A fire emergency broadcast system and method based on artificial intelligence
By collecting data and image recognition technology in real-time in high-rise buildings and dynamically adjusting the escape route, the existing fire emergency broadcast system has solved the problem of poor coverage and navigation in high-rise building fires, and improved the safety of personnel escape and monitoring accuracy.
Patent Information
- Application Number
- CN202410509818.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-25
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-04-25
AI Technical Summary
The existing fire emergency broadcasting system is difficult to effectively cover and guide evacuation in high-rise building fires, and the reliability of escape route navigation information is not high, resulting in uncertainty in the safety of personnel's life.
By dividing the building into multiple areas, collecting detection parameters such as smoke concentration, oxygen concentration and ambient temperature in real time, using artificial intelligence to judge the suspected fire and occurrence instructions, combining image recognition and deep learning technology to determine the fire area, dynamically adjust the escape route, and Intelligent guidance is provided through fire emergency broadcast equipment.
Improve the accuracy of fire monitoring and judgment, ensure that personnel obtain the latest escape information and choose the safest route to evacuate, maximize the safety of people's lives and property, monitor the spread of fire in real time and optimize the escape route to bypass high-risk areas.
Smart Images

Figure CN118400061B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of emergency broadcasting, and more specifically, to a fire emergency broadcasting system and method based on artificial intelligence. Background Art
[0002] In recent years, with the rapid urbanization process in my country, the population and building density in many cities have continued to increase; high-rise buildings have become the mainstream due to the small land area and large population, and skyscrapers are everywhere in major cities, but they also bring safety hazards;
[0003] On the one hand, high-rise buildings have complex internal structures and narrow fire passages, making it difficult to evacuate a large number of people. At the same time, high-altitude operations are time-consuming and labor-intensive, which poses a great challenge to firefighting and rescue. Once entering a high-rise building filled with thick smoke, the use time of life-support oxygen supply facilities is limited, which greatly increases the hidden dangers to the life safety of rescuers and trapped people. On the other hand, traditional emergency broadcasting equipment is arranged on lower floors, and it is difficult to cover the whole picture and provide guidance on high-rise fires. In addition, manual triggering and broadcast content updates are cumbersome. Once the fire spreads, it is particularly important to quickly update the evacuation sequence and evacuation route. However, manual operations are difficult to adjust in time, and a slight delay may cause people to be trapped and endanger their lives.
[0004] Of course, there are also intelligent emergency broadcast systems. For example, the Chinese patent with publication number CN112735075A discloses a fire management system based on artificial intelligence; including a fire sensor module, a fire monitoring module, a fire situation collection module, an intelligent analysis module, an indoor analysis module, a fire analysis module, an intelligent processing module, an automatic spraying module, an intelligent alarm module, an intelligent navigation module and an intelligent broadcasting module, a fire center module, a satellite monitoring module, a traffic information module and a route planning module; wherein the fire sensor module includes a temperature sensor, a smoke sensor and a body sensor; this invention can provide evacuation guidance to people in the room through a broadcast loudspeaker while automatically spraying and extinguishing the fire, thereby helping to minimize or avoid casualties;
[0005] However, in the above technology, it is only described that the intelligent navigation module forms the escape route navigation information according to the indoor analysis results, and the broadcasting loudspeaker provides evacuation guidance to the people in the room according to the escape route navigation information; however, the specific process of forming the escape route navigation information is not described, and the process of obtaining the indoor analysis results is also unknown; therefore, the reliability of the escape route navigation information cannot be guaranteed, and there is uncertainty in the life safety of people in an emergency when a fire occurs;
[0006] In view of this, the present invention proposes a fire emergency broadcast system and method based on artificial intelligence to solve the above problems. Summary of the invention
[0007] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a fire emergency broadcast method based on artificial intelligence, comprising:
[0008] Divide the building into n areas, where n is an integer greater than 1;
[0009] Collect detection parameters of n areas; the detection parameters include smoke concentration, oxygen concentration and ambient temperature;
[0010] Calculate the judgment coefficients corresponding to the n areas, determine whether to generate a suspected fire instruction, and mark the area corresponding to the suspected fire instruction as a suspected area;
[0011] Acquire a regional image of the suspected area;
[0012] Analyze the regional image to determine whether a fire occurrence instruction is generated, and mark the suspected area corresponding to the fire occurrence instruction as a fire area;
[0013] Mark the area where the passage in the building is located as the evacuation area, and mark the area in the building that is not marked as the evacuation area as the personnel area; regard multiple adjacent personnel areas without obstacles in the middle as a gathering area, and obtain m escape routes corresponding to each gathering area;
[0014] Analyze the m escape routes in each gathering area to obtain the best escape route for each gathering area;
[0015] Mark evacuation areas adjacent to the fire area as spread areas;
[0016] According to the spreading area, the m escape routes corresponding to each gathering area are retrieved;
[0017] The best escape route for each gathering area will be broadcast through the fire emergency broadcasting equipment.
[0018] Furthermore, the calculation method of the judgment coefficients corresponding to the n regions includes:
[0019]
[0020] Where PX i is the judgment coefficient of the ith region, WN i is the smoke density in the ith area, HW i is the ambient temperature of the ith region, YN i is the oxygen concentration in the i-th region, i∈[1,n], ω1, ω2, ω3 are the preset weight coefficients.
[0021] Furthermore, the method for determining whether to generate a suspected fire instruction includes:
[0022] Get wind speed and population density;
[0023] Methods for obtaining population density include:
[0024] Two parallel detection lines are set at the entrance and exit of the building. The detection line close to the entrance and exit of the building is marked as line one, and the detection line far from the entrance and exit of the building is marked as line two. The detection line is a diode line. When a human body passes through the detection line, the detection line will generate a digital signal.
[0025] If the first line generates a digital signal first and the second line generates a digital signal later, an entry instruction is generated;
[0026] If the second line generates a digital signal first and the first line generates a digital signal later, a leave instruction is generated;
[0027] Count the number of incoming and outgoing instructions respectively, and calculate the number of personnel;
[0028] The expression for the number of personnel is: RS = JS-LS;
[0029] In the formula, RS is the number of personnel, JS is the number of entry instructions, and LS is the number of leaving instructions;
[0030] Get the area of each region; calculate the total area of the region based on the area of each region;
[0031] The expression for the total area of the region is:
[0032] Where ZM is the total area of the region, QM i is the area of the ith region;
[0033] Obtain the population density through the ratio of the number of people to the total area of the region;
[0034] The expression of population density is:
[0035] Where RM is the population density;
[0036] Calculate fire thresholds based on occupancy density and wind speed;
[0037] The expression of fire threshold is:
[0038] In the formula, HZ is the fire threshold, FS is the wind speed, ε1 and ε2 are the preset proportional coefficients;
[0039] Compare the judgment coefficients corresponding to the n areas with the fire thresholds respectively;
[0040] If PX i <HZ, no fire suspected instruction will be generated;
[0041] If PX i ≥HZ, a suspected fire instruction is generated.
[0042] Furthermore, the method for analyzing the regional image includes:
[0043] Use the trained fire analysis model to identify the regional image and output the recognition results, which include fire and normal;
[0044] The fire analysis model training process includes:
[0045] Collect multiple area images in advance, mark each area image as a training image, annotate the environment in each training image, and the annotations include fire and normal; convert fire and normal into digital annotations respectively; divide the annotated training images into a training set and a test set; use the training set to train the fire analysis model, and use the test set to test the fire analysis model; preset an error threshold, and when the mean of the prediction errors of all training images in the test set is less than the error threshold, output the fire analysis model; the fire analysis model is a convolutional neural network model.
[0046] Furthermore, the method for determining whether to generate a fire occurrence instruction includes:
[0047] If the recognition result is normal, no fire occurrence instruction is generated;
[0048] If the recognition result is a fire, a fire occurrence instruction is generated.
[0049] Furthermore, the method for obtaining m escape routes corresponding to each gathering area includes:
[0050] Step 1: Randomly select a gathering area as the initial area, and define the area where the escape exit is located as the target area;
[0051] Step 2: Starting from the initial area, mark the initial area as a visited area, search the adjacent areas of the initial area, select one of the adjacent areas of the initial area as the next area, and mark the next area as a visited area; search the adjacent areas of the next area, select one of the adjacent areas of the next area that is not marked as a visited area as the new next area, and mark the new next area as a visited area; wherein the adjacent area to be searched is an evacuation area and is not a fire area;
[0052] Step 3: Loop step 2 until the adjacent areas of the next area are marked as visited areas, and then the loop ends. At this time, check whether the current area is the target area;
[0053] If the current area is the target area, the depth-first search ends and the route found by the depth-first search is used as the escape route;
[0054] If the current area is not the target area, the depth-first search ends, but the route found by the depth-first search is not used as an escape route.
[0055] Step 4: loop step 2 to step 3, obtain m escape routes corresponding to the gathering area of this loop, and mark the gathering area of this loop as the completed area;
[0056] Step 5: Repeat steps 1 to 4 to obtain m escape routes corresponding to each gathering area, and the gathering area randomly selected in step 1 is the gathering area that is not marked as a completed area.
[0057] Furthermore, the method of obtaining the best escape route for each gathering area includes:
[0058] Step a: Preset initialization temperature T max , minimum temperature T min , cooling coefficient δ and maximum number of iterations And let the current temperature T = T max ;
[0059] Step b: Randomly select a gathering area that has not obtained the best escape route and mark it as a simulation area. Randomly set a feasible solution χ, which is an escape route in the simulation area. The range of the feasible solution χ is the m escape routes in the simulation area.
[0060] Step c: determine the fitness function;
[0061] The expression of the fitness function is:
[0062] Where f is the fitness, y is the number of evacuation areas in the escape route, JL x is the number of evacuation areas between the xth evacuation area and the nearest fire area in the escape route, x∈[1,y], ψ1 and ψ2 are preset weight coefficients and are both greater than 0;
[0063] The number of evacuation areas in the escape route and the number of evacuation areas between each evacuation area in the escape route and the nearest fire area are obtained through the corresponding escape route;
[0064] Step d: Calculate the fitness f corresponding to the feasible solution χ; take the feasible solution χ as the current point, perform random perturbations in the neighborhood of the current point, obtain a new feasible solution χ′, and calculate the fitness f′ corresponding to the new feasible solution χ′;
[0065] Step e: Calculate the fitness difference f″, the expression of the fitness difference f″ is f″=ff′;
[0066] If the fitness difference f″>0, let χ=χ′, that is, assign the value of the new feasible solution χ′ to the feasible solution χ; if the fitness difference f″≤0, calculate the probability p′, and let χ=χ′ according to the probability p′; the expression of probability p′ is: In the formula, e is a natural constant;
[0067] Step f: Repeat steps d to e until the number of loops reaches the maximum number of iterations. When , the loop ends and enters step g;
[0068] Step g: Set the current temperature T = T × δ, that is, cool down the current temperature in step a, and assign the value after cooling to the current temperature; set the maximum number of iterations That is, assign the value of the maximum number of iterations after the reduction to the maximum number of iterations; if the maximum number of iterations after the reduction is not an integer, the maximum number of iterations after the reduction is rounded up so that the maximum number of iterations after the reduction is an integer;
[0069] Step h: Repeat steps d to g until the current temperature T < T min When , the loop ends, the escape route corresponding to the feasible solution χ is obtained and marked as the best escape route;
[0070] Step i: loop steps b to h until each gathering area obtains the corresponding optimal escape route, and the loop ends.
[0071] Furthermore, during the process of personnel escape, detection parameters of the spreading area are collected and analyzed; if the spreading area generates a suspected fire instruction, that is, the spreading area is marked as a suspected area, then the evacuation area adjacent to the spreading area marked as the suspected area is marked as the spreading area.
[0072] Furthermore, the method for reacquiring the m escape routes corresponding to each gathering area is different from the above-mentioned method for acquiring the m escape routes corresponding to each gathering area in that, when searching for an adjacent area as the next area, the adjacent area is neither a fire area nor a spread area; the remaining steps in reacquiring the m escape routes corresponding to each gathering area are consistent with the steps of acquiring the m escape routes corresponding to each gathering area.
[0073] An artificial intelligence-based fire emergency broadcast system, implementing the artificial intelligence-based fire emergency broadcast method, is characterized by comprising:
[0074] The area division module is used to divide the building into n areas, where n is an integer greater than 1;
[0075] A data acquisition module, used to collect detection parameters of n areas;
[0076] A fire judgment module is used to calculate the judgment coefficients corresponding to the n areas, determine whether to generate a suspected fire instruction, and mark the area corresponding to the suspected fire instruction as a suspected area;
[0077] An image acquisition module, used for acquiring regional images of suspected areas;
[0078] A fire determination module is used to analyze the regional image, determine whether to generate a fire occurrence instruction, and mark the suspected area corresponding to the fire occurrence instruction as a fire area;
[0079] The route planning module is used to mark the area where the passage in the building is located as the evacuation area, and mark the area in the building that is not marked as the evacuation area as the personnel area; multiple adjacent personnel areas without obstacles in the middle are regarded as a gathering area, and m escape routes corresponding to each gathering area are obtained;
[0080] The route selection module is used to analyze the m escape routes in each gathering area and obtain the best escape route for each gathering area;
[0081] The route broadcasting module is used to broadcast the best escape route for each gathering area through the fire emergency broadcasting equipment.
[0082] The technical effects and advantages of the fire emergency broadcasting system and method based on artificial intelligence of the present invention are as follows:
[0083] 1. Use sensors to monitor different areas in real time, automatically judge and confirm fires through image recognition, deep learning and other technologies, adjust the best escape route for each gathering area according to the real-time fire situation, and provide intelligent guidance through fire emergency broadcasting equipment; Use artificial intelligence technology to empower the traditional fire emergency broadcasting system, which not only improves the accuracy of fire monitoring and judgment, but also ensures that people in the building obtain the latest escape information and choose the safest route to evacuate, thereby realizing fully automatic and efficient fire emergency broadcasting guidance, and maximizing the protection of people's life and property safety in the event of a fire.
[0084] 2. This embodiment takes into account the dynamic factors of fire spread, can monitor and judge the fire spread in real time, dynamically identify the spread area according to the changes in fire spread, and re-optimize the escape routes of each gathering area based on the spread area, so as to ensure that the guided escape routes are adjusted in real time and bypass the area where the fire occurs, thereby minimizing the risk of people mistakenly entering high-risk areas during the escape process and threatening their personal safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0085] Figure 1This is a schematic diagram of an artificial intelligence-based fire emergency broadcasting system according to Embodiment 1 of the present invention;
[0086] Figure 2 This is a schematic diagram of the detection line setting of Example 1 of the present invention;
[0087] Figure 3 This is a schematic diagram of the aggregation area of Example 1 of the present invention;
[0088] Figure 4 This is a schematic diagram of an artificial intelligence-based fire emergency broadcasting system according to Embodiment 2 of the present invention;
[0089] Figure 5 This is a flow chart of a fire emergency broadcasting method based on artificial intelligence according to Embodiment 3 of the present invention;
[0090] Figure 6 This is a schematic diagram of an electronic device according to Embodiment 4 of the present invention;
[0091] Figure 7 This is a schematic diagram of a storage medium according to Embodiment 5 of the present invention.
[0092] Figure numbers: 1. Building entrance and exit; 2. First line; 3. Second line; 4. Passage; 5. Personnel area; 6. Gathering area; 7. Evacuation area. DETAILED DESCRIPTION
[0093] 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.
[0094] Embodiment 1:
[0095] See also Figure 1 As shown, the fire emergency broadcast system based on artificial intelligence described in this embodiment includes a region division module, a data acquisition module, a fire judgment module, an image acquisition module, a fire determination module, a route planning module, a route selection module and a route broadcast module; each module is connected by wire and / or wireless means to realize data transmission between modules;
[0096] The area division module is used to divide the building into n areas, where n is an integer greater than 1;
[0097] The methods for dividing n regions include:
[0098] Obtain the detection range corresponding to each sensor in the detection device, and sort the multiple detection ranges from large to small, take the detection range ranked last as the minimum detection range, and divide the n areas according to the minimum detection range; wherein the detection device includes an optical smoke sensor, an optical oxygen sensor and a temperature sensor, and an optical smoke sensor, an optical oxygen sensor and a temperature sensor are installed in each area;
[0099] A data acquisition module, used to collect detection parameters of n areas;
[0100] The detection parameters include smoke concentration, oxygen concentration and ambient temperature;
[0101] The smoke concentration is obtained by the optical smoke sensor. The smoke concentration can reflect the extent and location of the fire. The higher the smoke concentration, the greater the fire, and vice versa. At the same time, the location of the fire can be roughly determined based on the smoke concentration in different areas. If the smoke concentration is high in a specific area, the probability that the area is the location of the fire source is high.
[0102] The oxygen concentration is obtained by the optical smoke sensor. The oxygen concentration can reflect the severity of the fire. The lower the oxygen concentration, the bigger the fire. That is, the larger the fire scale, the more oxygen is consumed, and vice versa.
[0103] The ambient temperature is obtained by the temperature sensor. The ambient temperature can reflect the severity of the fire. The higher the ambient temperature, the greater the fire. At the same time, high temperature will accelerate the combustion speed, further aggravate the scale of the fire, and make the fire more likely to spread. The opposite is true.
[0104] A fire judgment module is used to calculate the judgment coefficients corresponding to the n areas, determine whether to generate a suspected fire instruction, and mark the area corresponding to the suspected fire instruction as a suspected area;
[0105] The calculation method of the judgment coefficient corresponding to n regions includes:
[0106]
[0107] Where PX i is the judgment coefficient of the ith region, WN i is the smoke density in the ith area, HW i is the ambient temperature of the ith region, YN i is the oxygen concentration in the i-th region, i∈[1,n], ω1, ω2, ω3 are preset weight coefficients;
[0108] The specific value of the weight coefficient in the formula can be set according to the actual situation. The weight coefficient reflects the influence of the detection parameter on the judgment coefficient. Those skilled in the art can preset the corresponding weight coefficient according to the actual influence of the detection parameter on the judgment coefficient, so as to accurately assess whether a fire occurs;
[0109] It should be noted that since the judgment coefficient is only used to preliminarily judge whether a fire has occurred, the calculated judgment coefficient does not need to be dimensioned, and the calculation of the judgment coefficient is dimensionless calculation;
[0110] Methods for determining whether to generate a suspected fire instruction include:
[0111] Obtain wind speed and population density; wind speed is obtained from the meteorological data of the building location, and the meteorological data of the building location is obtained from the website of the Meteorological Bureau;
[0112] Methods for obtaining population density include:
[0113] See also Figure 2 As shown, two parallel detection lines are set at the entrance and exit of the building, and the detection line close to the building entrance and exit 1 is marked as line 1 2, and the detection line far from the building entrance and exit 1 is marked as line 2 3; the detection line is a diode line. When a human body passes through the detection line, due to the high conductivity of the human body, a shielding effect will be formed to interfere with the electromagnetic field of the detection line, causing the detection line to generate a digital signal; the building entrance and exit 1 is such as an elevator entrance or a staircase entrance, etc.;
[0114] If line 12 generates a digital signal first and line 23 generates a digital signal later, an entry command is generated, indicating that the person first passes through line 12 and then through line 23, that is, the person enters the building;
[0115] If the digital signal is generated first by line 23 and then by line 12, a leaving instruction is generated, which means that the person passes through line 23 first and then line 12, that is, the person leaves the building;
[0116] Count the number of incoming and outgoing instructions respectively, and calculate the number of personnel;
[0117] The expression for the number of personnel is: RS = JS-LS;
[0118] In the formula, RS is the number of personnel, JS is the number of entry instructions, and LS is the number of leaving instructions;
[0119] Obtaining the area of each region, where the area of each region is measured in advance by a person skilled in the art; calculating the total area of the regions according to the area of each region;
[0120] The expression for the total area of the region is:
[0121] Where ZM is the total area of the region, QM i is the area of the ith region;
[0122] Obtain the population density through the ratio of the number of people to the total area of the region;
[0123] The expression of population density is:
[0124] Where RM is the population density;
[0125] Calculate fire thresholds based on occupancy density and wind speed;
[0126] The expression of fire threshold is:
[0127] In the formula, HZ is the fire threshold, FS is the wind speed, ε1 and ε2 are the preset proportional coefficients;
[0128] The specific value of the proportionality coefficient in the formula can be set according to actual conditions. The proportionality coefficient reflects the degree of influence of the population density and wind speed on the fire threshold. Those skilled in the art can preset the corresponding proportionality coefficient according to the actual degree of influence of the population density and wind speed on the fire threshold, so as to accurately calculate the fire threshold;
[0129] It should be noted that the greater the wind speed and the density of people, the lower the preset fire threshold. The reason is that the greater the wind speed, the faster the smoke spreads, resulting in a faster fire spread. A lower fire threshold needs to be preset to increase the sensitivity of fire detection. The greater the density of people, the more time it takes for people to escape. A lower fire threshold can detect the occurrence of fire earlier and provide more time for people to escape. The calculation of the fire threshold is dimensionless calculation.
[0130] Compare the judgment coefficients corresponding to the n areas with the fire thresholds respectively;
[0131] If PX i <HZ, no suspected fire command is generated, indicating that the area in the building is in a normal state and no fire has occurred;
[0132] If PX i ≥HZ, a suspected fire instruction is generated, indicating that the area in the building is in an abnormal state and a fire is suspected, which requires further analysis and judgment;
[0133] An image acquisition module, used for acquiring regional images of suspected areas;
[0134] The regional images are collected by image sensors installed in n areas. The reason for collecting regional images of suspected areas is that there may be misjudgments if only the detection parameters are collected to determine whether a fire has occurred in n areas. On the one hand, various sensors may fail during use, resulting in the collected detection parameters not being consistent with the actual detection parameters. On the other hand, when smoking, baking, welding, etc. occur in the building, the smoke concentration will increase. When a large number of people gather or chemical reactions occur in the building, the oxygen concentration will decrease. When it is hot in summer, a large number of electrical appliances are running in the building or the sunlight is strong, the ambient temperature will rise. All of the above situations may lead to the generation of suspected fire instructions, but no fire has occurred at this time.
[0135] A fire determination module is used to analyze the regional image, determine whether to generate a fire occurrence instruction, and mark the suspected area corresponding to the fire occurrence instruction as a fire area;
[0136] Methods for analyzing regional images include:
[0137] Use the trained fire analysis model to identify the regional image and output the recognition results, which include fire and normal;
[0138] The specific training process of the fire analysis model includes:
[0139] Collect multiple regional images in advance, mark each regional image as a training image, annotate the environment in each training image, and the annotations include fire and normal; convert fire and normal into digital annotations, for example, convert fire into 0 and convert normal into 1; divide the annotated training images into a training set and a test set, use 70% of the training images as the training set and 30% of the training images as the test set; use the training set to train the fire analysis model, and use the test set to test the fire analysis model; preset an error threshold, and when the mean of the prediction errors of all training images in the test set is less than the error threshold, output the fire analysis model; wherein the calculation formula for the mean prediction error is: Where Z P is the prediction error, P is the number of the training image, α P is the predicted annotation corresponding to the P-th group of training images, μP is the actual annotation corresponding to the P-th group of training images, and U is the number of training images in the test set; the error threshold is preset according to the accuracy required by the fire analysis model;
[0140] The fire analysis model mentioned above is specifically a convolutional neural network model;
[0141] The method for determining whether to generate a fire occurrence instruction includes:
[0142] If the recognition result is normal, no fire occurrence instruction is generated, indicating that no fire has occurred in the suspected area;
[0143] If the recognition result is a fire, a fire occurrence instruction is generated, indicating that a fire has occurred in the suspected area;
[0144] Route Planning Module, see Figure 3 As shown, it is used to mark the area where the passage 4 in the building is located as the evacuation area 7, and mark the area in the building that is not marked as the evacuation area 7 as the personnel area 5; multiple adjacent personnel areas 5 without obstacles in the middle are regarded as a gathering area 6, and m escape routes corresponding to each gathering area 6 are obtained; obstacles are such as walls, doors, windows, etc.; the thick solid line in the figure is the outline of the gathering area 6, the dotted line is the outline of the evacuation area 7, and the oblique line represents the passage 4 in the building;
[0145] It should be noted that the reason for treating multiple adjacent personnel areas 5 without obstacles in between as a gathering area 6 is that usually, the personnel in the building are not in the passage 4. If the building is an office building, the personnel are in the company. If the building is a residential building, the personnel are at home. The company and home are the gathering areas 6. If a fire occurs, the personnel will evacuate from the gathering area 6 to the outside of the building through the escape exit. At this time, the route planning is to plan the route of the personnel from the gathering area 6 to the area where the escape exit is located, and there is no need to plan the route within the gathering area 6.
[0146] The method for obtaining m escape routes corresponding to each gathering area 6 includes:
[0147] Step 1: Randomly select a gathering area 6 as the initial area, and define the area where the escape exit is located as the target area;
[0148] Step 2: Starting from the initial area, mark the initial area as a visited area, search the adjacent areas of the initial area, select one of the adjacent areas of the initial area as the next area, and mark the next area as a visited area; search the adjacent areas of the next area, select one of the adjacent areas of the next area that is not marked as a visited area as a new next area, and mark the new next area as a visited area; wherein the adjacent area to be searched is the evacuation area 7, and is not a fire area;
[0149] Step 3: Loop step 2 until the adjacent areas of the next area are marked as visited areas, and then the loop ends. At this time, check whether the current area is the target area;
[0150] If the current area is the target area, the depth-first search ends and the route found by the depth-first search is used as the escape route;
[0151] If the current area is not the target area, the depth-first search ends, but the route found by the depth-first search is not used as an escape route.
[0152] Step 4: loop step 2 to step 3, obtain m escape routes corresponding to the gathering area 6 of this loop, and the m escape routes are all different escape routes because the next area selected is different in each depth-first search; mark the gathering area 6 of this loop as the completed area;
[0153] Step 5: Repeat steps 1 to 4 to obtain m escape routes corresponding to each gathering area 6, and the gathering area 6 randomly selected in step 1 is the gathering area 6 that is not marked as a completed area;
[0154] A route selection module, used to analyze the m escape routes of each gathering area 6 and obtain the best escape route of each gathering area 6;
[0155] Methods for obtaining the best escape route for each gathering area6 include:
[0156] Step a: Preset initialization temperature T max , minimum temperature T min , cooling coefficient δ and maximum number of iterations And let the current temperature T = T max ;
[0157] Step b: randomly select a gathering area 6 that has not obtained the best escape route and mark it as a simulation area, randomly set a feasible solution χ, the feasible solution χ is an escape route in the simulation area, and the range of the feasible solution χ is the m escape routes in the simulation area;
[0158] Step c: determine the fitness function;
[0159] The expression of the fitness function is:
[0160] Where f is the fitness, y is the number of evacuation areas 7 in the escape route, JL x is the number of evacuation areas 7 between the xth evacuation area 7 and the nearest fire area in the escape route, x∈[1,y], ψ1 and ψ2 are preset weight coefficients and are both greater than 0;
[0161] The specific value of the weight coefficient in the formula can be set according to the actual situation. The weight coefficient reflects the influence of the number of evacuation areas 7 in the escape route and the sum of the number of evacuation areas 7 between each evacuation area 7 in the escape route and the nearest fire area on the safety of the route. Those skilled in the art can preset the corresponding weight coefficient according to the actual influence of the number of evacuation areas 7 in the escape route and the sum of the number of evacuation areas 7 between each evacuation area 7 in the escape route and the nearest fire area on the safety of the route, so as to accurately evaluate the safety of the route;
[0162] It should be noted that the number of evacuation areas 7 in the escape route and the number of evacuation areas 7 between each evacuation area 7 in the escape route and the nearest fire area are both obtained through the corresponding escape route, and the number of evacuation areas 7 in the escape route and the number of evacuation areas 7 between each evacuation area 7 in the escape route and the nearest fire area will affect the safety level of the corresponding escape route. The fewer the number of evacuation areas 7 in the escape route, the shorter the required escape time is, and the more the life safety of the personnel can be guaranteed, that is, the higher the safety level is, the smaller the corresponding fitness is, and vice versa; the more the number of evacuation areas 7 between each evacuation area 7 in the escape route and the nearest fire area is, the farther the distance between each evacuation area 7 and the nearest fire area is, and the safer the personnel are when escaping, that is, the higher the safety level is, the smaller the corresponding fitness is, and vice versa;
[0163] Step d: Calculate the fitness f corresponding to the feasible solution χ; take the feasible solution χ as the current point, perform random perturbations in the neighborhood of the current point, obtain a new feasible solution χ′, and calculate the fitness f′ corresponding to the new feasible solution χ′;
[0164] Step e: Calculate the fitness difference f″, the expression of the fitness difference f″ is f″=ff′;
[0165] If the fitness difference f″>0, let χ=χ′, that is, assign the value of the new feasible solution χ′ to the feasible solution χ; if the fitness difference f″≤0, calculate the probability p′, and let χ=χ′ according to the probability p′; the expression of probability p′ is: In the formula, e is a natural constant;
[0166] Step f: Repeat steps d to e until the number of loops reaches the maximum number of iterations. When , the loop ends and enters step g;
[0167] Step g: Set the current temperature T = T × δ, that is, cool down the current temperature in step a, and assign the value after cooling to the current temperature; set the maximum number of iterations That is, assign the value of the maximum number of iterations after the reduction to the maximum number of iterations; if the maximum number of iterations after the reduction is not an integer, the maximum number of iterations after the reduction is rounded up so that the maximum number of iterations after the reduction is an integer;
[0168] Step h: Repeat steps d to g until the current temperature T < T min When , the loop ends, the escape route corresponding to the feasible solution χ is obtained and marked as the best escape route;
[0169] Step i: looping steps b to h until each gathering area 6 obtains the corresponding optimal escape route, and the loop ends;
[0170] It should be noted that the initialization temperature T max , minimum temperature T min , cooling coefficient δ and maximum number of iterations As preset parameters, the preset parameters are preset by those skilled in the art. In the selection of the best historical escape route, for the same gathering area 6, multiple groups of different preset parameters are sequentially pre-set, and the best escape route is sequentially obtained by using the simulated annealing algorithm. The corresponding fitness is calculated according to the obtained multiple best escape routes, and the preset parameter corresponding to the minimum fitness is used as the preset parameter corresponding to the gathering area 6. The preset parameters corresponding to multiple gathering areas 6 are obtained by analogy, and the mean values of the multiple preset parameters (i.e., the mean value of the initialization temperature, the mean value of the lowest temperature, the mean value of the cooling coefficient, and the mean value of the maximum number of iterations) are used as the preset initialization temperature T in step a. max , minimum temperature T min , cooling coefficient δ and maximum number of iterations
[0171] A route broadcasting module, used to broadcast the best escape route for each gathering area 6 through the fire emergency broadcasting equipment;
[0172] A fire emergency broadcast device is installed in each gathering area 6, and the best escape route of each gathering area 6 is transmitted to the personnel in the corresponding gathering area 6 through the fire emergency broadcast device;
[0173] This embodiment uses sensors to monitor different areas in real time, automatically judges and confirms fires through image recognition, deep learning and other technologies, adjusts the best escape route for each gathering area 6 according to the real-time fire situation, and provides intelligent guidance through fire emergency broadcasting equipment; using artificial intelligence technology to empower traditional fire emergency broadcasting systems not only improves the accuracy of fire monitoring and judgment, but also ensures that people in the building obtain the latest escape information and choose the safest route to evacuate, thereby realizing fully automatic and efficient fire emergency broadcasting guidance, and maximizing the protection of people's life and property safety in the event of a fire.
[0174] Embodiment 2:
[0175] See also Figure 4 As shown, this embodiment further improves the design on the basis of embodiment 1. When a fire occurs in a building, if the fire is not put out in time, the fire will spread to more areas in the building during the escape process of personnel, that is, the area that was not originally marked as the fire area is marked as the fire area. If the personnel still escape according to the best escape route obtained in embodiment 1 at this time, the situation where the fire area appears in the best escape route will occur, that is, the personnel encounter the fire during the escape process, thereby endangering the life safety of the personnel; therefore, this embodiment provides a fire emergency broadcast system based on artificial intelligence, which also includes a fire spread judgment module and a route update module;
[0176] The fire spread judgment module is used to mark the evacuation area 7 adjacent to the fire area as a spread area, that is, the area where the fire is about to spread; during the process of personnel escape, the detection parameters of the spread area are collected and analyzed; if the spread area generates a suspected fire instruction, that is, the spread area is marked as a suspected area, then the evacuation area 7 adjacent to the spread area marked as the suspected area is marked as a spread area; and so on, during the process of personnel escape, it is judged whether the evacuation area 7 is marked as a spread area;
[0177] A route updating module is used to re-acquire m escape routes corresponding to each gathering area 6 according to the spreading area;
[0178] The method for re-obtaining the m escape routes corresponding to each gathering area 6 is different from the above method for obtaining the m escape routes corresponding to each gathering area 6 in that, when searching for an adjacent area as the next area, the adjacent area is neither a fire area nor a spreading area; the remaining steps in re-obtaining the m escape routes corresponding to each gathering area 6 are consistent with the above steps for obtaining the m escape routes corresponding to each gathering area 6;
[0179] This embodiment takes into account the dynamic factors of fire spread, can monitor and judge the fire spread in real time, dynamically identify the spread area according to the changes in fire spread, and re-optimize the escape routes of each gathering area 6 based on the spread area, thereby ensuring that the guided escape routes are adjusted in real time and bypass the area where the fire occurs, thereby minimizing the risk of people mistakenly entering high-risk areas during the escape process and threatening their personal safety.
[0180] Embodiment 3:
[0181] See also Figure 5 As shown, the part not described in detail in this embodiment can be found in the description of embodiments 1 and 2, and a fire emergency broadcast method based on artificial intelligence is provided, including:
[0182] Divide the building into n areas, where n is an integer greater than 1;
[0183] Collect detection parameters of n areas; the detection parameters include smoke concentration, oxygen concentration and ambient temperature;
[0184] Calculate the judgment coefficients corresponding to the n areas, determine whether to generate a suspected fire instruction, and mark the area corresponding to the suspected fire instruction as a suspected area;
[0185] Acquire a regional image of the suspected area;
[0186] Analyze the regional image to determine whether a fire occurrence instruction is generated, and mark the suspected area corresponding to the fire occurrence instruction as a fire area;
[0187] Mark the area where the passage 4 in the building is located as the evacuation area 7, and mark the area in the building that is not marked as the evacuation area 7 as the personnel area 5; regard multiple adjacent personnel areas 5 without obstacles in between as a gathering area 6, and obtain m escape routes corresponding to each gathering area 6;
[0188] Analyze the m escape routes of each gathering area 6 to obtain the best escape route of each gathering area 6;
[0189] Mark the evacuation area 7 adjacent to the fire area as the spreading area;
[0190] According to the spreading area, m escape routes corresponding to each gathering area 6 are retrieved;
[0191] The best escape route for each gathering area 6 is broadcasted through the fire emergency broadcasting equipment.
[0192] Furthermore, the calculation method of the judgment coefficients corresponding to the n regions includes:
[0193]
[0194] Where PX i is the judgment coefficient of the ith region, WN i is the smoke density in the ith area, HW i is the ambient temperature of the ith region, YN i is the oxygen concentration in the i-th region, i∈[1,n], ω1, ω2, ω3 are the preset weight coefficients.
[0195] Furthermore, the method for determining whether to generate a suspected fire instruction includes:
[0196] Get wind speed and population density;
[0197] Methods for obtaining population density include:
[0198] Two parallel detection lines are set at the entrance and exit of the building. The detection line close to the entrance and exit 1 is marked as line 2, and the detection line far from the entrance and exit 1 is marked as line 3. The detection line is a diode line. When a human body passes through the detection line, the detection line will generate a digital signal.
[0199] If the first line 2 generates a digital signal first and the second line 3 generates a digital signal later, an entry instruction is generated;
[0200] If the digital signal is generated first by line 2, line 3, and then by line 1, line 2, a leave instruction is generated;
[0201] Count the number of incoming and outgoing instructions respectively, and calculate the number of personnel;
[0202] The expression for the number of personnel is: RS = JS-LS;
[0203] In the formula, RS is the number of personnel, JS is the number of entry instructions, and LS is the number of leaving instructions;
[0204] Get the area of each region; calculate the total area of the region based on the area of each region;
[0205] The expression for the total area of the region is:
[0206] Where ZM is the total area of the region, QM i is the area of the ith region;
[0207] Obtain the population density through the ratio of the number of people to the total area of the region;
[0208] The expression of population density is:
[0209] Where RM is the population density;
[0210] Calculate fire thresholds based on occupancy density and wind speed;
[0211] The expression of fire threshold is:
[0212] In the formula, HZ is the fire threshold, FS is the wind speed, ε1 and ε2 are the preset proportional coefficients;
[0213] Compare the judgment coefficients corresponding to the n areas with the fire thresholds respectively;
[0214] If PX i <HZ, no fire suspected instruction will be generated;
[0215] If PX i ≥HZ, a suspected fire instruction is generated.
[0216] Furthermore, the method for analyzing the regional image includes:
[0217] Use the trained fire analysis model to identify the regional image and output the recognition results, which include fire and normal;
[0218] The fire analysis model training process includes:
[0219] Collect multiple area images in advance, mark each area image as a training image, annotate the environment in each training image, and the annotations include fire and normal; convert fire and normal into digital annotations respectively; divide the annotated training images into a training set and a test set; use the training set to train the fire analysis model, and use the test set to test the fire analysis model; preset an error threshold, and when the mean of the prediction errors of all training images in the test set is less than the error threshold, output the fire analysis model; the fire analysis model is a convolutional neural network model.
[0220] Furthermore, the method for determining whether to generate a fire occurrence instruction includes:
[0221] If the recognition result is normal, no fire occurrence instruction is generated;
[0222] If the recognition result is a fire, a fire occurrence instruction is generated.
[0223] Furthermore, the method for obtaining m escape routes corresponding to each gathering area 6 includes:
[0224] Step 1: Randomly select a gathering area 6 as the initial area, and define the area where the escape exit is located as the target area;
[0225] Step 2: Starting from the initial area, mark the initial area as a visited area, search the adjacent areas of the initial area, select one of the adjacent areas of the initial area as the next area, and mark the next area as a visited area; search the adjacent areas of the next area, select one of the adjacent areas of the next area that is not marked as a visited area as a new next area, and mark the new next area as a visited area; wherein the adjacent area to be searched is the evacuation area 7, and is not a fire area;
[0226] Step 3: Loop step 2 until the adjacent areas of the next area are marked as visited areas, and then the loop ends. At this time, check whether the current area is the target area;
[0227] If the current area is the target area, the depth-first search ends and the route found by the depth-first search is used as the escape route;
[0228] If the current area is not the target area, the depth-first search ends, but the route found by the depth-first search is not used as an escape route.
[0229] Step 4: loop step 2 to step 3, obtain m escape routes corresponding to the gathering area 6 of this loop, and mark the gathering area 6 of this loop as a completed area;
[0230] Step 5: Repeat steps 1 to 4 to obtain m escape routes corresponding to each gathering area 6, and the gathering area 6 randomly selected in step 1 is the gathering area 6 that is not marked as a completed area.
[0231] Furthermore, the method for obtaining the best escape route for each gathering area 6 includes:
[0232] Step a: Preset initialization temperature T max , minimum temperature T min , cooling coefficient δ and maximum number of iterations And let the current temperature T = T max ;
[0233] Step b: randomly select a gathering area 6 that has not obtained the best escape route and mark it as a simulation area, randomly set a feasible solution χ, the feasible solution χ is an escape route in the simulation area, and the range of the feasible solution χ is the m escape routes in the simulation area;
[0234] Step c: determine the fitness function;
[0235] The expression of the fitness function is:
[0236] Where f is the fitness, y is the number of evacuation areas 7 in the escape route, JL x is the number of evacuation areas 7 between the xth evacuation area 7 and the nearest fire area in the escape route, x∈[1,y], ψ1 and ψ2 are preset weight coefficients and are both greater than 0;
[0237] The number of evacuation areas 7 in the escape route and the number of evacuation areas 7 between each evacuation area 7 and the nearest fire area in the escape route are obtained through the corresponding escape route;
[0238] Step d: Calculate the fitness f corresponding to the feasible solution χ; take the feasible solution χ as the current point, perform random perturbations in the neighborhood of the current point, obtain a new feasible solution χ′, and calculate the fitness f′ corresponding to the new feasible solution χ′;
[0239] Step e: Calculate the fitness difference f″, the expression of the fitness difference f″ is f″=ff′;
[0240] If the fitness difference f″>0, let χ=χ′, that is, assign the value of the new feasible solution χ′ to the feasible solution χ; if the fitness difference f″≤0, calculate the probability p′, and let χ=χ′ according to the probability p′; the expression of probability p′ is: In the formula, e is a natural constant;
[0241] Step f: Repeat steps d to e until the number of loops reaches the maximum number of iterations. When , the loop ends and enters step g;
[0242] Step g: Set the current temperature T = T × δ, that is, cool down the current temperature in step a, and assign the value after cooling to the current temperature; set the maximum number of iterations That is, assign the value of the maximum number of iterations after the reduction to the maximum number of iterations; if the maximum number of iterations after the reduction is not an integer, the maximum number of iterations after the reduction is rounded up so that the maximum number of iterations after the reduction is an integer;
[0243] Step h: Repeat steps d to g until the current temperature T < T min When , the loop ends, the escape route corresponding to the feasible solution χ is obtained and marked as the best escape route;
[0244] Step i: loop steps b to h until each gathering area 6 obtains the corresponding optimal escape route, and the loop ends.
[0245] Furthermore, during the process of personnel escape, detection parameters of the spreading area are collected and analyzed; if the spreading area generates a suspected fire instruction, that is, the spreading area is marked as a suspected area, then the evacuation area 7 adjacent to the spreading area marked as the suspected area is marked as the spreading area.
[0246] Furthermore, the method for reacquiring the m escape routes corresponding to each gathering area 6 is different from the above-mentioned method for acquiring the m escape routes corresponding to each gathering area 6 in that, when searching for an adjacent area as the next area, the adjacent area is neither a fire area nor a spread area; the remaining steps in reacquiring the m escape routes corresponding to each gathering area 6 are consistent with the steps of acquiring the m escape routes corresponding to each gathering area 6.
[0247] Embodiment 4:
[0248] See also Figure 6 As shown, according to another aspect of the present application, an electronic device 500 is also provided. The electronic device 500 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are executed by one or more processors, the fire emergency broadcast method based on artificial intelligence as described above can be executed.
[0249] The method or system according to the embodiment of the present application can also be used by Figure 6 The electronic device architecture shown in FIG. Figure 6 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, ROM 503, RAM 504, a communication port 505 connected to a network, an input / output 506, a hard disk 507, etc. The storage device in the electronic device 500, such as ROM 503 or hard disk 507, may store an artificial intelligence-based fire emergency broadcast method provided by the present application. Furthermore, the electronic device 500 may also include a user interface 508. Of course, Figure 6 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 6 One or more components of an electronic device are shown.
[0250] Embodiment 5:
[0251] See also Figure 7 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, an artificial intelligence-based fire emergency broadcast method according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.
[0252] In addition, according to the implementation of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided by the present application, for example: a fire emergency broadcast method based on artificial intelligence. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.
[0253] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0254] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0255] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0256] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0257] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0258] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0259] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
[0260] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A fire emergency broadcasting method based on artificial intelligence, It is characterized in that include: Divide the building into n areas, where n is an integer greater than 1; Collect detection parameters of n areas; The detection parameters include smoke concentration, oxygen concentration and ambient temperature; Calculate the judgment coefficients corresponding to the n areas based on the detection parameters, determine whether to generate a suspected fire instruction, and mark the area corresponding to the suspected fire instruction as a suspected area; The method for determining whether to generate a suspected fire instruction includes: Get wind speed and population density; The method for obtaining the population density includes: Two parallel detection lines are set at the entrance and exit of the building. The detection line close to the entrance and exit (1) is marked as line one (2), and the detection line far from the entrance and exit (1) is marked as line two (3). The detection line is a diode line. When a human body passes through the detection line, the detection line will generate a digital signal. If the first line (2) generates a digital signal first and the second line (3) generates a digital signal later, an entry instruction is generated; If the second line (3) generates a digital signal first and the first line (2) generates a digital signal later, a leave instruction is generated; Count the number of incoming and outgoing instructions respectively, and calculate the number of personnel; The expression for the number of people is: ; In the formula, is the number of personnel, is the number of entry instructions, is the number of leaving instructions; Get the area of each region; calculate the total area of the region based on the area of each region; The expression for the total area of the region is: ; In the formula, is the total area of the region, is the area of the ith region; Obtain the population density through the ratio of the number of people to the total area of the region; The expression of population density is: ; In the formula, is the population density; Calculate fire thresholds based on occupancy density and wind speed; The expression of fire threshold is: ; In the formula, is the fire threshold, is the wind speed, , is the preset proportional coefficient; Compare the judgment coefficients corresponding to the n areas with the fire thresholds respectively; like , then no fire suspected instruction is generated; is the judgment coefficient of the ith region; like , a suspected fire instruction is generated; Acquire a regional image of the suspected area; Analyze the regional image to determine whether a fire occurrence instruction is generated, and mark the suspected area corresponding to the fire occurrence instruction as a fire area; The area where the passage (4) in the building is located is marked as an evacuation area (7), and the area in the building that is not marked as the evacuation area (7) is marked as a personnel area (5); multiple adjacent personnel areas (5) without obstacles in between are regarded as a gathering area (6), and m escape routes corresponding to each gathering area (6) are obtained; Analyze the m escape routes of each gathering area (6) to obtain the best escape route of each gathering area (6); Mark the evacuation area (7) adjacent to the fire area as the spreading area; According to the spreading area, the m escape routes corresponding to each gathering area (6) are retrieved; Analyze the m escape routes of each gathering area (6) to obtain the best escape route of each gathering area (6); Broadcast the best escape route for each gathering area (6) through the fire emergency broadcasting equipment; The method for obtaining the best escape route for each gathering area (6) comprises: Step a: Preset initialization temperature , minimum temperature , Temperature reduction coefficient and the maximum number of iterations , and let the current temperature ; Step b: Randomly select a gathering area (6) that does not have the best escape route and mark it as a simulation area. Randomly set a feasible solution , a feasible solution is an escape route in the simulation area, and the feasible solution is The range of is the m escape routes in the simulation area; Step c: determine the fitness function; The expression of the fitness function is: ; In the formula, For fitness, is the number of evacuation zones (7) in the escape route, is the number of evacuation areas (7) between the xth evacuation area (7) and the nearest fire area in the escape route, , , are preset weight coefficients and are all greater than 0; The number of evacuation areas (7) in the escape route and the number of evacuation areas (7) between each evacuation area (7) in the escape route and the nearest fire area are obtained through a corresponding escape route; Step d: Calculate feasible solutions The corresponding fitness ; with feasible solutions For the current point, perform random perturbations in the neighborhood of the current point to obtain a new feasible solution , and calculate the new feasible solution The corresponding fitness ; Step e: Calculate the fitness difference , fitness difference The expression is ; If the fitness difference , then let , the new feasible solution The value of is given to the feasible solution ; If the fitness difference , then calculate the probability , according to the probability make ; Probability The expression is: ; In the formula, e is a natural constant; Step f: Repeat steps d to e until the number of loops reaches the maximum number of iterations. When , the loop ends and enters step g; Step g: Set the current temperature , that is, cool down the current temperature in step a, and assign the value after cooling to the current temperature; let the maximum number of iterations , that is, assign the value of the reduced maximum number of iterations to the maximum number of iterations; if the reduced maximum number of iterations is not an integer, the reduced maximum number of iterations is rounded up to make it an integer; Step h: Repeat steps d to g until the current temperature reaches When , the loop ends and a feasible solution is obtained The corresponding escape route is marked as the best escape route; Step i: loop steps b to h until each gathering area (6) obtains the corresponding optimal escape route, and the loop ends.
2. According to the fire emergency broadcasting method based on artificial intelligence in claim 1, It is characterized in that The calculation method of the judgment coefficients corresponding to the n regions includes: ; In the formula, is the judgment coefficient of the ith region, is the smoke density in the ith area, is the ambient temperature of the ith region, is the oxygen concentration in the ith region, , , , is the preset weight coefficient.
3. According to the fire emergency broadcasting method based on artificial intelligence in claim 2, It is characterized in that The method for analyzing the regional image comprises: Use the trained fire analysis model to identify the regional image and output the recognition results, which include fire and normal; The fire analysis model training process includes: Collect multiple area images in advance, mark each area image as a training image, annotate the environment in each training image, and the annotations include fire and normal; convert fire and normal into digital annotations respectively; divide the digitally annotated training images into a training set and a test set; use the training set to train the fire analysis model, and use the test set to test the fire analysis model; preset an error threshold, and when the mean of the prediction errors of all training images in the test set is less than the error threshold, output the fire analysis model; the fire analysis model is a convolutional neural network model.
4. According to the artificial intelligence-based fire emergency broadcasting method of claim 3, It is characterized in that The method for determining whether to generate a fire occurrence instruction comprises: If the recognition result is normal, no fire occurrence instruction is generated; If the recognition result is a fire, a fire occurrence instruction is generated.
5. According to the artificial intelligence-based fire emergency broadcasting method of claim 4, It is characterized in that The method for obtaining m escape routes corresponding to each gathering area (6) comprises: Step 1: Randomly select a gathering area (6) as the initial area, and define the area where the escape exit is located as the target area; Step 2: Starting from the initial area, mark the initial area as a visited area, search the adjacent areas of the initial area, select one of the adjacent areas of the initial area as the next area, and mark the next area as a visited area; search the adjacent areas of the next area, select one of the adjacent areas of the next area that is not marked as a visited area as the new next area, and mark the new next area as a visited area; wherein the adjacent area to be searched is the evacuation area (7) and is not a fire area; Step 3: Loop step 2 until the adjacent areas of the next area are marked as visited areas, and then the loop ends. At this time, check whether the current area is the target area; If the current area is the target area, the depth-first search ends and the route found by the depth-first search is used as the escape route; If the current area is not the target area, the depth-first search ends, but the route found by the depth-first search is not used as an escape route. Step 4: loop step 2 to step 3, obtain m escape routes corresponding to the gathering area (6) of this loop, and mark the gathering area (6) of this loop as a completed area; Step 5: Repeat steps 1 to 4 to obtain m escape routes corresponding to each gathering area (6), and the gathering area (6) randomly selected in step 1 is the gathering area (6) that is not marked as a completed area.
6. According to claim 5, a fire emergency broadcast method based on artificial intelligence, It is characterized in that During the process of personnel escape, the detection parameters of the spreading area are collected and analyzed; if the spreading area generates a suspected fire instruction, that is, the spreading area is marked as a suspected area, then the evacuation area (7) adjacent to the spreading area marked as the suspected area is marked as the spreading area.
7. According to claim 6, a fire emergency broadcast method based on artificial intelligence, It is characterized in that The method for reacquiring the m escape routes corresponding to each gathering area (6) is different from the method for acquiring the m escape routes corresponding to each gathering area (6) in that, when searching for an adjacent area as the next area, the adjacent area is an evacuation area (7) and is neither a fire area nor a spreading area; the remaining steps in reacquiring the m escape routes corresponding to each gathering area (6) are consistent with the steps of acquiring the m escape routes corresponding to each gathering area (6).
8. An artificial intelligence-based fire emergency broadcast system, implementing an artificial intelligence-based fire emergency broadcast method according to any one of claims 1 to 7, It is characterized in that include: The area division module divides the building into n areas, where n is an integer greater than 1; Data acquisition module, collecting detection parameters of n areas; The fire judgment module calculates the judgment coefficients corresponding to the n areas based on the detection parameters, determines whether to generate a suspected fire instruction, and marks the area corresponding to the suspected fire instruction as a suspected area; An image acquisition module, which acquires regional images of suspected areas; The fire determination module analyzes the regional image to determine whether a fire occurrence instruction is generated, and marks the suspected area corresponding to the fire occurrence instruction as a fire area; The route planning module marks the area where the passage (4) in the building is located as an evacuation area (7), and marks the area in the building that is not marked as the evacuation area (7) as a personnel area (5); multiple adjacent personnel areas (5) without obstacles in between are regarded as a gathering area (6), and m escape routes corresponding to each gathering area (6) are obtained; A route selection module analyzes the m escape routes of each gathering area (6) and obtains the best escape route of each gathering area (6); The route broadcasting module broadcasts the best escape route for each gathering area (6) through the fire emergency broadcasting equipment.
