Fire management method and system based on big data processing

Abstract

The invention provides a fire management method and system based on big data processing. The method comprises the following steps: determining whether a fire occurs by using a judgment model; if it is determined that the fire occurs, an information processing model is used for processing the monitoring video of the internal place of the enterprise to determine a fire area, internal place layout information, initial flame information and the wind speed of the fire area; the average rescue time of the rescue workers is determined based on the determination model, and the flame size after the average rescue time is determined by using the flame information determination model based on the average rescue time of the rescue workers, the fire area, the internal place layout information, the flame information and the wind speed of the fire area; based on the flame size and the foam extinguishing agent information after the average rescue time, the use amount of the foam extinguishing agent is determined; and sending the usage amount of the foam extinguishing agent to a terminal of a rescue worker and notifying to start rescue. According to the method, the usage amount of the foam extinguishing agent can be managed when a fire occurs in an internal place of an enterprise.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise management, and in particular to a fire management method and system based on big data processing. Background Art

[0002] Fire is a form of uncontrolled combustion that poses a significant threat to a company's production and operations. The earlier a fire is discovered, the more serious damage it can cause.

[0003] In the prior art, smoke alarms are often used to manage fires in internal places of an enterprise. Smoke alarms mainly rely on detecting smoke particles in the air to sound an alarm. When a fire occurs, the smoke alarm will trigger an alarm. When a fire occurs, it is often extinguished by using fire extinguishing agents. Foam fire extinguishing agent is a commonly used fire extinguishing agent that is widely used in various fire scenarios. If a fire occurs in an internal place of an enterprise, in order to ensure the effective use of foam fire extinguishing agent, the amount of foam fire extinguishing agent needs to be determined in advance. The prior art often manually determines the amount of foam fire extinguishing agent based on the size of the fire. If the amount of foam fire extinguishing agent is determined too much, the time required to carry and prepare the foam fire extinguishing agent will be longer when extinguishing the fire, which may cause the fire to grow further over time. If the amount of foam fire extinguishing agent is determined too little, the fire may not be extinguished.

[0004] Therefore, how to manage the amount of foam fire extinguishing agent used when a fire occurs in an enterprise is an urgent problem to be solved. Summary of the Invention

[0005] The main technical problem solved by the present invention is how to manage the amount of foam fire extinguishing agent used when a fire occurs in an internal place of an enterprise.

[0006] According to a first aspect, the present invention provides a fire management method based on big data processing, comprising: obtaining surveillance videos of internal places of an enterprise, temperature sensor data of internal places of the enterprise, and smoke detector data of internal places of the enterprise; using a judgment model to determine whether a fire has occurred based on the surveillance videos of the internal places of the enterprise, the temperature sensor data of the internal places of the enterprise, and the smoke detector data of the internal places of the enterprise; if it is determined that a fire has occurred, using an information processing model to process the surveillance videos of the internal places of the enterprise to determine the fire area, internal place layout information, initial flame information, and wind speed in the fire area; based on the fire area, the location of the foam extinguishing agent, , the rescue personnel information uses a rescue time determination model to determine the average rescue time of the rescue personnel, the rescue personnel information includes the position of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers; based on the average rescue time of the rescue personnel, the fire area, the internal venue layout information, the initial flame information, and the wind speed of the fire area, a flame information determination model is used to determine the flame size after the average rescue time; based on the flame size after the average rescue time and the foam fire extinguishing agent information, the amount of foam fire extinguishing agent is determined; the amount of foam fire extinguishing agent is sent to the rescue personnel's terminal and the rescue personnel are notified to start the rescue.

[0007] Furthermore, the method further includes: if it is determined that no fire has occurred, using the judgment model again after an interval of ten minutes to determine whether a fire has occurred.

[0008] Furthermore, the internal place layout information includes the total area of the internal place, the height of the internal place, the floor material of the internal place, and multiple equipment information of the internal place, and the initial flame information includes flame shape, flame size, flame horizontal growth rate, and flame vertical growth rate.

[0009] Furthermore, the information processing model is a gated loop unit, the input of the information processing model is the surveillance video of the internal premises of the enterprise, and the output of the information processing model is the fire area, the internal premises layout information, the initial flame information, and the wind speed of the fire area.

[0010] Furthermore, the information processing model includes a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model. The input of the global processing sub-model is the surveillance video of the internal premises of the enterprise, and the output of the global processing sub-model is the segmented video of the fire area and the internal premises layout information. The input of the flame processing sub-model is the segmented video of the fire area, and the output of the flame processing sub-model is the fire area and the initial flame information. The input of the wind speed determination sub-model is the segmented video of the fire area and the initial flame information, and the output of the wind speed determination sub-model is the wind speed of the fire area.

[0011] According to the second aspect, the present invention provides a fire management system based on big data processing, including: an acquisition module for acquiring surveillance videos of internal places of an enterprise, temperature sensor data of internal places of an enterprise, and smoke detector data of internal places of an enterprise; a fire judgment module for using a judgment model to determine whether a fire has occurred based on the surveillance videos of the internal places of the enterprise, the temperature sensor data of the internal places of the enterprise, and the smoke detector data of the internal places of the enterprise; an information processing module for, if it is determined that a fire has occurred, using the information processing model to process the surveillance videos of the internal places of the enterprise to determine the fire area, internal place layout information, initial flame information, and wind speed in the fire area; a rescue time determination module for determining the fire area, internal place layout information, initial flame information, and wind speed in the fire area based on the fire area, foam extinguishing The fire agent position and rescuer information use a rescue time determination model to determine the average rescue time of the rescuers, and the rescuer information includes the position of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers; a flame size determination module is used to determine the flame size after the average rescue time based on the average rescue time of the rescuers, the fire area, the internal venue layout information, the initial flame information, and the wind speed of the fire area using a flame information determination model; a dosage determination module is used to determine the dosage of the foam fire extinguishing agent based on the flame size after the average rescue time and the foam fire extinguishing agent information; a sending module is used to send the dosage of the foam fire extinguishing agent to the rescuer's terminal and notify the rescuer to start the rescue.

[0012] Furthermore, the system is also used to: if it is determined that no fire has occurred, use the judgment model again after an interval of ten minutes to determine whether a fire has occurred.

[0013] Furthermore, the internal place layout information includes the total area of the internal place, the height of the internal place, the floor material of the internal place, and multiple equipment information of the internal place, and the initial flame information includes flame shape, flame size, flame horizontal growth rate, and flame vertical growth rate.

[0014] Furthermore, the information processing model is a gated loop unit, the input of the information processing model is the surveillance video of the internal premises of the enterprise, and the output of the information processing model is the fire area, the internal premises layout information, the initial flame information, and the wind speed of the fire area.

[0015] Furthermore, the information processing model includes a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model. The input of the global processing sub-model is the surveillance video of the internal premises of the enterprise, and the output of the global processing sub-model is the segmented video of the fire area and the internal premises layout information. The input of the flame processing sub-model is the segmented video of the fire area, and the output of the flame processing sub-model is the fire area and the initial flame information. The input of the wind speed determination sub-model is the segmented video of the fire area and the initial flame information, and the output of the wind speed determination sub-model is the wind speed of the fire area.

[0016] The present invention provides a fire management method and system based on big data processing, which includes obtaining surveillance videos of internal places of an enterprise, temperature sensor data of internal places of the enterprise, and smoke detector data of internal places of the enterprise; using a judgment model to determine whether a fire has occurred based on the surveillance videos of the internal places of the enterprise, the temperature sensor data of the internal places of the enterprise, and the smoke detector data of the internal places of the enterprise; if it is determined that a fire has occurred, using an information processing model to process the surveillance videos of the internal places of the enterprise to determine the fire area, internal place layout information, initial flame information, and wind speed in the fire area; using a rescue time determination model based on the fire area, the location of the foam extinguishing agent, and the information of the rescue personnel The method is characterized in that the average rescue time of rescuers is determined by a model, and the rescuer information includes the position of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers; the flame information is used to determine the flame size after the average rescue time based on the average rescue time of the rescuers, the fire area, the internal venue layout information, the initial flame information, and the wind speed of the fire area; the amount of foam fire extinguishing agent is determined based on the flame size after the average rescue time and foam fire extinguishing agent information; the amount of foam fire extinguishing agent is sent to the rescuer's terminal and the rescuer is notified to start the rescue. The method can manage the amount of foam fire extinguishing agent used when a fire occurs in an internal venue of an enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of a fire management method based on big data processing provided by an embodiment of the present invention;

[0018] Figure 2 A schematic diagram of a fire management system based on big data processing provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In an embodiment of the present invention, there is provided Figure 1 A fire management method based on big data processing is shown, and the fire management method based on big data processing includes steps S1 to S7:

[0020] Step S1, obtaining surveillance videos of places within the enterprise, temperature sensor data of places within the enterprise, and smoke detector data of places within the enterprise.

[0021] The surveillance video of an enterprise's internal premises is a real-time video stream obtained by camera equipment installed in the enterprise's internal premises.

[0022] The temperature sensor data of the enterprise's internal locations is the real-time temperature data obtained by the sensor devices installed in the enterprise's internal locations.

[0023] The smoke detector data of the enterprise's internal places is the real-time smoke concentration data obtained by the smoke detector devices installed in the enterprise's internal places.

[0024] Collecting surveillance videos, temperature sensor data, and smoke detector data from internal locations of an enterprise can provide information about the environment at the time of the fire and be used for fire diagnosis. This data will provide the basis for subsequent steps.

[0025] Step S2: Determine whether a fire has occurred using a judgment model based on the surveillance video of the enterprise's internal premises, the temperature sensor data of the enterprise's internal premises, and the smoke detector data of the enterprise's internal premises.

[0026] The judgment model is a gated cyclic unit model, the input of the judgment model is the surveillance video of the internal premises of the enterprise, the temperature sensor data of the internal premises of the enterprise, and the smoke detector data of the internal premises of the enterprise, and the output of the judgment model is whether a fire has occurred or not.

[0027] The Gated Recurrent Unit model includes the Gated Recurrent Unit (GRU). The GRU model controls the flow of information by introducing a gating mechanism. It incorporates two key gating units: the update gate and the reset gate, which dynamically adjust the importance of information through learning. The update gate controls the weight between the previous hidden state and the current input, while the reset gate controls whether to reset historical information to the default value. These two gating units enable the GRU model to adaptively retain old information and acquire new information, better capturing both short-term and long-term dependencies in the input sequence.

[0028] By using a gated recurrent unit model, we can effectively process sequential data such as video data, temperature sensor data, and smoke detector data. This allows us to determine whether a fire has occurred during fire monitoring and capture the dependencies between different time steps, improving the accuracy and efficiency of fire detection. The judgment model can be trained using gradient descent on training samples.

[0029] Step S3: If it is determined that a fire has occurred, the surveillance video of the internal premises of the enterprise is processed using an information processing model to determine the fire area, internal layout information, initial flame information, and wind speed in the fire area.

[0030] The information processing model is a gated loop unit, the input of the information processing model is the surveillance video of the internal premises of the enterprise, and the output of the information processing model is the fire area, the internal premises layout information, the initial flame information, and the wind speed of the fire area.

[0031] The fire zone is the specific area or location within the enterprise where the fire actually occurs.

[0032] The interior space layout information includes a total area of the interior space, a height of the interior space, a floor material of the interior space, and a plurality of equipment information of the interior space.

[0033] The total area of the internal premises refers to the overall floor space of the enterprise, usually in square meters (m 2 ) as the unit.

[0034] The height of the internal place refers to the height of the internal place of the enterprise, usually in meters (m).

[0035] The floor materials of internal places refer to the floor or ground covering materials of the internal places of the enterprise, such as concrete, wood, tiles, etc.

[0036] The plurality of equipment information of the internal location includes the types, sizes, and locations of various equipment existing in the internal location of the enterprise. For example, the types of various equipment include production equipment, electrical equipment, and fire-fighting equipment.

[0037] The initial flame information includes flame shape, flame size, flame horizontal growth speed, and flame vertical growth speed.

[0038] Flame morphology refers to the shape or appearance of the flame, such as fireball, jet, combustion bed, etc.

[0039] Flame size refers to the size or area of the flame, usually measured in square meters (m 2 ) as the unit.

[0040] The horizontal flame growth rate refers to the speed at which the flame spreads horizontally within the surveillance video of the enterprise's internal location, typically measured in meters per second (m / s). For example, if the surveillance video of the enterprise's internal location is 10 seconds long and the flame's horizontal growth rate is 0.1 m / s, the flame spreads horizontally by 0.1 m every second.

[0041] The vertical flame growth rate refers to the speed at which the flame spreads vertically within the surveillance video of the enterprise's internal premises, typically measured in meters per second (m / s). For example, if the surveillance video of the enterprise's internal premises is 10 seconds long and the vertical flame growth rate is 0.1 m / s, the flame spreads vertically by 0.1 m every second.

[0042] The wind speed in the fire area refers to the air flow speed near the fire area, usually measured in meters per second (m / s).

[0043] When processing surveillance video, the information processing model breaks it down into a series of continuous image frames, which are then fed into the information processing model frame by frame for processing. Through the model's recurrent structure, the information processing model can capture temporal changes and dynamic features in the video, including the fire area, internal layout information, initial flame information, and wind speed in the fire area.

[0044] In some embodiments, the information processing model includes a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model. The global processing sub-model, the flame processing sub-model, and the wind speed determination sub-model are all gated recurrent units. The global processing sub-model receives as input a surveillance video of the enterprise's internal premises, and outputs a segmented video of the fire area and information about the internal premises layout. The flame processing sub-model receives as input a segmented video of the fire area, and outputs the fire area and initial flame information. The wind speed determination sub-model receives as input a segmented video of the fire area and the initial flame information, and outputs the wind speed determination sub-model.

[0045] By dividing the information processing model into a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model, each sub-model focuses on solving a specific problem, improving the model's adaptability and expressiveness for different tasks. The global processing sub-model is responsible for processing the overall scene and extracting internal layout information. The flame processing sub-model is responsible for segmenting the fire area and extracting initial flame information. The wind speed determination sub-model is responsible for determining the wind speed in the fire area. By decomposing tasks into different sub-models, the specific information of each sub-task can be better captured and utilized.

[0046] Segmented videos of fire areas can provide visual features about the fire zone. For example, the movement patterns of flames and the dispersion of smoke in the video are correlated with wind speed. By analyzing and processing the fire zone video, the wind speed determination sub-model can extract wind speed clues from these features. During a fire, the heat and smoke generated by the flames cause changes in the air flow around the fire zone. The wind speed determination sub-model can analyze the segmented videos of the fire zone to detect air flow patterns and infer the wind speed within the fire zone. For example, the twisting shape of the flames and the direction of smoke movement can be used to estimate wind speed. As a gated recurrent unit, the wind speed determination sub-model has the ability to learn features and patterns. By training on a large amount of training data, the wind speed determination sub-model can learn the underlying relationship between the segmented videos of the fire zone and wind speed. During the inference phase, the model can predict the wind speed within the fire zone based on these learned features and patterns.

[0047] In some embodiments, if it is determined that no fire has occurred, the judgment model is used again to determine whether a fire has occurred after an interval of ten minutes.

[0048] Step S4, determining the average rescue time of rescuers using a rescue time determination model based on the fire area, the location of the foam extinguishing agent, and the rescuer information, wherein the rescuer information includes the location of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers.

[0049] The foam location refers to the storage or placement of foam fire extinguishing agents. These locations are typically pre-determined within a building to facilitate rapid deployment in the event of a fire. For example, dry powder fire extinguishers may be placed near emergency exits on each floor. The foam location can be pre-entered by staff.

[0050] Rescuer information includes each rescuer's location, age, running speed, and number of rescuers. The rescuer's location indicates their current location and is typically obtained through a Global Positioning System (GPS) or other positioning technology. Age information is used to assess each rescuer's physical condition and mobility. Running speed indicates how quickly the rescuer moves when reaching the fire area or other target location. The number of rescuers indicates the total number of people available for rescue.

[0051] The average rescue time for rescuers is the average time it takes for rescuers to arrive at the fire area. This value indicates the average time it takes for rescuers to arrive at the fire area after receiving the call.

[0052] The rescue time determination model is an artificial neural network model. The input of the rescue time determination model is the fire area, the location of the foam fire extinguishing agent, and the rescuer information. The output of the rescue time determination model is the average rescue time of the rescuers.

[0053] An artificial neural network (ANN) is a computational model that mimics the working principles of biological neural networks. It consists of a large number of interconnected artificial neurons, which transmit and process information through these connections. Through their multi-layered architecture and effective training strategies, ANNs enable more efficient and accurate modeling and processing of complex problems. ANNs can learn complex mappings from input to output, enabling prediction and processing of unknown data.

[0054] Artificial neural networks consist of an input layer, an output layer, and a hidden layer in between. Each layer is composed of many neurons. Each neuron receives input signals from the previous layer, performs calculations, and passes the results to the next layer after applying an activation function. During this process, the weights and biases between neurons are constantly adjusted to continuously optimize the network's output.

[0055] The rescue time determination model considers multiple factors, such as the location of the fire area, the location of the foam, and the information of the rescuers. By calculating the time it takes for rescuers to reach the fire area and combining the times of all rescuers, the average rescue time can be determined. This is done to assess the ability of rescuers to reach the fire area quickly after a fire breaks out, allowing for timely action and ensuring the safety of personnel and the success of firefighting efforts.

[0056] Step S5, determining the flame size after the average rescue time using a flame information determination model based on the average rescue time of the rescuers, the fire area, the internal location layout information, the initial flame information, and the wind speed of the fire area.

[0057] The flame size after the average rescue time is the flame size estimated by the flame information determination model after the average rescue time. Because rescue operations take time, and firefighting doesn't begin until rescuers arrive, it's important to estimate the flame size after the average rescue time to determine the amount of foam to be used.

[0058] The flame information determination model is an artificial neural network model. Its inputs are the average rescue time of the rescuers, the fire area, the internal layout information, the initial flame information, and the wind speed in the fire area. Its output is the flame size after the average rescue time.

[0059] An artificial neural network (ANN) is a computational model that mimics the working principles of biological neural networks. It consists of a large number of interconnected artificial neurons, which transmit and process information through these connections. Through their multi-layered architecture and effective training strategies, ANNs enable more efficient and accurate modeling and processing of complex problems. ANNs can learn complex mappings from input to output, enabling prediction and processing of unknown data.

[0060] Artificial neural networks include feedforward neural networks, recurrent neural networks, convolutional neural networks, etc.

[0061] The flame information determination model combines factors such as the average rescue time of rescuers, the layout information of the fire area, flame information, and the wind speed in the fire area to simulate and predict fires, and then determines the flame size after the average rescue time.

[0062] The flame information determination model simulates and predicts fires by combining factors such as the average rescue time for rescuers, the layout of the fire area, flame information, and wind speed in the fire area. This model estimates the size of the flame after the average rescue time. Specifically, the flame information determination model uses an artificial neural network model. By training and learning from a large amount of fire data, it develops a model capable of predicting fire development.

[0063] The model's inputs include information such as the average rescue time for rescuers, the fire area, interior layout, initial flame information, and wind speed in the fire area. This information reflects key factors influencing fire development. By feeding this information into the flame information determination model, the model can simulate and predict fires based on existing data and experience, thereby estimating the flame size after the average rescue time.

[0064] Step S6: determining the amount of foam fire extinguishing agent to be used based on the flame size and foam fire extinguishing agent information after the average rescue time.

[0065] Foam fire extinguishing agent information includes the type, composition, performance characteristics and instructions for use of the foam fire extinguishing agent.

[0066] In some embodiments, the amount of foam fire extinguishing agent used may be determined based on a pre-set matching relationship among the flame size after the average rescue time, the foam fire extinguishing agent information, and the amount of foam fire extinguishing agent used.

[0067] In some embodiments, the flame size and foam extinguishing agent information after the average rescue time can be constructed as a to-be-matched vector. By calculating the distance between the to-be-matched vector and each reference vector in a database, the amount of foam extinguishing agent corresponding to the reference vector whose distance is less than a threshold is determined as the current amount of foam extinguishing agent. The database is pre-constructed and includes reference vectors and the amount of foam extinguishing agent corresponding to the reference vectors. The reference vectors are constructed based on the flame size and foam extinguishing agent information after the average rescue time in historical data. The amount of foam extinguishing agent corresponding to the reference vectors is the amount of foam extinguishing agent determined in the historical data.

[0068] Step S7: sending the amount of the foam fire extinguishing agent to the rescuer's terminal and notifying the rescuer to start the rescue.

[0069] As an example, the calculated amount of foam fire extinguishing agent can be sent to a terminal device carried by rescuers via a network or wireless communication, so that the rescuers know the amount of fire extinguishing agent required and are informed to start the rescue mission.

[0070] In some embodiments, a position determination model can be used to output multiple recommended fire-fighting positions, and then the fire-fighting angle of each of the multiple recommended fire-fighting positions can be determined based on the angle determination model. The multiple recommended fire-fighting positions indicate where rescuers can use foam fire extinguishing agents to extinguish fires. The fire-fighting angle of each recommended fire-fighting position represents the angle between the flame and the ground when extinguishing the fire at this fire-fighting position. The position determination model and the angle determination model are artificial neural network models. The inputs of the position determination model are the flame size after the average rescue time, the fire area, the internal location layout information, the initial flame information, the wind speed in the fire area, and the rescuer information. The output of the position determination model is multiple recommended fire-fighting positions. The position determination model can comprehensively consider the fire area, the internal location layout information, the initial flame information, the wind speed in the fire area, and the rescuer information to output multiple recommended fire-fighting positions. Rescuers can perform firefighting operations based on multiple recommended firefighting positions. The angle determination model inputs include the recommended firefighting positions, the flame size after the average rescue time, the fire area, the internal layout information, the initial flame information, the wind speed in the fire area, and the rescuer information. The angle determination model outputs the firefighting angle for each recommended firefighting position. This allows rescuers to adjust the firefighting angle based on the output of the angle determination model for more effective firefighting. By using different firefighting positions and angles, flames can be extinguished more accurately, reducing the risk of fire spread.

[0071] Based on the same inventive concept, Figure 2 A schematic diagram of a fire management system based on big data processing provided by an embodiment of the present invention includes:

[0072] An acquisition module 21 is used to acquire surveillance videos of places within the enterprise, temperature sensor data of places within the enterprise, and smoke detector data of places within the enterprise;

[0073] a fire judgment module 22 for determining whether a fire has occurred using a judgment model based on surveillance video of the enterprise's internal premises, temperature sensor data of the enterprise's internal premises, and smoke detector data of the enterprise's internal premises;

[0074] An information processing module 23 is configured to, if a fire is determined to have occurred, use an information processing model to process the surveillance video of the enterprise's internal premises to determine the fire area, internal premises layout information, initial flame information, and wind speed in the fire area;

[0075] a rescue time determination module 24 for determining an average rescue time of rescuers using a rescue time determination model based on the fire area, the location of the foam extinguishing agent, and rescuer information, wherein the rescuer information includes the location of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers;

[0076] a flame size determination module 25 for determining the flame size after the average rescue time using a flame information determination model based on the average rescue time of the rescuers, the fire area, the internal location layout information, the initial flame information, and the wind speed of the fire area;

[0077] A dosage determination module 26 is used to determine the dosage of the foam fire extinguishing agent based on the flame size and foam fire extinguishing agent information after the average rescue time;

[0078] The sending module 27 is used to send the amount of the foam fire extinguishing agent to the terminal of the rescuer and notify the rescuer to start the rescue.

Claims

1. A fire management method based on big data processing, characterized in that: include: Obtain surveillance videos, temperature sensor data, and smoke detector data from within the enterprise; Determining whether a fire has occurred using a judgment model based on surveillance video of the enterprise's internal premises, temperature sensor data of the enterprise's internal premises, and smoke detector data of the enterprise's internal premises; If a fire is determined to have occurred, the surveillance video of the enterprise's internal premises is processed using an information processing model to determine the fire area, internal premises layout information, initial flame information, and wind speed in the fire area; Determining an average rescue time for rescuers using a rescue time determination model based on the fire area, the location of the foam fire extinguishing agent, and rescuer information, wherein the rescuer information includes the location of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers; Determining the flame size after the average rescue time using a flame information determination model based on the average rescue time of the rescue personnel, the fire area, the internal location layout information, the initial flame information, and the wind speed in the fire area, wherein the internal location layout information includes the total area of the internal location, the height of the internal location, the floor material of the internal location, and information about multiple devices in the internal location; and the initial flame information includes flame shape, flame size, flame horizontal growth rate, and flame vertical growth rate; Determining the amount of foam fire extinguishing agent to be used based on the flame size and foam fire extinguishing agent information after the average rescue time; Sending the amount of the foam fire extinguishing agent to the rescuer's terminal and notifying the rescuer to start the rescue; The method further includes: if it is determined that no fire has occurred, using the judgment model again after an interval of ten minutes to determine whether a fire has occurred.

2. The fire management method based on big data processing according to claim 1, characterized in that: The information processing model is a gated loop unit, the input of the information processing model is the surveillance video of the internal premises of the enterprise, and the output of the information processing model is the fire area, the internal premises layout information, the initial flame information, and the wind speed of the fire area.

3. The fire management method based on big data processing according to claim 2, characterized in that: The information processing model includes a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model. The input of the global processing sub-model is the surveillance video of the internal premises of the enterprise, and the output of the global processing sub-model is the segmented video of the fire area and the internal premises layout information. The input of the flame processing sub-model is the segmented video of the fire area, and the output of the flame processing sub-model is the fire area and the initial flame information. The input of the wind speed determination sub-model is the segmented video of the fire area and the initial flame information, and the output of the wind speed determination sub-model is the wind speed in the fire area.

4. A fire management system based on big data processing, characterized in that: include: An acquisition module is used to acquire surveillance videos of internal places of the enterprise, temperature sensor data of internal places of the enterprise, and smoke detector data of internal places of the enterprise; a fire judgment module, configured to determine whether a fire has occurred using a judgment model based on surveillance video of the enterprise's internal premises, temperature sensor data of the enterprise's internal premises, and smoke detector data of the enterprise's internal premises; an information processing module configured to, if a fire is determined to have occurred, use an information processing model to process surveillance video of the enterprise's internal premises to determine the fire area, internal premises layout information, initial flame information, and wind speed in the fire area, wherein the internal premises layout information includes the total area of the internal premises, the height of the internal premises, the floor material of the internal premises, and information about multiple pieces of equipment in the internal premises; and the initial flame information includes flame shape, flame size, horizontal flame growth rate, and vertical flame growth rate; a rescue time determination module, configured to determine an average rescue time for rescuers using a rescue time determination model based on the fire area, the location of the foam fire extinguishing agent, and rescuer information, wherein the rescuer information includes the location of each rescuer, the age of each rescuer, the running speed of each rescuer, and the number of rescuers; a flame size determination module, configured to determine the flame size after the average rescue time using a flame information determination model based on the average rescue time of the rescuers, the fire area, the internal location layout information, the initial flame information, and the wind speed of the fire area; A dosage determination module is used to determine the dosage of the foam fire extinguishing agent based on the flame size and foam fire extinguishing agent information after the average rescue time; A sending module, used for sending the amount of the foam fire extinguishing agent to the rescuer's terminal and notifying the rescuer to start the rescue; The system is further configured to: if it is determined that no fire has occurred, use the judgment model again after an interval of ten minutes to determine whether a fire has occurred.

5. The fire management system based on big data processing according to claim 4, characterized in that: The information processing model is a gated loop unit, the input of the information processing model is the surveillance video of the internal premises of the enterprise, and the output of the information processing model is the fire area, the internal premises layout information, the initial flame information, and the wind speed of the fire area.

6. The fire management system based on big data processing according to claim 5, characterized in that: The information processing model includes a global processing sub-model, a flame processing sub-model, and a wind speed determination sub-model. The input of the global processing sub-model is the surveillance video of the internal premises of the enterprise, and the output of the global processing sub-model is the segmented video of the fire area and the internal premises layout information. The input of the flame processing sub-model is the segmented video of the fire area, and the output of the flame processing sub-model is the fire area and the initial flame information. The input of the wind speed determination sub-model is the segmented video of the fire area and the initial flame information, and the output of the wind speed determination sub-model is the wind speed in the fire area.

Images