High-rise building-based fire spreading trend AI prediction method and system

Abstract

The embodiment of the invention relates to the technical field of fire spreading prediction, and particularly discloses a fire spreading trend AI prediction method and system based on a high-rise building. According to the embodiment of the invention, fire alarm information is received; acquiring smoke monitoring data of the target fire building, and performing smoke spreading prediction on the smoke monitoring data; selecting an auxiliary monitoring building; flame monitoring data transmitted by the auxiliary monitoring building are acquired, and flame spreading prediction is carried out; aI prediction of fire hazards and visual display of the fire spreading trend are carried out. Smoke monitoring data of a target fire building can be acquired, auxiliary monitoring analysis is performed, an auxiliary monitoring building is selected, flame monitoring data transmitted by the auxiliary monitoring building is acquired, and AI prediction of fire hazards and visual display of a fire spreading trend are performed by integrating a smoke prediction result and a flame prediction result. The defect of single data acquisition is overcome, and the accuracy of fire spreading trend prediction is effectively ensured in combination with flame monitoring of an auxiliary monitoring building.

Description

Technical Field

[0001] The present invention belongs to the technical field of fire spread prediction, and particularly relates to an AI prediction method and system for fire spread trend based on high-rise buildings. Background Art

[0002] Fire spread prediction is a process of pre-estimating and judging the development, diffusion, and evolution process of a fire in a specific environment using scientific methods, technical means, and mathematical models. Its core purpose is to predict the spread trend, speed, scope, and possible damage degree of a fire at different times and spaces by analyzing the initial conditions of the fire, environmental factors, etc., so as to provide a scientific basis for fire prevention and control, emergency response, and resource allocation.

[0003] In the prior art, when a fire occurs in a high-rise building, usually only the data of the building where the fire occurs can be obtained for fire spread analysis and trend prediction. The obtained data is single, and due to the damage caused by the fire to the building where the fire occurs, it is easy to cause data transmission delay, packet loss, or interruption, thus affecting the accuracy of fire spread trend prediction. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide an AI prediction method and system for fire spread trend based on high-rise buildings, aiming to solve the problems proposed in the background art.

[0005] To achieve the above purpose, the embodiments of the present invention provide the following technical solutions: An AI prediction method for fire spread trend based on high-rise buildings, the method specifically includes the following steps: Receive fire alarm information, determine the target fire building, and obtain the building layout data and building environment data of the target fire building; Obtain the smoke monitoring data of the target fire building, and based on the building layout data, perform smoke spread prediction on the smoke monitoring data to obtain a smoke prediction result; Perform auxiliary monitoring and analysis on the building environment data, and select auxiliary monitoring buildings; Obtain the flame monitoring data transmitted by the auxiliary monitoring buildings, and based on the building layout data, perform flame spread prediction on the flame monitoring data to obtain a flame prediction result; Integrate the smoke prediction result and the flame prediction result, perform AI prediction of fire danger, and based on the building layout data, perform visual display of the fire spread trend.

[0006] An AI prediction system for fire spread trend based on high-rise buildings, the system includes a fire alarm processing unit, a smoke spread prediction unit, an auxiliary building selection unit, a flame spread prediction unit, and a danger prediction display unit, wherein: A fire alarm processing unit, configured to receive fire alarm information, determine a target fire building, and obtain building layout data and building environment data of the target fire building; A smoke spread prediction unit, configured to obtain smoke monitoring data of the target fire building, and based on the building layout data, perform smoke spread prediction on the smoke monitoring data to obtain a smoke prediction result; An auxiliary building selection unit, configured to perform auxiliary monitoring and analysis on the building environment data to select an auxiliary monitoring building; A flame spread prediction unit, configured to obtain flame monitoring data transmitted by the auxiliary monitoring building, and based on the building layout data, perform flame spread prediction on the flame monitoring data to obtain a flame prediction result; A hazard prediction display unit, configured to comprehensively analyze the smoke prediction result and the flame prediction result, perform AI prediction of fire hazards, and based on the building layout data, perform visual display of the fire spread trend.

[0007] Compared with the prior art, the beneficial effects of the present invention are: In the embodiment of the present invention, by receiving fire alarm information; obtaining smoke monitoring data of a target fire building, performing smoke spread prediction on the smoke monitoring data; selecting an auxiliary monitoring building; obtaining flame monitoring data transmitted by the auxiliary monitoring building, performing flame spread prediction; performing AI prediction of fire hazards and visual display of the fire spread trend. It is possible to obtain smoke monitoring data of the target fire building, and perform auxiliary monitoring and analysis, select an auxiliary monitoring building, obtain flame monitoring data transmitted by the auxiliary monitoring building, and then comprehensively analyze the smoke prediction result and the flame prediction result, perform AI prediction of fire hazards and visual display of the fire spread trend, solve the defect of single data acquisition, and combine the flame monitoring of the auxiliary monitoring building to effectively ensure the accuracy of the fire spread trend prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings in the following descriptions are only some embodiments of the present invention.

[0009] Figure 1 Shows a flowchart of the method provided by the embodiment of the present invention.

[0010] Figure 2 Shows an application architecture diagram of the system provided by the embodiment of the present invention.

[0011] Figure 3 Shows a structural block diagram of the fire alarm processing unit in the system provided by the embodiment of the present invention.

[0012] Figure 4 The structural block diagram of the smoke spread prediction unit in the system provided by the embodiment of the present invention is shown.

[0013] Figure 5 The structural block diagram of the auxiliary building selection unit in the system provided by the embodiment of the present invention is shown. Detailed implementation manners

[0014] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0015] It can be understood that in the prior art, when a fire occurs in a high-rise building, usually only the data of the building where the fire occurs can be obtained for fire spread analysis and trend prediction. The obtained data is single, and due to the damage caused by the fire to the building where the fire occurs, it is easy to cause data transmission delay, packet loss or interruption, thereby affecting the accuracy of fire spread trend prediction.

[0016] To solve the above problems, the embodiment of the present invention receives fire alarm information, determines the target fire building, and obtains the building layout data and building environment data of the target fire building; obtains the smoke monitoring data of the target fire building, and based on the building layout data, predicts the smoke spread of the smoke monitoring data to obtain a smoke prediction result; conducts auxiliary monitoring analysis on the building environment data to select an auxiliary monitoring building; obtains the flame monitoring data transmitted by the auxiliary monitoring building, and based on the building layout data, predicts the flame spread of the flame monitoring data to obtain a flame prediction result; combines the smoke prediction result and the flame prediction result to conduct AI prediction of fire danger, and based on the building layout data, conducts visual display of the fire spread trend. It can obtain the smoke monitoring data of the target fire building, conduct auxiliary monitoring analysis, select an auxiliary monitoring building, obtain the flame monitoring data transmitted by the auxiliary monitoring building, and then combine the smoke prediction result and the flame prediction result to conduct AI prediction of fire danger and visual display of the fire spread trend, solving the defect of single data acquisition, and effectively ensuring the accuracy of fire spread trend prediction in combination with the flame monitoring of the auxiliary monitoring building.

[0017] Figure 1 The flowchart of the method provided by the embodiment of the present invention is shown.

[0018] Specifically, based on the AI prediction method for the fire spread trend of high-rise buildings, the method specifically includes the following steps: Step S101, receive fire alarm information, determine the target fire building, and obtain the building layout data and building environment data of the target fire building.

[0019] In an embodiment of the present invention, by receiving a fire alarm message, performing target recognition on the fire alarm message to determine the target fire building where the fire occurs, then obtaining the building layout data and building positioning data of the target fire building from a preset building backup database, and based on the building positioning data, performing positioning recognition on the target fire building in a preset electronic map, and then taking the positioning of the target fire building as the origin, performing environmental radiation analysis according to a preset environmental radius to obtain building environmental data.

[0020] Specifically, a flowchart of fire alarm message processing in the method provided by the embodiment of the present invention is shown.

[0021] Among them, in a preferred embodiment provided by the present invention, the steps of receiving the fire alarm message, determining the target fire building, and obtaining the building layout data and building environmental data of the target fire building specifically include the following steps: Receive a fire alarm message; Perform target recognition on the fire alarm message to determine the target fire building; Obtain the building layout data and building positioning data of the target fire building; Perform positioning recognition and environmental analysis according to the building positioning data to obtain building environmental data.

[0022] Among them, in a preferred embodiment provided by the present invention, the steps of performing positioning recognition and environmental analysis according to the building positioning data to obtain building environmental data specifically include the following steps: Obtain the BIM model of the target fire building, the building heat transfer coefficient, and the meteorological data of the target location according to the building positioning data. The meteorological data of the target location includes wind speed, temperature, and humidity data; Obtain the height of the target fire building from the BIM model of the target fire building, and perform multi-parameter fusion calculation on the wind speed, temperature, and humidity data and the height of the target fire building to obtain an initial adjustment parameter; Weight the initial adjustment parameter by the building heat transfer coefficient to obtain a temperature sensitivity coefficient, a time decay coefficient, and an environmental correction coefficient respectively; According to the BIM model of the target fire building, perform three-dimensional grid division on the internal space of the building, calculate the coordinate difference value of each grid point relative to the fire starting point, and mark the key areas in the BIM model of the target fire building to obtain a three-dimensional gradient field distribution map with marked spatial topological features. The key areas include the vertical channels of the longitudinal stairwells, the horizontal extensions of the transverse ventilation ducts, and the spatial blockages at the positions of the fire doors. Among them, the gradient value increases in the vertical channel part of the longitudinal stairwell, the gradient value decreases in the horizontal extension part of the transverse ventilation duct, and the gradient value is zeroed in the spatial blockage part at the position of the fire door; Obtain the neighboring buildings of the target fire building, and calculate the weight matrix of the threat value of the neighboring buildings according to the fire resistance rating of the materials of the neighboring buildings and the distance between the neighboring buildings of the target fire building; Construct a CFD benchmark model, and in the CFD benchmark model, adjust the air velocity components in each direction according to the weight matrix of the threat value of the neighboring buildings to obtain an air flow field model containing building structure characteristics; Input the environmental correction coefficient into the CFD benchmark model, use the temperature sensitivity coefficient to adjust the scaling ratio of the environmental correction coefficient, so that the air velocity automatically scales with the change of temperature and humidity, and forcibly set the velocity to zero boundary at the position of the fire door marked by the three-dimensional gradient field to obtain an optimized air flow field model containing building structure characteristics and real-time environmental parameters; Perform a spatial integration operation on the gradient field in the three-dimensional gradient field distribution map marked with spatial topological characteristics to quantify the diffusion potential energy of each area inside the building and obtain potential energy values; Obtain the air flow velocity data in the optimized air flow field model containing building structure characteristics and real-time environmental parameters, and use the air flow velocity data as the kinetic energy correction term; Superimpose the time decay coefficient on the potential energy value, and then calibrate it with the kinetic energy correction term to obtain a three-dimensional fire spread energy distribution cloud map with a timestamp; among them, the optimized air flow field model containing building structure characteristics and real-time environmental parameters, the weight matrix of the threat value of the neighboring buildings, and the three-dimensional fire spread energy distribution cloud map with a timestamp constitute the building environment dataset.

[0023] In the embodiment of the present invention, by converting meteorological data into dynamic coefficients, all calculation modules are affected by the dynamic coefficients. For example, windy weather will automatically increase the β value to accelerate the time response of the model. And, by constructing a BIM model and calculating the three-dimensional gradient field, and performing flow field constraints, the building structure characteristics are converted into physical constraints through the gradient field to prevent unreasonable diffusion such as the simulation results passing through load-bearing walls. Obtain the threat weight through building material data, and perform flow field correction based on this. The material characteristics of neighboring buildings change the operation accuracy distribution of the CFD model through the weight matrix. The final spatio-temporal energy cloud map can be decomposed into two display modes: a time-axis animation and a sectional heat map, supporting multi-dimensional decision analysis. The dynamic coefficient realizes minute-level environmental adaptation, the three-dimensional gradient field breaks through the limitation of two-dimensional plane analysis, and the threat weight matrix assigns different monitoring priorities to neighboring buildings.

[0024] Further, the AI prediction method for the fire spread trend of high-rise buildings further includes the following steps: Step S102, obtain the smoke monitoring data of the target fire building, and based on the building layout data, perform smoke spread prediction on the smoke monitoring data to obtain a smoke prediction result.

[0025] In an embodiment of the present invention, by obtaining the smoke monitoring data directly transmitted by the target fire building, performing time series division and period extraction on the smoke monitoring data to obtain smoke time series data, and then performing actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds in the target fire building, and further based on the building layout data, performing smoke spread prediction according to the multiple smoke spread directions and corresponding smoke spread speeds to obtain a smoke prediction result.

[0026] Specifically, a flowchart of performing smoke spread prediction in the method provided by the embodiment of the present invention is shown.

[0027] Among them, in a preferred embodiment provided by the present invention, the obtaining the smoke monitoring data of the target fire building, and based on the building layout data, performing smoke spread prediction on the smoke monitoring data to obtain a smoke prediction result specifically includes the following steps: Obtain the smoke monitoring data of the target fire building; Perform time series division and period extraction on the smoke monitoring data to obtain smoke time series data; Perform actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds; Based on the building layout data, perform smoke spread prediction according to the multiple smoke spread directions and multiple smoke spread speeds to obtain a smoke prediction result.

[0028] Among them, in a preferred embodiment provided by the present invention, the performing actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds specifically includes the following steps: Obtain the historical working day logs of smoke sensors in the target fire building, and obtain the dynamic weight coefficients of each sensor according to the number of false alarms and time intervals of the sensors in the historical working day logs; Obtain smoke time series data, and calculate the concentration change rate of the smoke time series data to obtain a concentration change rate curve graph with time marks; Decompose the three-dimensional gradient field distribution graph marked with spatial topological features into gradient field components in six directions: east, south, west, north, up, and down; According to the gradient field components decomposed in six directions and the concentration change rate curve graph with time marks, perform spatial superposition operation on the concentration change rate in each direction, and strengthen it in the vertical channel of the longitudinal staircase well and suppress it in the horizontal extension direction of the transverse ventilation duct to obtain the spatial superposition operation result of the concentration change rate of each sensor; Perform weighted average on the spatial superposition operation result of the concentration change rate of each sensor by using the dynamic weight coefficients of each sensor to obtain a six-direction smoke diffusion intensity matrix; The environmental threat direction correction coefficient is calculated using the weight matrix of the threat value of adjacent buildings and real-time meteorological data, and the six-direction smoke diffusion intensity matrix is adjusted using the environmental threat direction correction coefficient to obtain a smoke spread speed prediction table with environmental correction, and the smoke prediction result is obtained according to the smoke spread speed prediction table with environmental correction.

[0029] In the embodiment of the present invention, the data credibility is dynamically adjusted through the historical false alarm rate to avoid misjudgment caused by equipment failures. And the building structure characteristics are combined with real-time smoke data. For example: when a sharp increase in concentration is detected in the corridor on the third floor, the gradient characteristics of the ventilation ducts in this area are automatically associated, and it is predicted that the smoke will quickly spread along the ducts to the fifth floor, effectively adapting to the delay prediction in different environments. And, to ensure better results, environmental threats and meteorological data are introduced to correct the smoke spread speed in real time. For example: under the meteorological condition of southeast wind, when the adjacent supermarket is of wooden structure, the system increases the predicted value of the eastward spread speed by 58% and generates an instruction to evacuate the east parking lot first.

[0030] Further, the AI prediction method for the fire spread trend based on high-rise buildings further includes the following steps: Step S103, perform auxiliary monitoring and analysis on the building environment data to select an auxiliary monitoring building.

[0031] In the embodiment of the present invention, through building proximity recognition of the building environment data, multiple environmental adjacent buildings adjacent to the target fire building are determined, and through occurrence recognition of the fire alarm information, the occurrence floor and occurrence position of the fire in the target fire building are determined. Then, according to the occurrence floor and occurrence position, the auxiliary monitoring angle is determined, and then, according to the auxiliary monitoring angle, multiple environmental adjacent buildings are subjected to auxiliary monitoring matching, and an auxiliary monitoring building is selected from multiple environmental adjacent buildings.

[0032] Specifically, a flowchart showing the selection of an auxiliary monitoring building in the method provided by the embodiment of the present invention is shown.

[0033] Among them, in the preferred embodiment provided by the present invention, the performing auxiliary monitoring and analysis on the building environment data to select an auxiliary monitoring building specifically includes the following steps: Perform building proximity recognition on the building environment data to determine multiple environmental adjacent buildings; Perform occurrence recognition on the fire alarm information to determine the occurrence floor and occurrence position; Determine the auxiliary monitoring angle according to the occurrence floor and the occurrence position; Perform auxiliary monitoring matching on multiple environmental adjacent buildings according to the auxiliary monitoring angle, and select an auxiliary monitoring building.

[0034] Among them, in the preferred embodiment provided by the present invention, determining the auxiliary monitoring angle according to the occurrence floor and the occurrence position specifically includes the following steps: The absolute height of the floor where the fire starts extracted from the BIM model of the target fire building, and the horizontal projection distance between the target fire building and the adjacent building is obtained according to the geographic information system; The basic monitoring angle is calculated based on the absolute height of the floor where the fire starts and the horizontal projection distance of the adjacent building; Obtain the fire resistance rating of the adjacent building, calculate the threat values of all buildings based on the horizontal projection distance between the target fire building and the adjacent building, the fire resistance rating of the adjacent building, and the current wind direction angle, and normalize the threat values of all buildings to obtain the normalized threat weight; Obtain the azimuth angle of the adjacent building and calculate the matching degree between the azimuth angle of the adjacent building and the wind direction; Adjust the basic monitoring angle by using the matching degree between the azimuth angle of the adjacent building and the wind direction, and the normalized threat weight as correction terms to obtain the auxiliary monitoring angle.

[0035] In the embodiment of the present invention, through the dynamic coupling of building height and horizontal distance, the problem of vertical blind areas in traditional two-dimensional plane monitoring is solved. And the fire resistance rating of building materials is converted into weights, which can automatically prioritize the monitoring of wooden buildings. And the embedded azimuth matching enables the monitoring system to automatically deflect when the threat increases suddenly in the downwind direction.

[0036] Furthermore, the AI prediction method for the fire spread trend based on high-rise buildings further includes the following steps: Step S104, obtain the flame monitoring data transmitted by the auxiliary monitoring building, and based on the building layout data, perform flame spread prediction on the flame monitoring data to obtain a flame prediction result.

[0037] In the embodiment of the present invention, an auxiliary communication connection with the auxiliary monitoring building is established, and the auxiliary monitoring building is controlled for auxiliary monitoring to obtain the flame monitoring data transmitted by the auxiliary monitoring building. By performing time series division and period extraction on the flame monitoring data, flame time series data is obtained, and then actual spread analysis is performed on the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds of the fire in the target fire building. Furthermore, based on the building layout data, according to multiple flame spread directions and multiple flame spread speeds, flame spread prediction is performed to obtain a flame prediction result.

[0038] Specifically, the flowchart of performing flame spread prediction in the method provided by the embodiment of the present invention is shown.

[0039] Among them, in the preferred embodiment provided by the present invention, the steps of obtaining the flame monitoring data transmitted by the auxiliary monitoring building, predicting the flame spread based on the building layout data, and obtaining the flame prediction result specifically include the following steps: Obtain the flame monitoring data transmitted by the auxiliary monitoring building; Perform time series division and period extraction on the flame monitoring data to obtain flame time series data; Perform actual spread analysis on the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds; Based on the building layout data, perform flame spread prediction according to the multiple flame spread directions and multiple flame spread speeds to obtain the flame prediction result.

[0040] Among them, in the preferred embodiment provided by the present invention, the steps of performing actual spread analysis on the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds specifically include the following steps: Perform edge detection on the flame time series data, extract the flame front contour line, and calculate the vertical distance between the highest point of the flame and the building reference plane in each frame of the image to obtain a time-stamped flame height sequence; Filter the time-stamped flame height sequence by the sliding window average method to obtain a smoothed flame height curve; Perform first derivative calculation on the smoothed flame height curve to obtain the uncorrected vertical spread rate; Calculate the horizontal projection distance of the flame front according to the auxiliary monitoring angle and the real-time wind speed, and then perform differential operation on the horizontal projection distance of the flame front to obtain the uncorrected horizontal spread speed; Identify the gradient field characteristics of the area where the flame is currently located. If it is within the range of the stairwell, strengthen and correct the uncorrected vertical spread rate to obtain the structurally corrected vertical rate; Select the high-risk neighboring building orientations according to the threat weight matrix, and apply a diffusion resistance coefficient in the corresponding direction to obtain the azimuth suppression coefficient; Calculate the vector components in the four directions of east, south, west, and north according to the uncorrected horizontal spread speed and the azimuth suppression coefficient; calculate the vector components in the up and down directions according to the structurally corrected vertical rate to obtain a three-dimensional flame spread speed matrix; Compare the coordinates of the high-risk areas in the smoke prediction result, mark the overlapping areas with the smoke coverage area in the three-dimensional flame spread speed matrix, apply a chain reaction to the overlapping areas to obtain a fire spread heat map, and obtain the flame prediction result according to the fire spread heat map.

[0041] In the embodiments of the present invention, the present invention incorporates four types of data, namely infrared thermal imaging, building gradient field, adjacent threat matrix, and real-time weather, into a unified calculation framework, effectively improving the accuracy of prediction. Moreover, through differential processing of vertical enhancement and horizontal suppression, it accurately reflects the influence of building structures on the direction of fire spread. Additionally, through cross-verification of the smoke-flame coverage area, special risk points such as areas where flammable substances gather and electrical equipment rooms can be identified in advance.

[0042] Further, the AI prediction method for the fire spread trend based on high-rise buildings further includes the following steps: Step S105: Integrate the smoke prediction result and the flame prediction result to perform AI prediction of fire danger, and based on the building layout data, perform visual display of the fire spread trend.

[0043] In the embodiments of the present invention, the smoke prediction result and the flame prediction result are integrated to perform AI prediction of fire danger for the target fire building, obtain the spread trend data, and based on the building layout data, create a two-dimensional or three-dimensional visual background. Then, color matching is performed on the spread trend data, and the corresponding color matching results are recorded in real time according to the spread speed and the degree of disaster. Furthermore, according to the color matching results, the visual display of the fire spread trend of the target fire building is performed in the visual background.

[0044] Specifically, a flowchart of performing AI prediction of fire danger in the method provided by the embodiments of the present invention is shown.

[0045] Among them, in the preferred embodiment provided by the present invention, the step of integrating the smoke prediction result and the flame prediction result to perform AI prediction of fire danger and performing visual display of the fire spread trend based on the building layout data specifically includes the following steps: Integrate the smoke prediction result and the flame prediction result to perform AI prediction of fire danger and obtain the spread trend data; Create a visual background based on the building layout data; Perform color matching on the spread trend data and record the color matching results; According to the color matching results, perform visual display of the fire spread trend in the visual background.

[0046] Further, Figure 2 A diagram of the application architecture of the system provided by the embodiments of the present invention is shown.

[0047] Among them, in another preferred embodiment provided by the present invention, the AI prediction system for the fire spread trend based on high-rise buildings includes: The fire alarm processing unit 101 is used to receive fire alarm information, determine the target fire building, and obtain the building layout data and building environment data of the target fire building.

[0048] In the embodiment of the present invention, the fire alarm processing unit 101 receives fire alarm information, conducts target recognition on the fire alarm information to determine the target fire building where the fire occurs, then obtains the building layout data and building positioning data of the target fire building from the preset building backup database, and based on the building positioning data, conducts positioning recognition on the target fire building in the preset electronic map, and then takes the positioning of the target fire building as the origin, and conducts environmental radiation analysis according to the preset environmental radius to obtain building environment data.

[0049] Specifically, Figure 3 The structural block diagram of the fire alarm processing unit 101 in the system provided by the embodiment of the present invention is shown.

[0050] Among them, in the preferred embodiment provided by the present invention, the fire alarm processing unit 101 specifically includes: The alarm information receiving module 1011 is used to receive fire alarm information; The fire target recognition module 1012 is used to conduct target recognition on the fire alarm information to determine the target fire building; The building data acquisition module 1013 is used to obtain the building layout data and building positioning data of the target fire building; The positioning recognition and environment analysis module 1014 is used to conduct positioning recognition and environment analysis according to the building positioning data to obtain building environment data.

[0051] Furthermore, the AI prediction system for the fire spread trend based on high-rise buildings further includes: The smoke spread prediction unit 102 is used to obtain the smoke monitoring data of the target fire building, and based on the building layout data, conduct smoke spread prediction on the smoke monitoring data to obtain a smoke prediction result.

[0052] In the embodiment of the present invention, the smoke spread prediction unit 102 obtains the smoke monitoring data directly transmitted by the target fire building, conducts time series division and period extraction on the smoke monitoring data to obtain smoke time series data, then conducts actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds of the fire smoke in the target fire building, and further conducts smoke spread prediction based on the building layout data according to the multiple smoke spread directions and corresponding smoke spread speeds to obtain a smoke prediction result.

[0053] Specifically, Figure 4The block diagram of the smoke spread prediction unit 102 in the system provided by the embodiment of the present invention is shown.

[0054] Among them, in the preferred embodiment provided by the present invention, the smoke spread prediction unit 102 specifically includes: A smoke monitoring data acquisition module 1021, configured to acquire the smoke monitoring data of the target fire building; A time series division and period extraction module 1022, configured to perform time series division and period extraction on the smoke monitoring data to obtain smoke time series data; An actual spread analysis module 1023, configured to perform an actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds; A smoke spread prediction module 1024, configured to perform smoke spread prediction based on the building layout data according to the multiple smoke spread directions and the multiple smoke spread speeds to obtain a smoke prediction result.

[0055] Furthermore, the AI prediction system for the fire spread trend of high-rise buildings further includes: An auxiliary building selection unit 103, configured to perform auxiliary monitoring analysis on the building environment data and select an auxiliary monitoring building.

[0056] In the embodiment of the present invention, the auxiliary building selection unit 103 determines multiple environmentally adjacent buildings adjacent to the target fire building by performing building proximity recognition on the building environment data, and performs occurrence recognition on the fire alarm information to determine the occurrence floor and occurrence location of the fire in the target fire building. Then, according to the occurrence floor and occurrence location, an auxiliary monitoring angle is determined. Furthermore, according to the auxiliary monitoring angle, auxiliary monitoring matching is performed on the multiple environmentally adjacent buildings, and an auxiliary monitoring building is selected from the multiple environmentally adjacent buildings.

[0057] Specifically, Figure 5 The block diagram of the auxiliary building selection unit 103 in the system provided by the embodiment of the present invention is shown.

[0058] Among them, in the preferred embodiment provided by the present invention, the auxiliary building selection unit 103 specifically includes: A building proximity recognition module 1031, configured to perform building proximity recognition on the building environment data to determine multiple environmentally adjacent buildings; An occurrence recognition module 1032, configured to perform occurrence recognition on the fire alarm information to determine the occurrence floor and occurrence location; An auxiliary monitoring angle determination module 1033, configured to determine an auxiliary monitoring angle according to the occurrence floor and the occurrence location; The auxiliary monitoring matching module 1034 is configured to perform auxiliary monitoring matching on multiple adjacent buildings in the environment according to the auxiliary monitoring angle, and select an auxiliary monitoring building.

[0059] Further, the AI prediction system for the fire spread trend based on high-rise buildings further includes: The flame spread prediction unit 104 is configured to obtain the flame monitoring data transmitted by the auxiliary monitoring building, and perform flame spread prediction on the flame monitoring data based on the building layout data to obtain a flame prediction result.

[0060] In an embodiment of the present invention, the flame spread prediction unit 104 establishes an auxiliary communication connection with the auxiliary monitoring building, and performs auxiliary monitoring control on the auxiliary monitoring building to obtain the flame monitoring data transmitted by the auxiliary monitoring building. By performing time series division and period extraction on the flame monitoring data, flame time series data is obtained, and then actual spread analysis is performed on the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds of the fire in the target fire building. Furthermore, based on the building layout data, flame spread prediction is performed according to multiple flame spread directions and multiple flame spread speeds to obtain a flame prediction result.

[0061] The danger prediction and display unit 105 is configured to perform AI prediction of fire danger by integrating the smoke prediction result and the flame prediction result, and perform visual display of the fire spread trend based on the building layout data.

[0062] In an embodiment of the present invention, the danger prediction and display unit 105 integrates the smoke prediction result and the flame prediction result to perform AI prediction of fire danger on the target fire building to obtain spread trend data. And based on the building layout data, a two-dimensional or three-dimensional visual background is created, and then color matching is performed on the spread trend data. According to the spread speed and the degree of disaster, the corresponding color matching results are recorded in real time. Furthermore, according to the color matching results, the fire spread trend of the target fire building is visually displayed in the visual background.

[0063] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0064] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0065] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0066] The above embodiments only represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

[0067] The above is only the preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An AI prediction method for the fire spread trend based on high-rise buildings, characterized in that, The method specifically includes the following steps: Receive a fire alarm message, determine the target fire building, and obtain the building layout data and building environment data of the target fire building; Obtain the smoke monitoring data of the target fire building, and based on the building layout data, predict the smoke spread of the smoke monitoring data to obtain a smoke prediction result; Conduct auxiliary monitoring and analysis on the building environment data, and select an auxiliary monitoring building; Obtain the flame monitoring data transmitted by the auxiliary monitoring building, and based on the building layout data, predict the flame spread of the flame monitoring data to obtain a flame prediction result; Integrate the smoke prediction result and the flame prediction result, conduct an AI prediction of the fire hazard, and based on the building layout data, conduct a visual display of the fire spread trend; Among them, the steps of receiving a fire alarm message, determining the target fire building, and obtaining the building layout data and building environment data of the target fire building specifically include the following steps: Receive a fire alarm message; Conduct target recognition on the fire alarm message to determine the target fire building; Obtain the building layout data and building positioning data of the target fire building; Conduct positioning recognition and environmental analysis according to the building positioning data to obtain building environment data.

2. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 1, wherein The steps of conducting positioning recognition and environmental analysis according to the building positioning data to obtain building environment data specifically include the following steps: Obtain the BIM model of the target fire building, the building heat transfer coefficient, and the meteorological data of the target location according to the building positioning data. The meteorological data of the target location includes wind speed, temperature, and humidity data; Obtain the height of the target fire building from the BIM model of the target fire building, and perform multi-parameter fusion calculation on the wind speed, temperature, and humidity data and the height of the target fire building to obtain an initial adjustment parameter; Weight the initial adjustment parameter through the building heat transfer coefficient to obtain a temperature sensitivity coefficient, a time decay coefficient, and an environment correction coefficient respectively; According to the BIM model of the target fire building, conduct three-dimensional grid division on the internal space of the building, calculate the coordinate difference value of each grid point relative to the fire starting point, and mark the key areas in the BIM model of the target fire building to obtain a three-dimensional gradient field distribution map with marked spatial topological features. The key areas include the vertical channels of the longitudinal stairwells, the horizontal extensions of the transverse ventilation ducts, and the spatial blockages at the positions of the fire doors; among them, the gradient value of the vertical channel part of the longitudinal stairwell increases, the gradient value of the horizontal extension part of the transverse ventilation duct decreases, and the gradient value of the spatial blockage part at the position of the fire door is zero; Obtain the adjacent buildings of the target fire building, and calculate the weight matrix of the adjacent building threat value according to the fire resistance rating of the adjacent building materials and the distance between the adjacent buildings of the target fire building; Construct a CFD benchmark model, and in the CFD benchmark model, adjust the air velocity components in each direction according to the weight matrix of the adjacent building threat value to obtain an air flow field model including building structure characteristics; Input the environmental correction coefficient into the CFD benchmark model, use the temperature sensitivity coefficient to adjust the scaling ratio of the environmental correction coefficient, so that the air flow velocity automatically scales with the change of temperature and humidity, and forcefully set the velocity zero boundary at the position of the fire door marked by the three-dimensional gradient field, to obtain an optimized air flow field model containing building structure characteristics and real-time environmental parameters; Perform a spatial integration operation on the gradient field in the three-dimensional gradient field distribution map marking the spatial topological characteristics, quantify the diffusion potential energy of each region inside the building, and obtain the potential energy value; Obtain the air flow velocity data in the optimized air flow field model containing building structure characteristics and real-time environmental parameters, and use the air flow velocity data as the kinetic energy correction term; Superimpose the time decay coefficient on the potential energy value, and then calibrate it using the kinetic energy correction term to obtain a three-dimensional fire spread energy distribution cloud map with a time stamp; among them, the optimized air flow field model containing building structure characteristics and real-time environmental parameters, the weight matrix of the neighboring building threat value, and the three-dimensional fire spread energy distribution cloud map with a time stamp constitute the building environment data set.

3. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 2, wherein The obtaining of the smoke monitoring data of the target fire building and the performing of smoke spread prediction on the smoke monitoring data based on the building layout data to obtain the smoke prediction result specifically include the following steps: Obtain the smoke monitoring data of the target fire building; Perform time series division and period extraction on the smoke monitoring data to obtain smoke time series data; Perform actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds; Based on the building layout data, perform smoke spread prediction according to the multiple smoke spread directions and multiple smoke spread speeds to obtain the smoke prediction result.

4. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 3, wherein, The performing of actual spread analysis on the smoke time series data to determine multiple smoke spread directions and corresponding smoke spread speeds specifically includes the following steps: Obtain the historical working day logs of the smoke sensors in the target fire building, and obtain the dynamic weight coefficients of each sensor according to the number of false alarms and time intervals of the sensors in the historical working day logs; Obtain the smoke time series data, and calculate the concentration change rate of the smoke time series data to obtain a concentration change rate curve graph with time marks; Decompose the gradient field components of the three-dimensional gradient field distribution map marking the spatial topological characteristics in six directions: east, south, west, north, up, and down; According to the gradient field components decomposed in the six directions and the concentration change rate curve graph with time marks, perform a spatial superposition operation on the concentration change rate in each direction, and strengthen the vertical channel of the longitudinal stairwell and suppress the horizontal extension direction of the horizontal ventilation duct to obtain the spatial superposition operation result of the concentration change rate of each sensor; Perform weighted average on the spatial superposition operation result of the concentration change rate of each sensor using the dynamic weight coefficients of each sensor to obtain a six-direction smoke diffusion intensity matrix; Calculate the environmental threat direction correction coefficient using the weight matrix of the threat values of adjacent buildings and real-time meteorological data, and use the environmental threat direction correction coefficient to adjust the six-direction smoke diffusion intensity matrix to obtain a smoke spread speed prediction table with environmental correction, and obtain the smoke prediction result according to the smoke spread speed prediction table with environmental correction.

5. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 4, wherein, The auxiliary monitoring and analysis of the building environmental data and the selection of the auxiliary monitoring building specifically include the following steps: Perform building proximity recognition on the building environmental data to determine multiple environmental adjacent buildings; Perform occurrence recognition on the fire alarm information to determine the occurrence floor and occurrence location; Determine the auxiliary monitoring angle according to the occurrence floor and the occurrence location; Perform auxiliary monitoring matching on multiple environmental adjacent buildings according to the auxiliary monitoring angle, and select the auxiliary monitoring building.

6. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 5, wherein, The determination of the auxiliary monitoring angle according to the occurrence floor and the occurrence location specifically includes the following steps: Extract the absolute height of the fire floor from the BIM model of the target fire building, and obtain the horizontal projection distance between the target fire building and the adjacent building according to the geographic information system; Calculate the basic monitoring angle according to the absolute height of the fire floor and the horizontal projection distance of the adjacent building; Obtain the fire resistance rating of the adjacent building, calculate the threat values of all buildings according to the horizontal projection distance between the target fire building and the adjacent building, the fire resistance rating of the adjacent building, and the current wind direction angle, and normalize the threat values of all buildings to obtain the normalized threat weight; Obtain the azimuth angle of the adjacent building and calculate the matching degree between the azimuth angle of the adjacent building and the wind direction; Use the matching degree between the azimuth angle of the adjacent building and the wind direction and the normalized threat weight as correction terms to adjust the basic monitoring angle to obtain the auxiliary monitoring angle.

7. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 6, wherein The acquisition of the flame monitoring data transmitted by the auxiliary monitoring building and the prediction of the flame spread based on the building layout data to obtain the flame prediction result specifically include the following steps: Obtain the flame monitoring data transmitted by the auxiliary monitoring building; Perform time series division and period extraction on the flame monitoring data to obtain the flame time series data; Perform actual spread analysis on the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds; Based on the building layout data, perform flame spread prediction according to multiple flame spread directions and multiple flame spread speeds to obtain the flame prediction result.

8. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 7, wherein The actual spread analysis of the flame time series data to determine multiple flame spread directions and corresponding flame spread speeds specifically includes the following steps: Perform edge detection on the flame time series data, extract the flame front contour line, and calculate the vertical distance between the highest point of the flame and the building reference plane in each frame of the image to obtain a time-stamped flame height sequence; Filter the time-stamped flame height sequence by the sliding window average method to obtain a smoothed flame height curve; Perform the first derivative calculation on the smoothed flame height curve to obtain the uncorrected vertical spread rate; Calculate the horizontal projection distance of the flame front according to the auxiliary monitoring angle and the real-time wind speed, and then perform a differential operation on the horizontal projection distance of the flame front to obtain the uncorrected horizontal spread speed; Identify the gradient field characteristics of the area where the flame is currently located. If it is within the range of the stairwell, strengthen the correction of the uncorrected vertical spread rate to obtain the vertically corrected rate after structure correction; Select the azimuth of high-risk neighboring buildings according to the threat weight matrix, and apply a diffusion resistance coefficient in the corresponding direction to obtain the azimuth inhibition coefficient; Calculate the vector components in the four directions of east, south, west, and north according to the uncorrected horizontal spread speed and the azimuth inhibition coefficient; calculate the vector components in the up and down directions according to the vertically corrected rate after structure correction to obtain a three-dimensional flame spread speed matrix; Compare the coordinates of the high-risk areas in the smoke prediction results, and mark the overlapping areas with the smoke coverage area in the three-dimensional flame spread speed matrix, apply a chain reaction to the overlapping areas to obtain a fire spread heat map, and obtain the flame prediction result according to the fire spread heat map.

9. The AI prediction method for the fire spread trend based on high-rise buildings according to claim 8, characterized in that Based on the smoke prediction result and the flame prediction result, perform AI prediction of fire danger, and perform visual display of the fire spread trend based on the building layout data, which specifically includes the following steps: Based on the smoke prediction result and the flame prediction result, perform AI prediction of fire danger to obtain spread trend data; Create a visual background based on the building layout data; Perform color matching on the spread trend data and record the color matching result; According to the color matching result, perform visual display of the fire spread trend in the visual background.

10. An AI prediction system for the fire spread trend of high-rise buildings, which is applied to the AI prediction method for the fire spread trend of high-rise buildings according to any one of claims 1 to 9, characterized in that, The system includes a fire alarm processing unit, a smoke spread prediction unit, an auxiliary building selection unit, a flame spread prediction unit, and a danger prediction display unit, where: The fire alarm processing unit is used to receive fire alarm information, determine the target fire building, and obtain the building layout data and building environment data of the target fire building; The smoke spread prediction unit is used to obtain the smoke monitoring data of the target fire building, and perform smoke spread prediction on the smoke monitoring data based on the building layout data to obtain a smoke prediction result; The auxiliary building selection unit is used to perform auxiliary monitoring analysis on the building environment data and select an auxiliary monitoring building; The flame spread prediction unit is used to obtain the flame monitoring data transmitted by the auxiliary monitoring building, and perform flame spread prediction on the flame monitoring data based on the building layout data to obtain a flame prediction result; The danger prediction display unit is used to comprehensively analyze the smoke prediction result and the flame prediction result, perform AI prediction of fire danger, and perform visual display of the fire spread trend based on the building layout data.

Images