to be monitored transfers from an unburned state to a burned state ; A sprinkler operation strategy formulation module, configured to set a threshold for the probability of fire situation status transfer , and perform sprinkler operations on the = 1 and th grid areas in the internal area of the industrial plant to be monitored.

[0011] The device further includes a memory, a processor, and a computer program stored on the memory and executable on the processor.

[0012] The present invention has the following beneficial effects: Compared with the prior art, by dividing the internal area of the industrial plant to be monitored into λ×λ grids, defining the fire situation status of the grids and defining the fire type status, air vent status, distance status, obstacle status, and fire time status, obtaining historical fire statistical data, classifying the statistical data according to the fire type status, air vent status, distance status, obstacle status, and fire time status at the time of measurement, obtaining a classified statistical data sequence, calculating the probability that the fire situation status of each grid in the internal area of the industrial plant to be monitored transfers from an unburned state to a burned state, setting a threshold for the probability of fire situation status transfer, and performing sprinkler operations on the grid areas in the internal area of the industrial plant to be monitored where the fire situation status is 1 and the probability of fire situation status transfer is within the threshold range, the technical solution of the present invention can accurately make a sprinkler decision by using Internet of Things data and artificial intelligence algorithms when the fire source cannot be accurately located during a fire, and dynamically adjust the sprinkler strategy according to the fire situation. Description of the Drawings

[0013] Figure 1 is a flowchart of the method for intelligent fire safety decision-making based on the Internet of Things according to the present invention; Figure 2 is a structural block diagram of the device for intelligent fire safety decision-making based on the Internet of Things according to the present invention. Detailed Embodiments

[0014] The present invention will be further described below in conjunction with specific embodiments. Among them, the attached drawings are only for illustrative purposes, showing only schematic diagrams, not physical diagrams, and should not be construed as a limitation to the present invention. In order to better illustrate the specific embodiments of the present invention, some components in the attached drawings will be omitted, enlarged or reduced, which does not represent the size of the actual product.

[0015] The specific implementation process of the technical solution adopted by the present invention includes the following steps: Step 1: Divide the internal area of the industrial plant to be monitored into several grids of λ×λ. Real-time monitor the situation of the internal area of the industrial plant through an intelligent recognition camera. When a fire breaks out, locate the grid of the fire point area and change the fire situation status of the grid in the fire point area to 1, which is the over-fire state; Step 2: At a fixed sampling time Obtain the fire point image, where >0, and preprocess the fire point image to extract the fire point image with enhanced flame features. Respectively count the number of pixels with a gray value equal to 255 in the adjacent fire point images with enhanced flame features and , calculate the fire spread change rate , and the calculation formula is: , set a judgment threshold η, and classify the fire type status according to the set judgment threshold and the calculation result of the fire spread change rate . The specific classification principle is: When , the fire type status , is flashover fire; When , the fire type status , is non-flashover fire; When , the fire type status , is smoldering fire; In practice, if = 1s, it is appropriate for η to take 350% - 550%.

[0016] Step 3: Measure and obtain the wind direction of the th grid. If the th grid is in the upwind of the fire point area, its air outlet status = 0; if the th grid is in the downwind of the fire point area, its air outlet status = 1; Step 4: According to the grid of the located fire point area, obtain the distance value between the th grid and it. According to the set distance status classification thresholds , ,... Determine the distance status of the th grid, where ; specifically: When , the distance status of the th grid is ; When , the distance status of the th grid is ; And so on. When , the distance status of the th grid is ; When , the distance status of the th grid is .

[0017] Step 5: According to the detected fire occurrence time and the fire occurrence time interval, by setting the fire time status classification thresholds , ,..., Determine the fire time status of the th grid, where ; specifically: When , the fire time status of the th grid is ; When , the fire time status of the th grid is ; And so on. When , the fire time status of the th grid is ; When , the fire time status of the th grid is .

[0018] Step 6: Identify the goods situation in the th grid area inside the industrial plant to be monitored, calculate the fire exposure time and the fire resistance time of the th grid according to the obstacle type, and further calculate and determine the blocking status value of the th grid. The obtained blocking status value of the grid Arrange them in ascending order to form a sequence of blocking state values, and divide the sequence of blocking state values into equal subsequences, and respectively obtain the 1st percentile , the 2nd percentile ,..., the -1th percentile , and classify the blocking state values by using the obtained percentiles. Specifically: When , the blocking state value of the th grid is classified as first level; When , the blocking state value of the th grid is classified as second level; And so on, when , the blocking state value of the th grid is classified as level; When , the blocking state value of the th grid is classified as level; Step S225: According to the obtained classification result of the blocking state value , classify the blocking state of the th grid. Specifically: When the blocking state value of the th grid is classified as first level, its blocking state is ; When the blocking state value of the th grid is classified as second level, its blocking state is ; And so on, when the blocking state value of the th grid is classified as level, its blocking state is .

[0019] Among them, the fire exposure time and the fire resistance time of the th grid, the calculation formulas are respectively: ; In the formula, represents the radiant heat flux, , where is the emissivity; is the Stefan-Boltzmann constant; is the flame temperature; is the surface temperature of the barrier; represents the convective heat flux, , is the convective heat transfer coefficient; wherein, for the radiative heat flux, convective heat flux, flame temperature, and surface temperature of the barrier, they can be measured and obtained through corresponding sensors; ; In the formula, is the fire load density of the fire space, with the unit of MJ·m -2 ; is the thermal inertia coefficient of the boundary material; is the ventilation factor of the fire space; the ventilation factor of the fire space is calculated as follows: When the ventilation condition of the fire space is natural ventilation: ; In the formula, represents the area of the ventilation opening, with the unit of square meters; is the height of the ventilation opening, with the unit of meters; is the proportionality coefficient, and ≥1; When the number of ventilation openings in the fire space is greater than one: ; In the formula, represents the area of the th ventilation opening, with the unit of square meters; represents the height of the th ventilation opening, with the unit of meters; is an integer greater than 1; Under complex ventilation conditions, the calculation formula for the ventilation factor of the fire space is: ; In the formula, ; represents the ventilation opening area, with the unit of square meters; represents the height of the ventilation opening from the ground, with the unit of meters; represents the position coefficient. When the ignition point is located at the central position of the fire space, take = 0.55; when the ignition point is located at the position close to the wall of the fire space, take = 0.85; when the ignition point is located at the corner position of the fire space, take = 1; is expressed as the air flow rate in the fire space, with the unit of kilograms per second; is expressed as the ventilation time, with the unit of seconds; is expressed as the temperature of the ignition point, with the unit of degrees Celsius; is expressed as the ambient temperature, with the unit of degrees Celsius.

[0020] The blocking state value of the th grid, and the calculation formula is: .

[0021] Step 7: According to the historical fire statistics data, classify the statistical data according to the fire type state, air outlet state, distance state, obstacle state, and fire time state at the time of measurement, and obtain the classified statistical data sequence , and according to the sequence calculate the probability that the fire condition state of the th grid in the internal area of the industrial plant to be monitored transfers from the unburned state to the burned state , and the calculation formula is: ; in the formula, represents the statistical frequency of the fire condition state transferring from the unburned state to the burned state in the sequence ; represents the fire condition state transfer frequency in the sequence ; Step 8: Set the fire condition state transfer probability threshold , and for the = 1 and ≥ th grid area in the internal area of the industrial plant to be monitored, perform sprinkler operations.

[0022] Step 9: Repeat the above steps 1-8 until no fire event occurs in the internal area of the industrial plant to be monitored.

[0023] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent fire safety decision-making based on the Internet of Things, characterized in that, Including: Step 1: Divide the internal area of the industrial plant to be monitored into several grids of λ×λ, and define the fire condition of the th grid ={0, 1}, where when = 0, it indicates that the fire condition of the th grid is the unburned state; when = 1, it indicates that the fire condition of the th grid is the burned state; = 1, 2,... λ 2 ; Step 2: Define the fire type status ; Air outlet status ; Distance status ; Obstacle status ; Fire time status ; Among them, , , respectively represent that the fire type status is smoldering fire, flashover fire, non-flashover fire; , When it is, it respectively represents that the air outlet status is the upper air outlet and the lower air outlet of the ignition point; represents that the distance status is the th type; represents that the obstacle status is the th type; represents that the fire time status is the th type; Step 3: Obtain historical fire statistics data, classify the statistical data according to the fire type status, air outlet status, distance status, obstacle status, and fire time status at the time of measurement, and obtain the classified statistical data sequence ; Step 4: According to the sequence calculate the probability that the fire situation of the th grid in the internal area of the industrial plant to be monitored transfers from the unburned state to the burned state , and the calculation formula is: ； Wherein, is expressed as a sequence the statistical frequency of the fire situation state transferring from the unburned state to the burned state in the sequence; is expressed as a sequence the fire situation state transfer frequency in the sequence; Step 5: Set the threshold of the fire situation transition probability , for the interior of the industrial plant to be monitored = 1 and ≥ the th grid area to perform sprinkler operations.

2. The intelligent fire safety decision-making method based on the Internet of Things according to claim 1, wherein, In step two, the specific classification process steps of the fire type status are as follows: Step S211: At a fixed sampling time acquire an ignition point image, where > 0, and preprocess the ignition point image to extract an ignition point image with enhanced flame features; Step S212: Count the number of pixels with a grayscale value equal to 255 in the adjacent ignition point images after the flame characteristics are enhanced respectively and ; Step S213: According to the obtained number of pixels and calculate the rate of change of fire spread , and the calculation formula is: ; Step S214: Set a determination threshold η, and classify the fire type status according to the set determination threshold and the calculation result of the fire spread change rate The specific classification principle is as follows: When ≥ η, the fire type status is flashover fire; When 0.01η ≤ < η, the fire type status is non-flashover fire; When < 0.01η, the fire type status is smoldering fire.

3. The intelligent fire safety decision-making method based on the Internet of Things according to claim 1, wherein In step two, the specific classification process steps of the obstacle status are as follows: Step S221: Obtain the obstacle type in the th grid, and calculate the fire exposure time of the th grid and the fire resistance time according to the obstacle type. The calculation formulas are as follows: ​ ； In the formula, is expressed as the radiant heat flux, , where is the emissivity; is the Stefan-Boltzmann constant; is the flame temperature; is the temperature of the surface of the barrier; is expressed as the convective heat flux, , is the convective heat transfer coefficient; ； In the formula, is the fire load density of the fire space, with the unit of MJ·m -2 ; is the thermal inertia coefficient of the boundary material; is the ventilation factor of the fire space; Step S222: Determine the blocking state value of the nth grid according to the calculation results of the fire exposure time and the fire resistance time . The calculation formula is: ; ; Step S223: The obtained blocking state values of the grid are sorted in ascending order to form a sequence of blocking state values, and the sequence of blocking state values is divided into equal subsequences, and the 1st quantile , the 2nd quantile ,..., the -1th quantile are obtained respectively; Step S224: Classify the blocking state values using the obtained quantiles Specifically, it is as follows: When , the blocking state value of the th grid is classified as first level; When the blocking state value of the nth grid is classified as secondary; And so on, when the blocking state value of the grid is classified as level; When the blocking state value of the th grid is classified as Step S225: Classify the blocking status of the nth grid according to the obtained blocking status value and classification result, specifically: When the blocking state value of the th grid is classified as the first level, its blocking state is ; When the blocking state value of the th grid is classified as secondary, its blocking state is ; And so on, when the blocking state value of the th grid is classified as level, its blocking state is .

4. The intelligent fire safety decision-making method based on the Internet of Things according to claim 1, characterized in that In step two, the distance status and the fire time status are determined according to the threshold method, specifically: Step S231: Set the distance status classification threshold according to the distance from the ignition point , ,... , and set the fire time status classification threshold according to the time interval from the fire occurrence time , ,... ; where ; ; Step S232: Set the distance value between the th grid and the ignition point, and classify the distance state by using the set distance state classification threshold, specifically: When the distance status of the th grid is ; When the distance status of the th grid is ; And so on, when the distance status of the th grid is ; When , the distance state of the th grid is ; Step S233: Set the th grid and the time interval from the fire occurrence to be , and classify the fire time status by using the set classification threshold of the fire time status, specifically: When the fire time status of the th grid is When the fire time status of the th grid is And so on, when the fire time status of the th grid is When , the fire time status of the th grid is .

5. The method for intelligent fire safety decision-making based on the Internet of Things according to claim 3, characterized in that When the ventilation condition of the fire space is natural ventilation, the calculation method of the ventilation factor of the fire space is as follows: ； In the formula, represents the area of the ventilation opening, in square meters; is the height of the ventilation opening, in meters; is a proportionality coefficient, and ≥ 1; When the number of ventilation openings in the fire space is greater than one, the ventilation factor of the fire space is calculated as follows: ； In the formula, represents the area of the th vent, in square meters; represents the height of the th vent, in meters; is an integer greater than 1; Under complex ventilation conditions, the calculation method of the ventilation factor of the fire space is as follows: ； In the formula, ; represents the ventilation opening area, with the unit of square meters; represents the height of the ventilation opening from the ground, with the unit of meters; represents the position coefficient. When the ignition point is at the central position of the fire space, take = 0.55; when the ignition point is at the position close to the wall in the fire space, take = 0.85; when the ignition point is at the corner position of the fire space, take = 1; represents the air flow rate in the fire space, with the unit of kilograms per second; represents the ventilation time, with the unit of seconds; represents the temperature of the ignition point, with the unit of degrees Celsius; represents the ambient temperature, with the unit of degrees Celsius.

6. The intelligent fire safety decision-making device based on the Internet of Things is characterized in that, The device is used to implement the steps of the Internet of Things-based intelligent fire safety decision-making method described in any one of claims 1-5, including: A fire recognition module, which is used to acquire images of the internal area of the industrial plant to be monitored, and when a fire is recognized, acquire the fire situation status of the nth grid and the fire type status ; Fire real-time data acquisition module, which is used to collect, when a fire occurs, the distance value between the th grid area and the location of the ignition point, the ignition time interval, the type of obstacle, and the location situation with respect to the ignition point, and determine the air outlet state, distance state, obstacle state, and fire time state of the th grid according to the collected data; Historical fire data acquisition module, which is used to obtain historical fire statistical data, classify the statistical data according to the fire type status, air outlet status, distance status, obstacle status, and fire time status at the time of measurement, and obtain the classified statistical data sequence ; A data processing module, which is used to calculate, according to the sequence the probability that the fire condition of the nth grid in the internal area of the industrial plant to be monitored transfers from the unburned state to the burned state ; The sprinkler operation strategy formulation module is used to set the threshold of the fire situation state transition probability , and perform sprinkler operations on the and th grid areas inside the industrial plant to be monitored 7. The intelligent fire safety decision-making device based on the Internet of Things according to claim 6, characterized in that, The device further includes a memory, a processor, and a computer program stored on the memory and executable on the processor. Wherein, when the processor executes the program, it can implement the steps of the Internet of Things-based intelligent fire safety decision-making method described in any one of claims 1-5.

Images