A building safety monitoring system based on image recognition

Abstract

The present invention relates to the field of safety supervision, and discloses a building safety supervision system based on image recognition, comprising: an image acquisition module, which acquires images of balcony areas of each household corresponding to each building based on a monitoring camera; an image processing module, which processes and recognizes the acquired balcony area images of each household through a pre-trained convolutional neural network model, thereby identifying the balcony area that extends outward for drying and marking it as a key area; a safety analysis module, which calculates the safety factor according to a first mode if the number of key areas of the current building is 0 and the number of debris on the ground of the building within a monitoring cycle is 0; otherwise, the safety factor is calculated according to a second mode, the safety factor is compared with a preset safety factor interval, and a safety judgment is made according to the comparison result; a safety warning module, which selects whether to issue a warning according to the result of the safety judgment; and a monitoring management module, which controls the monitoring coverage of the monitoring camera of the building according to the warning information.

Description

Technical Field

[0001] The present invention relates to the field of safety supervision, and in particular to a building safety supervision system based on image recognition. Background Art

[0002] Building safety supervision mainly includes fire safety supervision, public security management, property management supervision, public facilities and equipment management supervision, etc. The necessity of building safety supervision system is reflected in improving the level of building safety management, enhancing safety prevention capabilities, improving emergency response speed, optimizing building resource allocation, and improving user experience and satisfaction.

[0003] Although the existing building safety supervision system has improved the safety management level of buildings to a certain extent, it still has some defects. For example, the existing safety supervision means mainly rely on traditional access control systems, fire alarm systems and video surveillance systems, and normal building surveillance cameras are generally installed at the height of the first floor of the building, so it is impossible to effectively supervise the hazards of high-altitude objects thrown from the building. Although there are special cameras with high-altitude object throwing detection functions in the existing technology, there are still some defects. For example, if the camera is improperly installed or the monitoring range is too large, it may infringe on the privacy rights of residents and cause dissatisfaction and complaints from residents.

[0004] Therefore, the problem of how to balance residents' privacy and effective identification of objects thrown from high places is an urgent problem that needs to be solved. Summary of the invention

[0005] The purpose of the present invention is to provide a building safety supervision system based on image recognition to solve the technical problem of being unable to balance the privacy of residents and the effective identification of high-altitude dropped objects.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A building safety monitoring system based on image recognition, comprising:

[0008] The image acquisition module collects images of the balcony area of ​​each household in each building based on the surveillance camera;

[0009] The image processing module processes and recognizes the collected balcony area images of each household through a pre-trained convolutional neural network model, thereby identifying the balcony area that extends out for drying clothes and marking it as a key area;

[0010] Safety analysis module: if the number of key areas of the current building is 0 and the number of debris on the ground of the building is 0 within a monitoring period, the safety factor is calculated according to the first mode; otherwise, the safety factor is calculated according to the second mode, the safety factor is compared with the preset safety factor range, and a safety judgment is made according to the comparison result;

[0011] The safety warning module determines whether to issue a warning based on the results of the safety assessment;

[0012] The monitoring management module controls the monitoring coverage of the monitoring cameras in the building according to the early warning information.

[0013] As a further technical solution, the method for obtaining the safety factor is:

[0014] First mode: Q i =1;

[0015] Second mode: in,

[0016] Among them, Q i is the safety factor of the current building, m is the number of equally divided monitoring units in a monitoring cycle, j is the jth monitoring unit, and p j is the frequency of high-altitude dropping in the jth monitoring unit, w(k1, k2) is the judgment function about the debris on the ground of the building, ω1, ω2, ω3 are weight coefficients, S i is the clutter coefficient of the current building floor, Δf i is the ratio of the debris area on the floor of the building to the total floor area, n a is the number of key areas in the current building, q i is the environmental impact coefficient, g i is the number of green plants placed outside the balcony of the current building, and θ is the preset fixed coefficient.

[0017] As a further technical solution, the safety factor Q i With the preset safety factor interval [Q imin , Q imax ] to make a comparison, and the process of making a safety judgment based on the comparison results is as follows:

[0018] If Q i ∈[Q imin , Q imax ], the safety is judged as medium risk;

[0019] If Q i ∈(0, Q imin ), the safety is judged as low risk;

[0020] If Q i ∈(Q imax , +∞), the safety is judged as high risk;

[0021] Among them, early warnings will be issued for buildings that are judged to be medium-risk or high-risk, while no early warnings will be issued for buildings that are judged to be low-risk.

[0022] As a further technical solution, the working process of the monitoring management module is:

[0023] For buildings that are judged to be high-risk, the surveillance camera's monitoring coverage is adjusted to the maximum to provide full surveillance coverage of the building;

[0024] For buildings judged to be at medium risk, the surveillance cameras are controlled to adjust the monitoring coverage of key areas to the maximum, and the remaining areas are adjusted to the minimum monitoring coverage;

[0025] For buildings judged to be of low risk, the monitoring coverage of the surveillance cameras will be adjusted to the minimum.

[0026] As a further technical solution, the process of obtaining the environmental impact coefficient is as follows:

[0027] Get the weather information of the current building, including wind speed V i and wind direction information;

[0028] By formula:

[0029]

[0030] Calculate the environmental impact coefficient q of the current building i ;

[0031] Among them, γ is the angle between the wind direction and the building surface of the current building, A and B are exponential coefficients, ρ is the preset proportionality coefficient.

[0032] The smaller the angle between the wind direction and the building surface of the current building, the closer the wind direction and the building surface of the current building are to being parallel. When objects are thrown from high altitude, the range of the objects falling is larger under the influence of wind force and wind direction. Therefore, the more harmful objects are thrown from high altitude, the higher the environmental impact coefficient q. i The larger it is, the lower the overall safety factor.

[0033] As a further technical solution, the process of obtaining the judgment function w(k1, k2) about the debris on the floor of the building is:

[0034] Use surveillance cameras to capture images of the current building ground;

[0035] Based on the image recognition software, the types of debris in the image are identified, where the debris types are divided into two categories, one is ignitable objects and the other is non-ignitable objects;

[0036] Mark the debris that is an igniter in the image in red, mark the debris that is not an igniter in green, connect the outer green marks to form a closed area, and obtain the area of ​​the closed area k2;

[0037] Draw a circle with a preset radius d centered on the red mark to obtain the total red influence area k1;

[0038]

[0039] Among them, is the weight factor.

[0040] As a further technical solution, the system further includes a fire linkage alarm module for performing a linkage alarm for a fire caused by an ignition object falling into a critical area.

[0041] As a further technical solution, the working process of the fire linkage alarm module is as follows:

[0042] Detect whether there is an ignition object falling into the critical area through a monitoring camera. If it exists, perform a fire risk analysis; if it does not exist, do not perform the next step of fire risk analysis;

[0043] The process of performing a fire risk analysis is as follows:

[0044] Continuously obtain images of the items being dried in the critical area within multiple unit times to observe the diffusion of the ignition object and the state change of the items being dried;

[0045] Obtain the curve x(t) of the diffusion area of the ignition object changing with time after it falls, and at the same time obtain the curve y(t) of the smoke diameter at the falling position of the ignition object changing with time;

[0046] Through the formula:

[0047]

[0048] Calculate to obtain the fire risk coefficient G;

[0049] Compare the fire risk coefficient G with the preset fire risk threshold G0:

[0050] If G≥G0, it is determined that the probability of a fire occurring is high, and the fire linkage alarm is immediately triggered; if G<G0, it is determined that the probability of a fire occurring is low, and the fire linkage alarm is not triggered;

[0051] Among them, ρ δ is the reference coefficient, N is the number of unit times, δ is the δth unit time, μ1, μ2, μ3 are conversion coefficients, t0, t1 are the start time and end time of the δth unit time respectively, and V i is the wind speed.

[0052] The beneficial effects of the present invention:

[0053] The present invention collects images of the balcony area of ​​each household corresponding to each building through an image acquisition module, thereby providing image information for tracking objects thrown from high places, and then processes and identifies the collected images of the balcony area of ​​each household through a pre-trained convolutional neural network model, thereby identifying the balcony area extended outward for drying clothes and marking it as a key area, thereby pre-dividing the monitoring coverage area. Obviously, the key area requires the largest range of monitoring, and objects thrown from high places pose the greatest risk to the key area. Therefore, it is necessary to focus on safety monitoring and reduce the proportion of residents' privacy. For other areas, focus on residents' privacy and reduce the monitoring scope, thereby achieving the effect of taking into account both effective safety monitoring and residents' privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The present invention will be further described below in conjunction with the accompanying drawings.

[0055] Figure 1 It is a system structure block diagram of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] See also Figure 1 As shown, the present invention is a building safety monitoring system based on image recognition, comprising:

[0058] The image acquisition module collects images of the balcony area of ​​each household in each building based on the surveillance camera;

[0059] The image processing module processes and recognizes the collected balcony area images of each household through a pre-trained convolutional neural network model, thereby identifying the balcony area that extends out for drying clothes and marking it as a key area;

[0060] Safety analysis module: if the number of key areas of the current building is 0 and the number of debris on the ground of the building is 0 within a monitoring period, the safety factor is calculated according to the first mode; otherwise, the safety factor is calculated according to the second mode, the safety factor is compared with the preset safety factor range, and a safety judgment is made according to the comparison result;

[0061] The safety warning module determines whether to issue a warning based on the results of the safety assessment;

[0062] The monitoring management module controls the monitoring coverage of the monitoring cameras in the building according to the early warning information.

[0063] The method for obtaining the safety factor is:

[0064] First mode: Q i =1;

[0065] Second mode: in,

[0066] Among them, Q i is the safety factor of the current building, m is the number of equally divided monitoring units in a monitoring cycle, j is the jth monitoring unit, and p j is the frequency of high-altitude dropping in the jth monitoring unit, w(k1, k2) is the judgment function about the debris on the ground of the building, ω1, ω2, ω3 are weight coefficients, S i is the clutter coefficient of the current building floor, Δf i is the ratio of the debris area on the floor of the building to the total floor area, n a is the number of key areas in the current building, q i is the environmental impact coefficient, g i is the number of green plants placed outside the balcony of the current building, and θ is the preset fixed coefficient.

[0067] The safety factor Q i and the preset safety factor interval [Q imin , Q imax ] to make a comparison, and the process of making a safety judgment based on the comparison results is as follows:

[0068] If Q i ∈[Q imin , Q imax ], the safety is judged as medium risk;

[0069] If Q i ∈(0, Q imin ), the safety is judged as low risk;

[0070] If Q i ∈(Q imax , +∞), the safety is judged as high risk;

[0071] Among them, early warnings will be issued for buildings that are judged to be medium-risk or high-risk, while no early warnings will be issued for buildings that are judged to be low-risk.

[0072] In this embodiment, in order to solve the technical problems raised in the background technology, the image acquisition module is used to collect images of the balcony area of ​​each household corresponding to each building, so as to provide image information for tracking high-altitude objects, and then the pre-trained convolutional neural network model is used to process and identify the collected balcony area images of each household, so as to identify the balcony area extended for drying and mark it as a key area, so as to pre-divide the monitoring coverage area. Obviously, the key area requires the largest range of monitoring, and high-altitude objects pose the greatest risk to the key area. Therefore, it is necessary to focus on safety monitoring and reduce the proportion of residents' privacy. For the remaining areas, focus on residents' privacy and reduce the monitoring range, so as to achieve the effect of taking into account both effective safety monitoring and residents' privacy.

[0073] At the same time, a specific analysis is performed through the security analysis module to calculate the safety factor of each building. Obviously, the higher the safety factor, the higher the probability of high-altitude throwing. Therefore, the focus on residents' privacy is greater at this time, so it is sufficient to maintain the minimum range of monitoring. Specifically, if the number of key areas of the current building is 0 and the number of debris on the ground of the building is 0 within a monitoring cycle, the safety factor is calculated according to the first mode; otherwise, the safety factor is calculated according to the second mode, and the safety factor is compared with the preset safety factor range. A safety judgment is made based on the comparison result, and whether to issue an early warning is selected based on the result of the safety judgment. Finally, the monitoring coverage of the surveillance camera of the building is controlled according to the early warning information, so as to adapt each building to the focus on security monitoring and residents' privacy according to the actual situation, thereby promoting the safety monitoring effect of the entire building and improving the living experience of residents.

[0074] In addition, a specific method for obtaining the safety factor is provided. In the first mode, if there is no debris on the ground of the building for a long time and there is no clothes hanging outside, the probability of fire risk caused by high-altitude throwing is low. Therefore, in the first mode: Q i =1; In the second mode, there are debris on the floor of the building and some items are hung out to dry, so the danger and possibility of high-altitude throwing is greater, so the second mode: in,

[0075] According to the above formula, we can see that the standard deviation The method shows the frequency of high-altitude dropping of the current building. Obviously, the higher the frequency of high-altitude dropping, the lower the safety factor of the building. The frequency of high-altitude dropping is the number of times high-altitude dropping occurs in the monitoring unit. The situation of high-altitude dropping is captured and identified by the existing surveillance camera, which will not be elaborated here. It can be seen that the more key areas there are in the current building, the greater the risk of high-altitude throwing, and therefore the lower the safety factor;

[0076] If the current building's floor clutter coefficient is larger, S i , it means that the more debris falls from the building, the lower the safety factor of the building. It can be seen that if the ratio of the debris area on the floor of the building to the total floor area is Δf i The larger the value, the greater the density of discarded debris, so the safety factor of the building is lower. i The larger the value, the greater the impact on the safety factor; if the number of green plants placed outside the balcony of the current building is g i The more green plants there are, the higher the possibility that they will passively fall off during strong winds or earthquakes, and therefore the lower the safety factor. Through the above technical solution, multiple parameters of building safety monitoring are utilized, and environmental influences and objective conditions are comprehensively considered to establish an assessment model for building safety, so as to improve the accuracy of building safety assessments, and provide effective support for real-time adjustment of the focus on residents' privacy and security monitoring.

[0077] The working process of the monitoring management module is as follows:

[0078] For buildings that are judged to be high-risk, the surveillance camera's monitoring coverage is adjusted to the maximum to provide full surveillance coverage of the building;

[0079] For buildings judged to be at medium risk, the surveillance cameras are controlled to adjust the monitoring coverage of key areas to the maximum, and the remaining areas are adjusted to the minimum monitoring coverage;

[0080] For buildings that are judged to be low-risk, the monitoring coverage of the surveillance camera is controlled to be adjusted to the minimum. Through the above technical solution, a specific monitoring management process is provided, which focuses on comprehensive security monitoring of high-risk buildings, focuses on security monitoring of key areas of medium-risk buildings, and maintains the minimum monitoring range for low-risk buildings.

[0081] The process of obtaining the environmental impact coefficient is as follows:

[0082] Get the weather information of the current building, including wind speed V i and wind direction information;

[0083] By formula:

[0084]

[0085] Calculate the environmental impact coefficient q of the current building i ;

[0086] Among them, γ is the angle between the wind direction and the building surface of the current building, A and B are exponential coefficients, ρ is the preset proportionality coefficient.

[0087] In this embodiment, when the angle between the wind direction and the building surface of the current building is smaller, it means that the current wind direction and the building surface of the current building are closer to being parallel. When there are objects thrown from high altitude, the range of the objects falling is larger under the influence of wind force and wind direction. Therefore, the harm of objects thrown from high altitude is higher, and the environmental impact coefficient q is i The larger the environmental impact coefficient, the lower the overall safety factor; the environmental impact model is established by combining wind speed and wind direction to dynamically update the impact of the environment on the safety factor. Obviously, the larger the environmental impact coefficient, the greater the danger of high-altitude throwing of objects to the building, and therefore the lower the safety factor.

[0088] The process of obtaining the judgment function w(k1, k2) about the debris on the floor of the building is:

[0089] Use surveillance cameras to capture images of the current building ground;

[0090] Based on the image recognition software, the types of debris in the image are identified, where the debris types are divided into two categories, one is ignitable objects and the other is non-ignitable objects;

[0091] Mark the debris that is an igniter in the image in red, mark the debris that is not an igniter in green, connect the outer green marks to form a closed area, and obtain the area of ​​the closed area k2;

[0092] Draw a circle with the red mark as the center and the preset radius d to obtain the total red affected area k1;

[0093]

[0094] in, is the weight factor.

[0095] In this embodiment, a specific calculation formula for the judgment function is provided. It can be seen from the above formula that the more igniting materials there are in the debris discarded under the building in the historical data, the higher the probability of fire caused by high-altitude throwing, and the greater the impact on the safety of the building. Therefore, it increases in the form of an exponential. If there are no igniting materials, although there is a danger, the probability of causing a fire is greatly reduced. The above technical solution effectively refers to the impact of the type of debris dropped from the building on the probability of causing a fire, thereby achieving the purpose of more accurately calculating and evaluating the safety factor of each building.

[0096] The system also includes a fire linkage alarm module, which is used to provide a linkage alarm for fires caused by igniters falling into key areas.

[0097] The working process of the fire linkage alarm module is:

[0098] Detect whether there is an ignition source falling into the critical area through the monitoring camera. If there is, conduct a fire risk analysis; if not, do not proceed to the next step of the fire risk analysis.

[0099] The process of conducting a fire risk analysis is as follows:

[0100] Continuously obtain images of the drying items in the critical area within multiple unit time intervals to observe the spread of the ignition source and the state changes of the dried items.

[0101] Obtain the curve x(t) of the spread area of the ignition source changing with time after it falls, and at the same time obtain the curve y(t) of the smoke diameter at the falling position of the ignition source changing with time.

[0102] Through the formula:

[0103]

[0104] Calculate to obtain the fire risk coefficient G.

[0105] Compare the fire risk coefficient G with the preset fire risk threshold G0:

[0106] If G≥G0, it is judged that the probability of a fire occurring is high, and immediately trigger a fire linkage alarm; if G<G0, it is judged that the probability of a fire occurring is low, and do not trigger a fire linkage alarm.

[0107] Among them, ρ δ is a reference coefficient, N is the number of unit time intervals, δ is the δth unit time interval, μ1, μ2, μ3 are conversion coefficients, t0, t1 are the start time and end time of the δth unit time interval respectively, and V i is the wind speed.

[0108] In this embodiment, in order to reduce the situation where each subsystem of the existing safety supervision module is independent and cannot conduct linkage alarm in a timely manner, a fire linkage alarm module is set up. By identifying whether there is an ignition source falling from the images monitored by the monitoring camera head in real time, if there is, it means there is a possibility of a fire, so it is necessary to focus on monitoring the critical area where it falls, and then conduct a fire risk analysis, in order to provide effective support for realizing timely linkage alarm between the safety monitoring system and the fire alarm system, reduce the probability of false fire alarms and missed alarms, improve the response time of the corresponding disaster relief strategies, and a specific fire risk analysis method is given. By obtaining the curve x(t) of the spread area of the ignition source changing with time after it falls, and at the same time obtaining the curve y(t) of the smoke diameter at the falling position of the ignition source changing with time; then substituting into the formula: The fire risk coefficient G is calculated; obviously, if the igniter diffuses faster and over a larger area after it falls, the probability of a fire is obviously greater. If smoke is generated after the igniter falls and the diameter of the smoke becomes larger and larger, it means that it is always in a state of ignition and presents an increasing trend. An accurate fire risk model is established by combining the above two parameters and wind speed. The larger the fire risk coefficient G, the higher the probability of a fire. Once it exceeds the predetermined fire risk threshold, the fire linkage alarm is immediately triggered to improve the joint working efficiency between multiple systems.

[0109] It should be noted that the calculation formulas and various parameters involved in the calculation in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.

[0110] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A building safety monitoring system based on image recognition, characterized in that: include: The image acquisition module collects images of the balcony area of ​​each household in each building based on the surveillance camera; The image processing module processes and recognizes the collected balcony area images of each household through a pre-trained convolutional neural network model, thereby identifying the balcony area that extends out for drying clothes and marking it as a key area; Safety analysis module: if the number of key areas of the current building is 0 and the number of debris on the ground of the building is 0 within a monitoring period, the safety factor is calculated according to the first mode; otherwise, the safety factor is calculated according to the second mode, the safety factor is compared with the preset safety factor range, and a safety judgment is made according to the comparison result; The safety warning module determines whether to issue a warning based on the results of the safety assessment; The monitoring management module controls the monitoring coverage of the monitoring cameras in the building according to the early warning information.

2. The building safety monitoring system based on image recognition according to claim 1 is characterized in that: The method for obtaining the safety factor is: First mode: Qi=1; Second mode: in, Among them, Qi is the safety factor of the current building, m is the number of equally divided monitoring units in a monitoring period, j is the jth monitoring unit, pj is the frequency of high-altitude throwing in the jth monitoring unit, w(k1, k2) is the judgment function about the debris on the ground of the building, ω1, ω2, ω3 are weight coefficients, Si is the debris coefficient of the current building ground, Δfi is the ratio of the debris area of ​​the building ground to the total ground area, na is the number of key areas of the current building, qi is the environmental impact coefficient, gi is the number of green plants placed outside the balcony of the current building, and θ is a preset fixed coefficient.

3. The building safety supervision system based on image recognition according to claim 1 or 2, characterized in that: The safety factor Qi is compared with the preset safety factor interval Qimin, Qimax], and the process of making a safety judgment based on the comparison result is as follows: If Qi∈[Qimin, Qimax], the safety is judged as medium risk; If Qi∈(0, Qimin), the safety is judged as low risk; If Qi∈(Qimax, +∞), the safety is judged as high risk; Among them, early warnings will be issued for buildings that are judged to be medium-risk or high-risk, while no early warnings will be issued for buildings that are judged to be low-risk.

4. The building safety monitoring system based on image recognition according to claim 3 is characterized in that: The working process of the monitoring management module is as follows: For buildings that are judged to be high-risk, the surveillance camera's monitoring coverage is adjusted to the maximum to provide full surveillance coverage of the building; For buildings judged to be at medium risk, the surveillance cameras are controlled to adjust the monitoring coverage of key areas to the maximum, and the remaining areas are adjusted to the minimum monitoring coverage; For buildings judged to be of low risk, the monitoring coverage of the surveillance cameras will be adjusted to the minimum.

5. The building safety monitoring system based on image recognition according to claim 2 is characterized in that: The process of obtaining the environmental impact coefficient is as follows: Get the weather information of the current building area, including wind speed Vi and wind direction information; By formula: Calculate the environmental impact coefficient qi of the current building; Among them, γ is the angle between the wind direction and the building surface of the current building, A and B are exponential coefficients, ρ is the preset proportionality coefficient.

6. The building safety monitoring system based on image recognition according to claim 2 is characterized in that: The process of obtaining the judgment function w(k1, k2) about the debris on the floor of the building is: Use surveillance cameras to capture images of the current building ground; Based on the image recognition software, the types of debris in the image are identified, where the debris types are divided into two categories, one is ignitable objects and the other is non-ignitable objects; Mark the sundries that are ignition sources in the image with red marks, and mark the sundries that are not ignition sources with green marks. Connect the green marks on the periphery to form a closed area, and obtain the area k2 of the closed area; Draw a circle with a preset radius d centered on the red mark to obtain the total area k1 of the red influence; in, is the weight factor.

7. The building safety monitoring system based on image recognition according to claim 1 is characterized in that: The system further includes a fire linkage alarm module for performing a linkage alarm for a fire caused by an ignition source falling into a key area.

8. The building safety monitoring system based on image recognition according to claim 7 is characterized in that: The working process of the fire linkage alarm module is as follows: Detect whether there is an ignition source falling into the key area through a monitoring camera. If it exists, perform a fire risk analysis; if it does not exist, do not perform the next step of the fire risk analysis; The process of performing a fire risk analysis is as follows: Continuously obtain images of the items being dried in the key area within multiple unit times to observe the diffusion situation of the ignition source and the state change of the items being dried; Obtain the curve x(t) of the diffusion area of the ignition source changing with time after the ignition source falls, and at the same time obtain the curve y(t) of the smoke diameter at the falling position of the ignition source changing with time; Through the formula: Calculate to obtain the fire risk coefficient G; Compare the fire risk coefficient G with a preset fire risk threshold G0: If G≥G0, it is judged that the probability of a fire occurring is high, and the fire linkage alarm is immediately triggered; if G<G0, it is judged that the probability of a fire occurring is low, and the fire linkage alarm is not triggered; Among them, ρ δ is the reference coefficient, N is the number of unit time, δ is the δth unit time, μ1, μ2, μ3 are conversion coefficients, t0, t1 are the start time and end time of the δth unit time respectively, V i is the wind speed.

Images