Fire-fighting emergency early warning processing and fire extinguishing integrated control system based on fireproof door

Abstract

The invention relates to the technical field of fire-fighting early warning, in particular to a fire-fighting emergency early-warning processing and fire-extinguishing integrated control system based on a fireproof door, which comprises a fire source positioning unit used for acquiring a temperature field, a smoke concentration field and gas components of each fireproof area in a target building structure, and sending the temperature field, the smoke concentration field and the gas components to a fire-fighting emergency early-warning processing and fire-extinguishing integrated control unit; determining a fire source center coordinate of the fireproof area according to the temperature field; and the fire behavior determination unit is used for determining a fire behavior intensity field of the fireproof area according to the temperature field, the smoke concentration field and the gas components. According to the invention, the temperature field, the smoke concentration field and the gas component of each fireproof area of the target building are obtained through the fire source positioning unit, the center coordinate of the fire source can be accurately positioned, the intensity of the fire behavior is judged, and the fire occurrence position and the fire behavior development condition can be quickly and accurately identified in the process.

Description

Technical Field

[0001] This invention relates to the field of fire early warning technology, specifically to an integrated fire emergency early warning, processing, and extinguishing control system based on fire doors. Background Technology

[0002] Fire doors are specially designed to prevent the spread of fire and protect the safety of people inside buildings. They are usually made of fire-resistant materials, can withstand high temperatures and maintain their integrity for a certain period of time, and prevent the spread of flames, smoke and toxic gases. Fire doors are widely used in residential, commercial buildings, industrial facilities and other buildings, especially in places where it is necessary to separate fire-resistant areas or ensure safe escape.

[0003] Traditional fire source location and intensity assessment systems rely on manual inspection or simple sensor data collection, making accurate fire source location and intensity assessment difficult. Delays and errors in fire source location can lead to untimely and inaccurate fire information, affecting the efficiency of firefighting and emergency response. Furthermore, traditional firefighting systems often require manual operation or a gradual response; for example, firefighting equipment may need to be manually activated or parameters manually adjusted, making it difficult to quickly activate and accurately extinguish fires in their early stages. Therefore, fires spread more rapidly, resulting in greater losses. Moreover, traditional systems are mostly static; once a fire occurs, the system cannot dynamically monitor fire changes in real time or automatically adjust firefighting strategies based on fire intensity or spread. Adjustments to firefighting equipment are often reactive, leading to a lack of timely and effective responses as the fire spreads. Finally, traditional systems typically lack escape route planning functions effectively integrated with the building structure. People may not know the safest and fastest escape routes during a fire, leading to chaotic evacuation and increasing the risk of casualties. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide an integrated fire emergency early warning and fire extinguishing control system based on fire doors.

[0005] The technical solution adopted to solve the above-mentioned technical problems is: an integrated fire emergency early warning, processing, and extinguishing control system based on fire doors, comprising: A fire source locating unit is used to acquire the temperature field, smoke concentration field and gas composition of each fire protection zone within the target building structure, and to determine the coordinates of the fire source center of the fire protection zone based on the temperature field. A fire intensity determination unit is used to determine the fire intensity field of the fire protection area based on the temperature field, the smoke concentration field, and the gas composition. A coverage determination unit is used to determine the fire coverage area of ​​the fire prevention zone based on the coordinates of the fire source center and the fire intensity field. An escape planning unit is used to construct an abstract structural diagram based on the target building structure, and to determine the escape path of the target building structure based on the fire coverage area of ​​the fire-prevention zone and the abstract structural diagram. The fire extinguishing control unit is used to determine the control parameters of the preset fire extinguishing equipment in the fire protection area according to the fire coverage area of ​​the fire protection area, and to control the fire extinguishing equipment in the fire protection area to extinguish the fire according to the control parameters.

[0006] Preferably, the abstract structural diagram includes a set of structural nodes and a set of structural connecting edges, wherein the structural nodes in the set of structural nodes correspond to the fire-resistant zones within the target building structure, and the structural connecting edges in the set of structural connecting edges correspond to the fire doors between the fire-resistant zones.

[0007] Preferably, the control parameters include nozzle pressure control parameters, nozzle extinguishing agent flow rate, and duration.

[0008] Preferably, determining the coordinates of the fire source center in the fireproof zone based on the temperature field includes: The temperature gradient field of the fire-resistant zone is calculated based on the temperature field of the fire-resistant zone, wherein the calculation formula for the temperature gradient field is as follows: ; in, Indicates the spatial location of the fire prevention zone and time Temperature gradient field below, Indicates the spatial location of the fire prevention zone and time The temperature field below; The squared modulus of the temperature gradient of the fire-resistant zone is calculated based on the temperature gradient field of the fire-resistant zone. The formula for calculating the squared modulus of the temperature gradient of the fire-resistant zone is as follows: ; in, Indicates the spatial location of the fire prevention zone and time The squared magnitude of the temperature gradient below; The coordinates of the fire source center in the fire protection zone are determined based on the square of the modulus of the temperature gradient in the fire protection zone. The formula for calculating the coordinates of the fire source center in the fire protection zone is as follows: ; in, This indicates the coordinates of the center of the fire source in the fire prevention zone.

[0009] Preferably, determining the fire intensity field of the fire-resistant zone based on the temperature field, the smoke concentration field, and the gas composition includes: Based on the temperature field, feature extraction is performed on the fire protection area to obtain the temperature field gradient features of the fire protection area; The energy proportion of the smoke concentration field is extracted based on wavelet packet decomposition to obtain the smoke concentration field characteristics of the fireproof area; The equivalent combustion efficiency is calculated based on the changes in gas composition to obtain the gas composition characteristics of the fireproof area; The attention weights are calculated based on the weights of the temperature field gradient characteristics, smoke concentration field characteristics, and gas composition characteristics of the fire protection area according to the attention mechanism. The attention weights are adjusted based on the gas composition characteristics of the fireproof area to obtain new attention weights; The temperature field gradient features, smoke concentration field features, and gas composition features of the fire protection area are weighted and summed according to the new attention weights to obtain the fused features of the fire protection area. The 3D convolutional neural network is used to predict the fire intensity of the fire protection area through the fused features, so as to obtain the fire intensity field of the fire protection area.

[0010] Preferably, the formula for calculating the temperature field gradient characteristics of the fireproof zone is as follows: ; in, This indicates the temperature field gradient characteristics of the fire-resistant zone. Represents a non-linear activation function. and Represents the learnable weights and biases; The formula for calculating the smoke concentration field characteristics of the fire-resistant zone is as follows: ; in, This indicates the characteristics of the smoke concentration field in the fire-prevention area. In wavelet packet decomposition, the first... The energy of each decomposition layer, and , Indicates the spatial location of the fire prevention zone and time The smoke concentration field below, Indicates the first Wavelet packet decomposition coefficients; The formula for calculating the gas composition characteristics is as follows: ; in, Indicates the gas composition characteristics of the fire-resistant area. , and Indicates the concentrations of carbon dioxide, carbon monoxide, and methane. Indicates the ignition concentration of alkane; The formula for calculating the attention weight is as follows: ; in, Indicates the first Spatiotemporal attention weights for various features This represents the activation function. and This represents the parameters obtained through learning. Indicates the first Type of characteristics; The formula for calculating the new attention weight is as follows: ; in, This represents the new attention weights. Indicates the learning rate. This represents the rate of change of equivalent combustion efficiency. Indicates an indicator function, when ,but Otherwise, .

[0011] Preferably, determining the fire coverage area of ​​the fire prevention zone based on the coordinates of the fire source center and the fire intensity field includes: The fire intensity field of the fire-protected area is compared with a preset fire intensity threshold to determine the boundary of the fire-covered area. The calculation formula for the boundary of the fire-covered area is as follows: ; in, Indicates the boundary of the area covered by the fire. Indicates the spatial location of the fire prevention zone and time The intensity of the fire field below, This indicates the preset fire intensity threshold. The fire coverage area of ​​the fire prevention zone is determined based on the coordinates of the fire source center and the boundary of the fire coverage area.

[0012] Preferably, determining the escape route of the target building structure based on the fire coverage area of ​​the fire-prevention zone and the abstract structural diagram includes: Semantic encoding is performed on the structural nodes in the abstract structural diagram. The semantic encoding includes coordinate features, spatial type encoding, maximum temperature value, exit distance, and fire coverage area. The coordinate features represent the position of the structural node in the three-dimensional space of the target building structure. The maximum temperature value represents the maximum temperature value of the structural node. The exit distance represents the shortest distance from the structural node to the nearest fire door. The weights of the structural connection edges in the abstract structural graph are dynamically calculated to obtain the dynamic weights of the structural connection edges. A semantically enhanced graph is constructed based on the semantic encoding of the structural nodes in the abstract structural graph and the dynamic weights of the structural connection edges; A Markov decision process model is performed on the target building structure based on the semantic augmentation graph to obtain a Markov decision process model. The Markov decision process model includes a state space, an action space, and a reward function. The state space is the semantic encoding of the structural nodes of the semantic augmentation graph. The action space is the transition from the current node to the adjacent node. Each action represents path selection, i.e., selecting the next structural node from the current state. The reward function is the negative value of the dynamic weight of the structural connection edge of the semantic augmentation graph. The Markov decision process model is solved using proximal strategy optimization to obtain the escape path of the target building structure.

[0013] Preferably, the formula for calculating the dynamic weight of the structural connection edge is as follows: ; in, Represents structural nodes and structural nodes Dynamic weights between them and Represents structural nodes and structural nodes The coordinates of the location This represents the preset thermal risk sensitivity coefficient. Represents structural nodes and structural nodes The average of the maximum temperature values, This indicates the area covered by the fire.

[0014] Preferably, the control parameters of the preset fire extinguishing equipment within the fire protection zone are determined based on the fire coverage area of ​​the fire protection zone, including: Based on the fire coverage area of ​​the fire prevention zone, the effectiveness of the fire extinguishing equipment is modeled to obtain the fire extinguishing effectiveness model of the fire extinguishing equipment, wherein the expression of the fire extinguishing effectiveness model is as follows: ; in, This represents a model illustrating the fire extinguishing effectiveness of fire extinguishing equipment. Indicates the first in the fire extinguishing equipment Extinguishing agent flow rate of each nozzle This indicates the preset extinguishing agent efficiency coefficient; Define the fire extinguishing objective function of the fire extinguishing equipment, wherein the fire extinguishing objective function is as follows: ; in, This represents the objective function of the fire extinguishing equipment. and This indicates the preset fire suppression weight. Indicates the first in the fire extinguishing equipment Pressure control parameters for each nozzle; Constraints are defined based on the fire extinguishing effectiveness model of the fire extinguishing equipment, wherein the expressions for the constraints are as follows: ; in, The minimum threshold representing the fire extinguishing effect of the nozzle. This indicates the total number of nozzles in the fire extinguishing equipment; The fire extinguishing objective function is solved by applying the alternating direction multiplier method under constraints to obtain the control parameters of the fire extinguishing equipment within the fire protection area.

[0015] The beneficial effects of the present invention are as follows: (1) The present invention obtains the temperature field, smoke concentration field and gas composition of each fire protection area of ​​the target building through the fire source positioning unit, which can accurately locate the center coordinates of the fire source and judge the intensity of the fire. This process can quickly and accurately identify the location of the fire and the development of the fire, providing important data support for subsequent processing. In addition, the coverage determination unit determines the coverage area of ​​the fire based on the center coordinates of the fire source and the fire intensity field, thereby realizing dynamic monitoring of the spread of the fire. This enables the fire extinguishing system to adjust the fire extinguishing strategy in real time to deal with different fire scenarios and minimize the losses caused by the fire. (2) The present invention combines the abstract diagram of the building structure with the fire coverage area through the escape planning unit, which can provide a safe escape route for people in the danger zone when a fire occurs. This can not only improve the escape efficiency, but also reduce the number of people injured. The invention provides clear escape guidance for people inside the building, and determines the fire coverage area through the fire extinguishing control unit. Combined with the preset control parameters of the fire extinguishing equipment, it automatically adjusts and starts the fire extinguishing equipment. This process does not require human intervention and can quickly extinguish the fire in the early stage of the fire, greatly improving the fire extinguishing efficiency and reducing the possibility of fire spread. (3) Through the coordinated work of multiple functional modules such as fire source location, fire analysis, escape planning and fire extinguishing control, the invention can achieve integrated emergency response. The seamless connection between different modules ensures the timeliness and accuracy of fire response, while reducing human error or delay. In addition, it can not only achieve timely fire warning and effective fire extinguishing, but also provide comprehensive fire safety management support for buildings. Through real-time monitoring and data feedback, building managers can adjust the emergency plan and fire-fighting facility configuration according to the specific fire situation. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall system architecture in one embodiment of the present invention.

[0017] Attached reference numerals: 1. Fire source location unit; 2. Fire intensity determination unit; 3. Coverage determination unit; 4. Escape planning unit; 5. Fire extinguishing control unit. Detailed Implementation

[0018] Example 1, as Figure 1 As shown, the present invention proposes an integrated fire emergency early warning, processing, and extinguishing control system based on fire doors, comprising: Fire source locating unit 1 is used to acquire the temperature field, smoke concentration field and gas composition of each fire protection zone in the target building structure, and to determine the coordinates of the fire source center of the fire protection zone based on the temperature field. Fire intensity determination unit 2 is used to determine the fire intensity field of the fire protection zone based on the temperature field, smoke concentration field and gas composition. Coverage determination unit 3 is used to determine the fire coverage area of ​​the fire protection zone based on the coordinates of the fire source center and the fire intensity field. Escape planning unit 4 is used to construct an abstract structural diagram based on the target building structure and to determine the escape path of the target building structure based on the fire coverage area of ​​the fire prevention zone and the abstract structural diagram. Fire extinguishing control unit 5 is used to determine the control parameters of the preset fire extinguishing equipment in the fire protection area according to the fire coverage area of ​​the fire protection area, and to control the fire extinguishing equipment in the fire protection area to extinguish the fire according to the control parameters.

[0019] In this invention, the temperature field refers to the temperature distribution at different locations within the fire protection zone, with the center of the fire source typically being the area with the highest temperature concentration; the smoke concentration field refers to the smoke concentration distribution at different locations within the fire protection zone, with higher smoke concentrations in fire-prone areas, thus monitoring smoke concentration can help locate the fire source; the fire intensity field is the spatial distribution of fire intensity in different areas of the fire, with high fire intensity near the fire source and relatively weaker fire intensity in areas far from the fire source, and the fire intensity can be calculated by analyzing data such as temperature, smoke concentration, and gas composition; the fire coverage area refers to the area affected by the fire, including the high-temperature area around the fire source, the smoke diffusion area, and the area damaged by the fire, and by analyzing the fire intensity field, it is possible to determine which areas need to be... Special attention and protection are required; based on the abstract diagram of the building structure, escape routes inside the building are planned. This helps design reasonable evacuation routes by analyzing the fire-covered area and the abstract structural diagram of the building; the abstract structural diagram is a simplified representation of the building, usually removing details and retaining only the spatial layout of important components such as rooms, passages, doors, and windows. Through this abstract diagram, escape routes and safety passages can be quickly assessed; escape routes refer to the routes that people can quickly and safely evacuate during a fire. The escape planning unit designs the optimal evacuation routes based on the fire's impact area and the building structure; fire extinguishing equipment refers to various equipment used to extinguish fires, such as fire extinguishers, automatic sprinkler systems (such as sprinkler systems), fire extinguishers, fire extinguishing systems, etc.

[0020] Example 2: The present invention proposes an integrated fire emergency early warning and fire extinguishing control system based on fire doors. Compared with Example 1, this example further includes: an abstract structural diagram including a set of structural nodes and a set of structural connection edges, wherein the structural nodes in the set of structural nodes correspond to the fire protection zones within the target building structure, and the structural connection edges in the set of structural connection edges correspond to the fire doors between the fire protection zones.

[0021] In an optional embodiment, the control parameters include nozzle pressure control parameters, nozzle extinguishing agent flow rate, and duration.

[0022] In an optional embodiment, determining the coordinates of the fire source center in the fire-resistant zone based on the temperature field includes: The temperature gradient field of the fire protection zone is calculated based on the temperature field of the fire protection zone. The formula for calculating the temperature gradient field is as follows: ; in, Indicates the spatial location of the fire prevention zone and time Temperature gradient field below, Indicates the spatial location of the fire prevention zone and time The temperature field below; The squared modulus of the temperature gradient in the fire-resistant zone is calculated based on the temperature gradient field of the fire-resistant zone. The formula for calculating the squared modulus of the temperature gradient in the fire-resistant zone is as follows: ; in, Indicates the spatial location of the fire prevention zone and time The squared magnitude of the temperature gradient below; The coordinates of the fire source center in the fire protection zone are determined based on the square of the modulus of the temperature gradient within the fire protection zone. The formula for calculating the coordinates of the fire source center in the fire protection zone is as follows: ; in, This indicates the coordinates of the center of the fire source in the fire prevention zone.

[0023] It's important to note that the temperature field describes how the temperature at each location within a fire-resistant area changes with space and time; the temperature gradient field, on the other hand, represents the rate and direction of temperature change—in other words, it shows how temperature changes spatially. If the temperature changes rapidly in certain areas, then the temperature gradient in those areas will also be larger. By calculating the "square modulus" of the temperature gradient field, we obtain a measure of the intensity of the temperature change; simply put, the larger this value, the more drastic the temperature change. Ignition sources are often located where temperature changes are most drastic, so the location with the largest square modulus of the temperature gradient is likely the center of the ignition source. Finally, by analyzing the square modulus of the temperature gradient across the entire area, we can find the location with the strongest temperature change; this location is the likely center of the ignition source. In other words, ignition sources are usually located where temperature changes are most drastic. Imagine a fire-prevention zone where the temperature distribution can be described by a function. For example, temperature variations in space and time might be due to temperature sources at different locations, resulting in temperature fluctuations over time. Assume the temperature change at a certain location is caused by several factors, such as changes in the surrounding air temperature or the temperature of burning objects. By analyzing the temperature changes at each location with spatial coordinates (e.g., x, y, z axes), we can derive the temperature gradient at each location. This tells us how the temperature at a certain location changes rapidly. Next, we calculate the intensity of temperature change at each location to find the area with the strongest temperature gradient. This area is likely the location of the fire source. Finally, we determine the location of the fire source by finding the location with the largest squared magnitude of the temperature gradient, which is usually the origin of the fire.

[0024] In an optional embodiment, determining the fire intensity field of the fire-resistant zone based on the temperature field, smoke concentration field, and gas composition includes: Features of the fire protection area are extracted based on the temperature field to obtain the temperature field gradient features of the fire protection area. The energy proportion of the smoke concentration field is extracted based on wavelet packet decomposition to obtain the characteristics of the smoke concentration field in the fire prevention area. The equivalent combustion efficiency is calculated based on the changes in gas composition in order to obtain the gas composition characteristics of the fireproof area; The attention weights are calculated by weighting the temperature field gradient characteristics, smoke concentration field characteristics, and gas composition characteristics of the fire protection area based on the attention mechanism. The attention weights are adjusted based on the gas composition characteristics of the fire prevention area to obtain new attention weights; The temperature field gradient features, smoke concentration field features, and gas composition features of the fire protection zone are weighted and summed according to the new attention weights to obtain the fused features of the fire protection zone. Fire intensity is predicted in fire-prevention zones by fusing features using a 3D convolutional neural network, thus obtaining the fire intensity field of the fire-prevention zone.

[0025] It should be noted that the temperature gradient field reflects temperature changes in different spaces and times. By analyzing these changes, we can derive temperature gradient characteristics, which can help us identify potential ignition sources within a fire-prevention area. Temperature changes are usually one of the most obvious indicators during a fire, and ignition sources are often located in areas of most drastic temperature changes. Changes in the smoke concentration field are also crucial for determining the occurrence of a fire. Through wavelet packet decomposition, the smoke concentration field can be decomposed into different frequency components, allowing for further analysis of which frequency components are most critical for fire source warning. High-concentration smoke is often closely related to ignition sources; analyzing changes in smoke can more accurately locate ignition sources. Changes in gas composition (such as carbon dioxide, oxygen, and smoke) during combustion can also serve as an important indicator for predicting fire intensity. By calculating the equivalent combustion efficiency, the combustion efficiency of the ignition source can be estimated. Intensity; the higher the combustion efficiency, the stronger the fire is usually; in this step, we combine temperature field gradient features, smoke concentration field features, and gas composition features, and use an attention mechanism to calculate their weights; the attention mechanism can automatically adjust the weights according to the importance of different features, so that the system pays more attention to those features that have a greater impact on fire prediction; according to the changes in gas composition features, we can adjust these weights to ensure that the model focuses more accurately on the most critical factors for fire prediction; if a feature has a greater impact on fire source prediction at the current moment, its weight will be increased, and vice versa; finally, the system will perform a weighted summation of each feature according to the adjusted attention weights to obtain a fused feature; these fused features bring together information such as temperature, smoke, and gas, and can effectively describe the current status of the fire-prevention area; by using a 3D convolutional neural network (3D We input these fused features into the model to predict the intensity of a fire. 3D CNNs can handle features in both spatial and temporal dimensions, helping the system identify fire intensity across different times and locations based on temperature, smoke concentration, and gas variations. For example, suppose we use this method for fire prediction in a forest fire prevention zone. First, we collect data on temperature, smoke concentration, and gas composition using sensors, and extract features from the smoke concentration field using wavelet packet decomposition. Next, combining the temperature field and gas composition data, we calculate the weights of different features using an attention mechanism, focusing on areas with significant temperature variations. Then, we adjust the weights to ensure that the impact of smoke concentration and gas variations on fire prediction is better reflected. Finally, by inputting all these features into a 3D convolutional neural network, the system can predict the location of the fire source and the intensity of the fire, thus providing accurate support for fire prevention and emergency response.

[0026] In an optional embodiment, the formula for calculating the temperature field gradient characteristics of the fire-resistant zone is as follows: ; in, This indicates the temperature field gradient characteristics of the fire-resistant zone. Represents a non-linear activation function. and Represents the learnable weights and biases; The formula for calculating the smoke concentration field characteristics of a fire-prevention zone is as follows: ; in, This indicates the characteristics of the smoke concentration field in the fire-prevention area. In wavelet packet decomposition, the first... The energy of each decomposition layer, and , Indicates the spatial location of the fire prevention zone and time The smoke concentration field below, Indicates the first Wavelet packet decomposition coefficients; The formula for calculating the gas composition characteristics is as follows: ; in, Indicates the gas composition characteristics of the fire-resistant area. , and Indicates the concentrations of carbon dioxide, carbon monoxide, and methane. Indicates the ignition concentration of alkane; The formula for calculating attention weights is as follows: ; in, Indicates the first Spatiotemporal attention weights for various features This represents the activation function. and This represents the parameters obtained through learning. Indicates the first Type of characteristics; The new formula for calculating attention weights is as follows: ; in, This represents the new attention weights. Indicates the learning rate. This represents the rate of change of equivalent combustion efficiency. Indicates an indicator function, when ,but Otherwise, .

[0027] In an optional embodiment, determining the fire coverage area of ​​the fire prevention zone based on the coordinates of the fire source center and the fire intensity field includes: The fire intensity field of the fire protection zone is compared with a preset fire intensity threshold to determine the boundary of the fire-covered area. The calculation formula for the boundary of the fire-covered area is as follows: ; in, Indicates the boundary of the area covered by the fire. Indicates the spatial location of the fire prevention zone and time The intensity of the fire field below, This indicates the preset fire intensity threshold. The fire coverage area of ​​the fire prevention zone is determined based on the coordinates of the fire source center and the boundary of the fire-covered area.

[0028] It should be noted that the fire intensity threshold is a preset limit value used to distinguish between areas with more severe fires and areas with less severe fires. Areas where the fire intensity exceeds this threshold are usually considered dangerous areas where a fire has already occurred or is spreading. The boundary of the fire coverage area refers to the farthest extent to which the fire has spread, usually referring to areas where the fire intensity reaches or exceeds the preset fire intensity threshold. This boundary is of great significance in fire prevention strategies because it represents the extent of the fire's spread and directly affects key areas for fire rescue, evacuation, and fire prevention.

[0029] In an optional embodiment, determining the escape route of the target building structure based on the fire coverage area of ​​the fire-prevention zone and an abstract structural diagram includes: Semantic encoding is performed on the structural nodes in the abstract structural diagram. The semantic encoding includes coordinate features, spatial type encoding, maximum temperature value, exit distance, and fire coverage area. The coordinate features represent the position of the structural node in the three-dimensional space of the target building structure, the maximum temperature value represents the maximum temperature of the structural node, and the exit distance represents the shortest distance from the structural node to the nearest fire door. The weights of the structural connection edges in the abstract structural graph are dynamically calculated to obtain the dynamic weights of the structural connection edges. A semantically enhanced graph is constructed based on the semantic encoding of structural nodes and the dynamic weights of structural connection edges in the abstract structural graph; A Markov decision process model is performed on the target building structure based on the semantic augmentation graph to obtain the Markov decision process model. The Markov decision process model includes a state space, an action space, and a reward function. The state space is the semantic encoding of the structural nodes of the semantic augmentation graph. The action space is the transition from the current node to the adjacent node. Each action represents path selection, that is, selecting the next structural node from the current state. The reward function is the negative value of the dynamic weight of the structural connection edge of the semantic augmentation graph. The Markov decision process model is solved based on proximal strategy optimization to obtain the escape path of the target building structure.

[0030] In an optional embodiment, the dynamic weights of the structural connection edges are calculated using the following formula: ; in, Represents structural nodes and structural nodes Dynamic weights between them and Represents structural nodes and structural nodes The coordinates of the location This represents the preset thermal risk sensitivity coefficient. Represents structural nodes and structural nodes The average of the maximum temperature values, This indicates the area covered by the fire.

[0031] It should be noted that dynamic weights mean that the weights of edges change over time or due to other factors. In a fire, the weights of edges may be related to the impact of the fire, the difficulty of escape, or the danger, such as smoke concentration or fire intensity. Semantic augmented graphs are graphs constructed by combining the semantic encoding of nodes and the dynamic weights of edges. The nodes and edges in the graph not only contain spatial and connectivity information but also fire-related safety information and dangers, providing richer context. Markov decision processes are mathematical models used to describe decision-making processes in a given environment. Proximal policy optimization is a reinforcement learning algorithm used to solve Markov decision process models. It maximizes cumulative rewards by updating the policy while maintaining the stability and effectiveness of the policy. PPO avoids drastic policy changes by updating the "proximal" side of the policy, ensuring the stability of the training process.

[0032] In an optional embodiment, the control parameters of preset fire extinguishing equipment within the fire protection zone are determined based on the fire coverage area of ​​the fire protection zone, including: Based on the fire coverage area of ​​the fire prevention zone, the effectiveness of the fire extinguishing equipment is modeled to obtain the fire extinguishing effectiveness model of the fire extinguishing equipment. The expression of the fire extinguishing effectiveness model is as follows: ; in, This represents a model illustrating the fire extinguishing effectiveness of fire extinguishing equipment. Indicates the first in the fire extinguishing equipment Extinguishing agent flow rate of each nozzle This indicates the preset extinguishing agent efficiency coefficient; Define the fire extinguishing objective function for the fire extinguishing equipment, where the fire extinguishing objective function is as follows: ; in, This represents the objective function of the fire extinguishing equipment. and This indicates the preset fire suppression weight. Indicates the first in the fire extinguishing equipment Pressure control parameters for each nozzle; The constraints are defined based on the fire extinguishing effectiveness model of the fire extinguishing equipment, and the expressions for the constraints are as follows: ; in, The minimum threshold representing the fire extinguishing effect of the nozzle. This indicates the total number of nozzles in the fire extinguishing equipment; The fire extinguishing objective function is solved by applying the alternating direction multiplier method and constraints to obtain the control parameters of the fire extinguishing equipment within the fire protection area.

[0033] It should be noted that the fire extinguishing efficiency model refers to a mathematical model used to describe the fire extinguishing capability of fire extinguishing equipment. The fire extinguishing efficiency model is an expression representing the fire extinguishing capability of fire extinguishing equipment (including each nozzle). The minimum threshold of the fire extinguishing effect of the nozzle represents the minimum fire extinguishing effect requirement that each nozzle must meet, that is, the efficiency of each nozzle cannot be lower than this value to ensure the basic fire extinguishing effect. The Alternating Direction Multiplier Method (ADMM) is an optimization algorithm, usually used to solve optimization problems with constraints. In this problem, ADMM is used to solve the fire extinguishing objective function while considering the constraints. ADMM decomposes a large-scale optimization problem into multiple smaller subproblems, which facilitates the solution. The key idea of ​​ADMM is to gradually approach the optimal solution by alternating updates between multiple subproblems.

[0034] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A fire emergency early warning, processing, and extinguishing integrated control system based on fire doors, characterized in that, include: Fire source locating unit (1), the fire source locating unit (1) is used to obtain the temperature field, smoke concentration field and gas composition of each fire protection zone in the target building structure, and determine the fire source center coordinates of the fire protection zone according to the temperature field; Fire intensity determination unit (2), the fire intensity determination unit (2) is used to determine the fire intensity field of the fire protection area based on the temperature field, the smoke concentration field and the gas composition; Coverage determination unit (3), the coverage determination unit (3) is used to determine the fire coverage area of ​​the fire prevention zone according to the coordinates of the fire source center and the fire intensity field; Escape planning unit (4), the escape planning unit (4) is used to construct an abstract structural diagram based on the target building structure, and to determine the escape path of the target building structure based on the fire coverage area of ​​the fire prevention zone and the abstract structural diagram; Fire extinguishing control unit (5) is used to determine the control parameters of the preset fire extinguishing equipment in the fire protection area according to the fire coverage area of ​​the fire protection area, and control the fire extinguishing equipment in the fire protection area to extinguish the fire according to the control parameters.

2. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 1, characterized in that, The abstract structural diagram includes a set of structural nodes and a set of structural connecting edges. The structural nodes in the set of structural nodes correspond to the fire-resistant zones within the target building structure, and the structural connecting edges in the set of structural connecting edges correspond to the fire doors between the fire-resistant zones.

3. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 2, characterized in that, The control parameters include the nozzle pressure control parameters, the nozzle extinguishing agent flow rate, and the duration.

4. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 3, characterized in that, Determining the coordinates of the fire source center in the fire protection zone based on the temperature field includes: The temperature gradient field of the fire-resistant zone is calculated based on the temperature field of the fire-resistant zone, wherein the calculation formula for the temperature gradient field is as follows: ； in, Indicates the spatial location of the fire prevention zone and time Temperature gradient field below, Indicates the spatial location of the fire prevention zone and time The temperature field below; The squared modulus of the temperature gradient of the fire-resistant zone is calculated based on the temperature gradient field of the fire-resistant zone. The formula for calculating the squared modulus of the temperature gradient of the fire-resistant zone is as follows: ； in, Indicates the spatial location of the fire prevention zone and time The squared magnitude of the temperature gradient below; The coordinates of the fire source center in the fire protection zone are determined based on the square of the modulus of the temperature gradient in the fire protection zone. The formula for calculating the coordinates of the fire source center in the fire protection zone is as follows: ； in, This indicates the coordinates of the center of the fire source in the fire prevention zone.

5. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 4, characterized in that, Determining the fire intensity field of the fire-resistant zone based on the temperature field, the smoke concentration field, and the gas composition includes: Based on the temperature field, feature extraction is performed on the fire protection area to obtain the temperature field gradient features of the fire protection area; The energy proportion of the smoke concentration field is extracted based on wavelet packet decomposition to obtain the smoke concentration field characteristics of the fireproof area; The equivalent combustion efficiency is calculated based on the changes in gas composition to obtain the gas composition characteristics of the fireproof area; The attention weights are calculated based on the weights of the temperature field gradient characteristics, smoke concentration field characteristics, and gas composition characteristics of the fire protection area according to the attention mechanism. The attention weights are adjusted based on the gas composition characteristics of the fireproof area to obtain new attention weights; The temperature field gradient features, smoke concentration field features, and gas composition features of the fire protection area are weighted and summed according to the new attention weights to obtain the fused features of the fire protection area. The 3D convolutional neural network is used to predict the fire intensity of the fire protection area through the fused features, so as to obtain the fire intensity field of the fire protection area.

6. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 5, characterized in that, The formula for calculating the temperature field gradient characteristics of the fire-resistant zone is as follows: ； in, This indicates the temperature field gradient characteristics of the fire-resistant zone. Represents a non-linear activation function. and Represents the learnable weights and biases; The formula for calculating the smoke concentration field characteristics of the fire-resistant zone is as follows: ； in, This indicates the characteristics of the smoke concentration field in the fire-prevention area. In wavelet packet decomposition, the first... The energy of each decomposition layer, and , Indicates the spatial location of the fire prevention zone and time The smoke concentration field below, Indicates the first Wavelet packet decomposition coefficients; The formula for calculating the gas composition characteristics is as follows: ； in, Indicates the gas composition characteristics of the fire-resistant area. , and Indicates the concentrations of carbon dioxide, carbon monoxide, and methane. Indicates the ignition concentration of alkane; The formula for calculating the attention weight is as follows: ； in, Indicates the first Spatiotemporal attention weights for various features This represents the activation function. and This represents the parameters obtained through learning. Indicates the first Type of characteristics; The formula for calculating the new attention weight is as follows: ； in, This represents the new attention weights. Indicates the learning rate. This represents the rate of change of equivalent combustion efficiency. Indicates an indicator function, when ,but Otherwise, .

7. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 6, characterized in that, The fire coverage area of ​​the fire prevention zone is determined based on the coordinates of the fire source center and the fire intensity field, including: The fire intensity field of the fire-protected area is compared with a preset fire intensity threshold to determine the boundary of the fire-covered area. The calculation formula for the boundary of the fire-covered area is as follows: ； in, Indicates the boundary of the area covered by the fire. Indicates the spatial location of the fire prevention zone and time The intensity of the fire field below, This indicates the preset fire intensity threshold. The fire coverage area of ​​the fire prevention zone is determined based on the coordinates of the fire source center and the boundary of the fire coverage area.

8. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 7, characterized in that, Based on the fire coverage area of ​​the fire-prevention zone and the abstract structural diagram, the escape routes of the target building structure are determined, including: Semantic encoding is performed on the structural nodes in the abstract structural diagram. The semantic encoding includes coordinate features, spatial type encoding, maximum temperature value, exit distance, and fire coverage area. The coordinate features represent the position of the structural node in the three-dimensional space of the target building structure. The maximum temperature value represents the maximum temperature value of the structural node. The exit distance represents the shortest distance from the structural node to the nearest fire door. The weights of the structural connection edges in the abstract structural graph are dynamically calculated to obtain the dynamic weights of the structural connection edges. A semantically enhanced graph is constructed based on the semantic encoding of the structural nodes in the abstract structural graph and the dynamic weights of the structural connection edges; A Markov decision process model is performed on the target building structure based on the semantic augmentation graph to obtain a Markov decision process model. The Markov decision process model includes a state space, an action space, and a reward function. The state space is the semantic encoding of the structural nodes of the semantic augmentation graph. The action space is the transition from the current node to the adjacent node. Each action represents path selection, i.e., selecting the next structural node from the current state. The reward function is the negative value of the dynamic weight of the structural connection edge of the semantic augmentation graph. The Markov decision process model is solved using proximal strategy optimization to obtain the escape path of the target building structure.

9. A fire emergency early warning, processing, and extinguishing integrated control system based on fire doors according to claim 8, characterized in that, The formula for calculating the dynamic weight of the structural connection edge is as follows: ； in, Represents structural nodes and structural nodes Dynamic weights between them and Represents structural nodes and structural nodes The coordinates of the location This represents the preset thermal risk sensitivity coefficient. Represents structural nodes and structural nodes The average of the maximum temperature values, This indicates the area covered by the fire.

10. The integrated fire emergency early warning, processing, and extinguishing control system based on fire doors according to claim 9, characterized in that, The control parameters of the preset fire extinguishing equipment within the fire protection zone are determined based on the fire coverage area of ​​the fire protection zone, including: Based on the fire coverage area of ​​the fire prevention zone, the effectiveness of the fire extinguishing equipment is modeled to obtain the fire extinguishing effectiveness model of the fire extinguishing equipment, wherein the expression of the fire extinguishing effectiveness model is as follows: ； in, This represents a model illustrating the fire extinguishing effectiveness of fire extinguishing equipment. Indicates the first in the fire extinguishing equipment Extinguishing agent flow rate of each nozzle This indicates the preset extinguishing agent efficiency coefficient; Define the fire extinguishing objective function of the fire extinguishing equipment, wherein the fire extinguishing objective function is as follows: ； in, This represents the fire extinguishing objective function of the fire extinguishing equipment. and This indicates the preset fire suppression weight. Indicates the first in the fire extinguishing equipment Pressure control parameters for each nozzle; Constraints are defined based on the fire extinguishing effectiveness model of the fire extinguishing equipment, wherein the expressions for the constraints are as follows: ； in, The minimum threshold representing the fire extinguishing effect of the nozzle. This indicates the total number of nozzles in the fire extinguishing equipment; The fire extinguishing objective function is solved by applying the alternating direction multiplier method under constraints to obtain the control parameters of the fire extinguishing equipment within the fire protection area.

Images