A fire emergency management system and fire prediction method

Abstract

The present application provides a fire emergency management system and fire prediction method, which monitors fires in a target area and collects fire images of the target area; performs regional segmentation on the fire image to obtain multiple flame areas in the target area, determines the concentration gradient of the smoke in each flame area based on the color distribution of the smoke in each flame area, and determines the smoke spread characteristics in the target area based on the concentration gradient of all smoke; obtains fire environment data during the occurrence of the fire in the target area, and determines the heat transfer relationship generated when the smoke flows in the target area based on the fire evolution model combined with the fire environment data; predicts the thermal runaway state of the fire through the heat transfer relationship and the smoke spread characteristics, and then obtains a confident fire situation in the target area; the fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area. Using the solution of the present application, a confident early warning of fire spread can be achieved in complex fire scenarios.

Description

Technical Field

[0001] The present application relates to the field of emergency management technology, and more specifically, to a fire emergency management system and a fire prediction method. Background Art

[0002] With the acceleration of urbanization and the development of industrialization, the frequent occurrence of fire accidents poses a significant threat to people's lives and property. In recent years, with the development of technologies such as computer vision, the Internet of Things, big data analysis, and artificial intelligence, fire management has gradually moved towards intelligent and automated directions. Image processing, sensor technology, and data fusion have provided new solutions for fire emergency management.

[0003] However, in the existing technology, traditional fire prediction methods mostly rely on manual inspections, on-site feedback and simple alarm systems. These methods usually have shortcomings such as information lag, slow response speed, and difficulty in comprehensive coverage. They are also prone to errors when dealing with large-scale fires. Especially for the prediction of fire spread and the identification of dangerous areas, existing technologies often rely on manual experience and on-site perception, lack systematic and intelligent data support and decision-making basis, and lack the ability to accurately evaluate based on real-time dynamic data of fires. Since the spread of fire is usually affected by complex environmental factors such as wind speed, temperature, humidity, etc., the prediction results of fire spread are inaccurate. Especially for complex fire scenes, existing technologies find it difficult to provide real-time and accurate fire spread risk prediction and emergency management strategies; therefore, how to achieve confidence warning of fire spread in complex fire scenes has become a difficult problem facing the industry. Summary of the Invention

[0004] The present application provides a fire emergency management system and a fire prediction method, which can realize a confidence warning of fire spread in complex fire scenarios.

[0005] In a first aspect, the present application provides a fire prediction method for use in a fire emergency management system for fire emergency management, the method comprising the following steps:

[0006] Conduct fire monitoring in the target area and collect fire images in the target area;

[0007] performing region segmentation on the fire image to obtain multiple flame regions in a target area, determining a concentration gradient of smoke in each flame region based on a color distribution of smoke in each flame region, and determining a smoke spread characteristic in the target area based on the concentration gradients of all smoke;

[0008] Obtaining fire environment data during a fire in a target area, and determining the heat transfer relationship generated by smoke flow in the target area based on a pre-built fire evolution model combined with the fire environment data;

[0009] Predicting the thermal runaway state of the fire in the target area by using the heat transfer relationship and the smoke spread characteristics, thereby obtaining a confident fire situation in the target area;

[0010] The fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area.

[0011] In some embodiments, performing region segmentation on the fire image to obtain multiple flame regions in the target area specifically includes:

[0012] Binarizing the fire image to obtain a binary image;

[0013] A plurality of flame regions in a target area are segmented from the fire image based on the binary image.

[0014] In some embodiments, determining the concentration gradient of smoke in each flame region based on the color distribution of smoke in each flame region specifically includes:

[0015] Performing color distribution analysis on the smoke in each flame area to obtain a change trend of the smoke in each flame area;

[0016] The concentration gradient of smoke in the target area is determined by all the changing trends.

[0017] In some embodiments, performing color distribution analysis on the smoke in each flame region to obtain a change trend of the smoke in each flame region specifically includes:

[0018] Convert each flame area from RGB color space to HSV color space;

[0019] Extracting the color distribution of smoke in each flame region based on the HSV color space of each flame region;

[0020] The change trend of the smoke in each flame area is determined by the color distribution of the smoke in each flame area.

[0021] In some embodiments, determining the heat transfer relationship generated by smoke flow in the target area based on the pre-built fire evolution model and the fire environment data specifically includes:

[0022] performing time alignment on the fire environment data to obtain a time aligned sequence;

[0023] determining the heat distribution of smoke flow during fire spread in the target area based on a pre-built fire evolution model and the time-aligned sequence;

[0024] The heat transfer relationship generated when the smoke flows in the target area is determined by the heat distribution of the smoke flow.

[0025] In some embodiments, predicting the thermal runaway state of a fire in a target area by using the heat transfer relationship and the smoke spread characteristics, and then obtaining a confident fire situation in the target area specifically includes:

[0026] Determine the heat energy release at each location in the target area based on the distribution of combustibles in the target area;

[0027] Predicting a thermal runaway state of a fire in a target area based on the smoke spread characteristics, the respective heat energy release amounts, and the heat transfer relationship;

[0028] A confident fire situation in the target area is generated based on the thermal runaway state.

[0029] In some embodiments, generating a confident fire situation in the target area based on the thermal runaway state specifically includes:

[0030] Initialize a prediction model;

[0031] Using the thermal runaway state as an input to the prediction model;

[0032] Outputting the risk probability of fire spread at each location in the target area through the prediction model;

[0033] Generate a confident fire situation in the target area based on the risk probability of all fires spreading.

[0034] In a second aspect, the present application provides a fire emergency management system, which includes a fire prediction unit, wherein the fire prediction unit includes:

[0035] An acquisition module is used to monitor fires in a target area and acquire fire images of the target area;

[0036] a processing module configured to segment the fire image into multiple flame regions in a target area, determine a concentration gradient of the smoke in each flame region based on a color distribution of the smoke in each flame region, and determine a smoke spread characteristic in the target area based on the concentration gradients of all the smoke;

[0037] The processing module is further configured to obtain fire environment data during the occurrence of a fire in the target area, and determine the heat transfer relationship generated by smoke flow in the target area based on a pre-built fire evolution model combined with the fire environment data;

[0038] The processing module is further configured to predict a thermal runaway state of a fire in a target area by using the heat transfer relationship and the smoke spread characteristics, thereby obtaining a reliable fire situation in the target area;

[0039] The execution module is used to control the fire alarm to send a fire alarm to the fire command center based on the confident fire situation in the target area.

[0040] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned fire prediction method.

[0041] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions or codes. When the instructions or codes are run on a computer, the computer implements the above-mentioned fire prediction method when executing the instructions or codes.

[0042] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0043] In the present application, fire monitoring is performed on a target area, and a fire image of the target area is collected; the fire image is segmented to obtain multiple flame areas in the target area, the concentration gradient of the smoke in each flame area is determined based on the color distribution of the smoke in each flame area, and the smoke spread characteristics in the target area are determined based on the concentration gradient of all smoke; fire environment data during the occurrence of the fire in the target area is obtained, and the heat transfer relationship generated when the smoke flows in the target area is determined based on a pre-built fire evolution model combined with the fire environment data; the thermal runaway state of the fire in the target area is predicted based on the heat transfer relationship and the smoke spread characteristics, and a confident fire situation in the target area is obtained; the fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area.

[0044] It can be seen that in this application, firstly, the smoke spread characteristics in the target area are determined based on the concentration gradient of all smoke, which can accurately capture the degree of smoke spread during the fire process, so as to accurately predict the evolution process of fire spread in the target area in complex fire scenes; secondly, the heat transfer relationship generated when the smoke flows in the target area is determined based on the pre-constructed fire evolution model combined with the fire environment data, which can provide an important basis for the heat propagation of fire spread in the target area, judge the fire intensity and potential expansion risk in the target area, so as to help identify high-temperature dangerous areas, thereby effectively predicting the spread range and speed of the fire in the target area, and avoiding missing dangerous areas; then, the thermal runaway state of the fire in the target area is predicted by the heat transfer relationship and the smoke spread characteristics, which can realize a multi-dimensional comprehensive evaluation of fire spread prediction and provide a more comprehensive prediction of fire spread path and risk. The fire alarm system can effectively enhance the accuracy and reliability of fire spread warning, thereby providing more accurate warning and decision-making support for emergency management in complex fire scenarios. Subsequently, based on the thermal runaway state, a confident fire situation in the target area is generated, which can quantitatively describe the risk of fire spreading in the target area, provide an accurate fire risk probability, and then clearly present the probability and risk level of fire spread, so that the fire command center can grasp the dynamic development of the fire in the target area in real time. Finally, the fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area. The fire command center can accurately identify the key areas of fire spread in the target area, formulate targeted emergency plans, and reasonably allocate fire resources, thereby improving the efficiency and effectiveness of fire emergency management. In summary, the scheme can realize the confidence warning of fire spread in complex fire scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0046] Figure 1 is an exemplary flow chart of a fire prediction method according to some embodiments of the present application;

[0047] Figure 2 is an exemplary flow chart for determining a change trend of smoke according to some embodiments of the present application;

[0048] Figure 3 is an exemplary flow chart for determining a fire situation according to some embodiments of the present application;

[0049] Figure 4 is a structural diagram of a fire prediction unit according to some embodiments of the present application;

[0050] Figure 5 It is a structural diagram of a computer device for implementing a fire prediction method according to some embodiments of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0052] refer to Figure 1 This figure is an exemplary flow chart of a fire prediction method according to some embodiments of the present application, which is used in a fire emergency management system for fire emergency management. The fire prediction method 100 mainly includes the following steps:

[0053] In step 101, a target area is monitored for fire, and a fire image of the target area is collected.

[0054] In a specific implementation, a fire detector with an integrated camera, such as a Nest Protect smoke detector (with an integrated camera function), is used to monitor the target area for fire. When the fire detector detects a fire signal, the optical camera in the fire detector is used to collect fire images of the target area at a fixed frame rate (e.g., 5 frames per second). The fire detector with an integrated camera is an integrated device that combines an optical camera, a smoke sensor, and a temperature sensor. In other embodiments, other methods may also be used for collection, which is not specifically limited here.

[0055] It should be noted that the fire image in this application refers to an image of the fire situation in the target area.

[0056] In step 102, the fire image is segmented to obtain multiple flame regions in the target area. The concentration gradient of the smoke in each flame area is determined based on the color distribution of the smoke in each flame area. The smoke spread characteristics in the target area are determined based on the concentration gradients of all the smoke.

[0057] In some embodiments, performing region segmentation on the fire image to obtain multiple flame regions in the target area may be achieved by using the following steps:

[0058] Binarizing the fire image to obtain a binary image;

[0059] A plurality of flame regions in a target area are segmented from the fire image based on the binary image.

[0060] In a specific implementation, the fire image is binarized to obtain a binary image, which can be achieved in the following manner: dynamically selecting a threshold value for distinguishing the flame area from the background according to the brightness difference in the fire image through an adaptive threshold method, and then setting the pixels above the threshold in the fire image to 255 (representing the flame area), and the pixels below the threshold to 0 (representing the background area), thereby obtaining a binary image. In other embodiments, other methods can also be used for implementation, which are not limited here; based on the binary image, multiple flame areas of the target area are segmented from the fire image, which can be achieved in the following manner: using a connected domain analysis algorithm (such as a depth-first search or a breadth-first search) to obtain a plurality of flame areas. Search), traverse each pixel in the binary image, check whether the adjacent pixels around it (usually in 8 directions) are in the same area (that is, the pixel value is 255), and if so, connect the area of ​​the adjacent pixels with the area of ​​the current pixel until all adjacent pixels are visited, and then mark all pixels belonging to the same area as a connected domain, and finally segment multiple flame areas of the target area from the fire image based on the positions of each marked connected domain in the fire image, wherein each flame area is the area part of the connected domain in the binary image at the corresponding position in the fire image. In other embodiments, other methods can also be used for determination, which is not limited here.

[0061] It should be noted that the binary image in this application refers to a black and white image converted from a fire image; in addition, the flame area refers to the area of ​​the flame position in the fire image.

[0062] In some embodiments, determining the concentration gradient of smoke in each flame region based on the color distribution of smoke in each flame region can be achieved by using the following steps:

[0063] Performing color distribution analysis on the smoke in each flame area to obtain a change trend of the smoke in each flame area;

[0064] The concentration gradient of smoke in the target area is determined by all the changing trends.

[0065] In some embodiments, reference Figure 2 As shown in FIG. 1 , this figure is an exemplary flow chart for determining a change trend of smoke in some embodiments of the present application. In this embodiment, color distribution analysis of the smoke in each flame region is performed to obtain a change trend of the smoke in each flame region. The following steps can be used to achieve this:

[0066] First, in step 1021, each flame region is converted from the RGB color space to the HSV color space;

[0067] Next, in step 1022, the color distribution of the smoke in each flame region is extracted based on the HSV color space of each flame region;

[0068] Finally, in step 1023 , the change trend of the smoke in each flame area is determined based on the color distribution of the smoke in each flame area.

[0069] In specific implementation, converting each flame area from RGB color space to HSV color space can be achieved in the following manner, namely: converting each flame area from RGB image to HSV image through a color space conversion algorithm (such as cv2.cvtColor() function in OpenCV library). In other embodiments, other methods can also be used for determination, which is not limited here; extracting the color distribution of smoke in each flame area based on the HSV color space of each flame area can be achieved in the following manner, namely: using the color features and texture features of smoke in historical experience data through a machine learning algorithm to identify the smoke in each flame area, and then calculating the hue (used to reflect the color type of smoke), saturation (used to reflect the purity or vividness of the smoke color), and brightness of the smoke in each flame area based on the HSV color space of each flame area. (used to reflect the brightness of the smoke color), and the set of results calculated by the machine learning algorithm is used as the color distribution of the smoke in each flame area. In other embodiments, other methods can also be used for determination, which is not limited here; determining the change trend of the smoke in each flame area through the color distribution of the smoke in each flame area can be achieved in the following manner, namely: using an existing evaluation algorithm (for example: cross-validation, genetic algorithm and ensemble learning, etc.) based on the color distribution (i.e., hue, saturation, brightness) of the smoke in each flame area to evaluate the change trend of the smoke in each flame area (such as color change, area change, brightness change, etc.), and then using the set of all trend values ​​corresponding to each flame area obtained by the evaluation as the change trend of the smoke in each flame area. In other embodiments, other methods can also be used for determination, which is not limited here.

[0070] It should be noted that the changing trend of smoke in the present application represents the changing situation of smoke in the flame area, which can reflect the changing process of expansion, concentration, and thinning of fire smoke in the target area; the HSV color space represents the color space used for smoke analysis in the flame area; the color distribution of smoke represents the distribution of smoke color in the flame area.

[0071] In addition, in a specific implementation, determining the concentration gradient of smoke in the target area from all the changing trends can be achieved in the following manner, namely: a deep learning framework can be used to construct a convolutional neural network model, and the convolutional neural network model is used to train a large number of image data sets containing smoke changing trends to learn the quantitative relationship between the changing trend of fire smoke (such as color change, area change, brightness change and other trend values) and the smoke concentration change rate, so as to input all the changing trends in this embodiment into the convolutional neural network model, and finally use the calculation result of the convolutional neural network model as the concentration gradient of smoke in the target area. In other embodiments, other methods can also be used for determination, which is not limited here.

[0072] It should be noted that the smoke spread characteristic in the present application represents the degree of smoke spread in the target area. The larger the value corresponding to the smoke spread characteristic, the higher the degree of smoke spread in the target area. The smaller the value corresponding to the smoke spread characteristic, the lower the degree of smoke spread in the target area. As a preferred embodiment, the smoke spread characteristic in the target area can be determined based on the concentration gradient of all smoke in the following manner, that is, the average value of the concentration gradient of all smoke can be used as the smoke spread characteristic in the target area. In other embodiments, other methods can also be used for determination, which is not limited here.

[0073] In step 103, fire environment data of the target area during the fire occurrence process is obtained, and the heat transfer relationship generated by smoke flow in the target area is determined based on the pre-built fire evolution model combined with the fire environment data.

[0074] In specific implementation, the fire environment data of the target area during the fire can be obtained in the following ways, namely: through integrated sensors (including environmental temperature and humidity sensors, oxygen concentration sensors, infrared sensors, smoke sensors, wind speed sensors and other sensors), the environmental temperature and humidity, oxygen concentration, flame temperature, smoke concentration, wind speed and other data values ​​of the target area during the fire are collected in real time according to the specified collection frequency, and machine learning methods (such as isolation forest, support vector machine, neural network, etc.) are used to perform trustworthy processing on all collected data values. The trustworthy processing may include: eliminating abnormal data, interpolation and correction, etc., to truly reflect the target area. The entire process of the fire in the domain is finally obtained by the trusted processing of all the above data values ​​as the fire environment data during the fire in the target area. The specified acquisition frequency can be set according to specific needs and the fire situation at the target area. For example, in the early stage of the fire, the fire environment data changes rapidly, so a higher acquisition frequency needs to be set, such as once per second (5Hz). In the later stage of the fire, the change of the fire environment data gradually stabilizes, and a lower acquisition frequency can be set, such as once per minute (0.016Hz). In other embodiments, other methods can also be used to obtain data, which is not specifically limited here.

[0075] It should be noted that the fire environment data in this application represents a collection of physical quantities related to the development of a fire. Therefore, the fire environment data directly affects the spread trend of the fire in the target area.

[0076] In some embodiments, determining the heat transfer relationship generated by smoke flow in the target area based on the pre-built fire evolution model combined with the fire environment data can be achieved by the following steps:

[0077] performing time alignment on the fire environment data to obtain a time aligned sequence;

[0078] determining the heat distribution of smoke flow during fire spread in the target area based on a pre-built fire evolution model and the time-aligned sequence;

[0079] The heat transfer relationship generated when the smoke flows in the target area is determined by the heat distribution of the smoke flow.

[0080] In a specific implementation, the fire environment data is time-aligned to obtain a time-aligned sequence, which can be achieved in the following manner: all data values ​​in the fire environment data can be aligned according to the corresponding collection time (i.e., data values ​​with the same collection time are synchronized) through an existing time series analysis method (e.g., Pandas in the Python library), so that all data values ​​are aligned at the same collection time point, thereby obtaining a time-aligned sequence. In other embodiments, other methods can also be used for implementation, which is not limited here. It should be noted that the time-aligned sequence in this application represents a set of fire environment data after time alignment.

[0081] It should be noted that the fire evolution model in this application refers to a machine learning model that can describe the dynamic changes of a fire in a fire area, wherein the fire dynamic changes represent the distribution of heat as the fire spreads in the fire area as the smoke flows. As a preferred embodiment, the fire evolution model can be constructed in the following manner: first, a fire environment data record library of a historical fire area, which contains time-aligned sequences of data values ​​such as ambient temperature and humidity, oxygen concentration, flame temperature, smoke concentration, and wind speed collected during the fire, is used as samples for the machine learning model. The samples are then divided into a training set and a test set, for example, 80% of the samples are used as the training set and 20% of the samples are used as the test set. Then, the samples in the training set are used as input to the machine learning model, and the model is trained using a supervised learning algorithm. The fire dynamic changes obtained through training (i.e., the distribution of heat as the smoke flows in the fire area as the fire spreads) are used as output of the machine learning model. The obtained fire dynamic changes are further verified using samples in the test set, and the verified model is finally used as the fire evolution model. In other embodiments, other methods may also be used for construction, which are not limited here.

[0082] In a specific implementation, determining the heat distribution of smoke flow during fire spread in a target area based on a pre-constructed fire evolution model and the time-aligned sequence can be achieved in the following manner: first, inputting the time-aligned sequence into the pre-constructed fire evolution model, thereby outputting the dynamic changes of the fire in the target area through the fire evolution model. Second, the heat value of the smoke flow during fire spread at each location in the target area can be calculated using a heat conduction model (such as the Fourier heat conduction equation) or a smoke thermodynamic model in combination with the dynamic changes of the fire. Ultimately, the set of all heat values ​​is used as the heat distribution of the smoke flow during fire spread in the target area. In other embodiments, other methods may be used for determination, which are not limited here. Determining the heat transfer relationship generated by smoke flow in the target area through the heat distribution of the smoke flow can be achieved in the following manner: first, calculating the absolute difference in heat values ​​between different locations in the heat distribution of the smoke flow. Ultimately, the set of all absolute differences is used as the heat transfer relationship generated by smoke flow in the target area. In other embodiments, other methods may be used for determination, which are not limited here.

[0083] It should be noted that the heat transfer relationship in the present application represents the intensity of the change in heat transferred by the smoke flow at the corresponding position in the target area. The larger the value corresponding to the heat transfer relationship, the greater the intensity of the change in heat transferred by the smoke flow at the corresponding position in the target area. The smaller the value corresponding to the heat transfer relationship, the smaller the intensity of the change in heat transferred by the smoke flow at the corresponding position in the target area. Therefore, the heat transfer relationship can intuitively reflect the degree of heat propagation in the fire area, provide key heat change information for the prediction of fire spread, and assist in the prediction and analysis of the fire spread process. In addition, the heat distribution of the smoke flow in the present application represents the heat situation of the smoke flow propagation during the fire spread in the target area.

[0084] In step 104, the thermal runaway state of the fire in the target area is predicted based on the heat transfer relationship and the smoke spread characteristics, thereby obtaining a confident fire situation in the target area.

[0085] In some embodiments, the following steps may be used to predict the thermal runaway state of a fire in a target area using the heat transfer relationship and the smoke spread characteristics, and then obtain a confident fire situation in the target area:

[0086] Determine the heat energy release at each location in the target area based on the distribution of combustibles in the target area;

[0087] Predicting a thermal runaway state of a fire in a target area based on the smoke spread characteristics, the respective heat energy release amounts, and the heat transfer relationship;

[0088] A confident fire situation in the target area is generated based on the thermal runaway state.

[0089] In specific implementation, the following method can be used to determine the amount of heat energy released at each position in the target area through the distribution of combustibles in the target area, namely: a drone equipped with a multispectral camera can be used in combination with a geographic information system and remote sensing image processing technology to collect information such as the type, quantity and density of combustibles (such as wood, oil, paper, etc.) at each position in the target area (i.e., the distribution of combustibles in the target area). Then, the intensity of infrared radiation released at each position in the target area can be determined based on the distribution of combustibles in the target area through thermoluminescence imaging, and the amount of heat energy released at each position in the target area can be calculated in combination with the physical model of thermal radiation. In other embodiments, other methods can also be used for determination, which is not limited here; The prediction of the thermal runaway state of the fire in the target area based on the smoke spread characteristics, the respective heat energy release amounts and the heat transfer relationship can be achieved in the following manner, namely: for each position in the target area, the influence weights of the smoke spread characteristics, the heat energy release amounts and the heat transfer relationship on the degree of thermal runaway of the fire are preset using historical experience. In other embodiments, other methods may also be used to set the influence weights, which are not limited here. The weighted average of the calculated smoke spread characteristics, the heat energy release amounts at the corresponding positions and the absolute differences at the corresponding positions in the heat transfer relationship is then used as the out-of-control value of the fire at the corresponding positions. The out-of-control value of the fire at each position in the target area can be obtained in the above manner, and finally the set of all out-of-control values ​​is used as the thermal runaway state of the fire in the target area.

[0090] It should be noted that the heat energy release amount in the present application represents the heat energy release capacity at the corresponding position in the target area. The larger the heat energy release amount, the greater the heat energy release capacity at the corresponding position in the target area, and the smaller the heat energy release amount, the smaller the heat energy release capacity at the corresponding position in the target area. In addition, the thermal runaway state in the present application represents the degree of uncontrolled fire spread during the fire spread in the target area. The larger the uncontrollable value in the thermal runaway state, the greater the degree of uncontrolled fire spread during the fire spread in the target area. Conversely, the smaller the uncontrollable value in the thermal runaway state, the smaller the degree of uncontrolled fire spread during the fire spread in the target area. Therefore, the thermal runaway state can predict the expansion path and risk of the fire in the target area, and provide a key basis for fire prediction, spread path analysis and risk assessment.

[0091] In some embodiments, generating a confident fire situation in a target area based on the thermal runaway state may be achieved by the following steps:

[0092] Initialize a prediction model;

[0093] Using the thermal runaway state as an input to the prediction model;

[0094] Outputting the risk probability of fire spread at each location in the target area through the prediction model;

[0095] Generate a confident fire situation in the target area based on the risk probability of all fires spreading.

[0096] It should be noted that the prediction model in this application is a deep learning model for predicting the fire spread path and spread risk. The prediction model can automatically learn the complex patterns and laws in the fire spread process based on the input thermal runaway state, and then accurately predict the risk probability of fire spread at different locations in the target area; the prediction model learns the mapping relationship between thermal runaway state and fire situation in historical data through a large amount of fire experimental data and actual case analysis, and constructs a function that can accurately predict the fire situation for the newly input thermal runaway state. Among them, the algorithm framework of the prediction model can adopt convolutional neural network, recurrent neural network, etc. In other embodiments, the algorithm framework of the prediction model can also adopt other algorithm structures, which are not limited here.

[0097] In specific implementation, the output of the risk probability of fire spread at each position in the target area by the prediction model can be achieved in the following manner, namely: the prediction model automatically learns the complex patterns and laws in the fire spread process according to the input thermal runaway state through the back propagation algorithm or the gradient descent method, and then calculates the risk probability of fire spread at each position in the target area. The value of the risk probability is between 0 and 1. In other embodiments, other methods can also be used for implementation, which is not limited here; generating a confident fire situation in the target area based on the risk probability of all fire spreads can be achieved in the following manner, namely: first, the risk probability corresponding to the risk level of fire spread can be calculated based on actual conditions, historical data analysis or expert experience. The risk probability range is set, where the risk level of fire spread is divided into high risk, medium risk, and low risk. For example, the risk probability range corresponding to high risk can be set to (0.7, 1], the risk probability range corresponding to medium risk can be set to [0.3, 0.7], and the risk probability range corresponding to low risk can be set to [0, 0.3). Then, according to the above risk probability range and the risk probability of fire spread at each location in the target area, the risk level of fire spread at each location in the target area is marked. For example, the high-risk area can be marked in red, the medium-risk area can be marked in yellow, and the low-risk area can be marked in green. Finally, the location map generated after the target area is marked is used as the confident fire situation in the target area for reference. Figure 3 As shown, this figure is an exemplary flow chart for determining the fire situation in some embodiments of the present application. Other methods can also be used to implement it in other embodiments, which is not limited here.

[0098] It should be noted that the risk probability in the present application indicates the risk level of fire spread at the corresponding position in the target area. The greater the risk probability, the greater the risk of fire spread at the corresponding position in the target area. Conversely, the smaller the risk probability, the smaller the risk of fire spread at the corresponding position in the target area. In addition, the fire situation in the present application indicates the risk level of fire spread at different positions in the target area, where the risk level is divided into: high risk, medium risk, and low risk, which will not be repeated here.

[0099] In step 105, the fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area.

[0100] In specific implementation, the fire alarm can send a fire alarm to the fire command center based on the confident fire situation in the target area in the following way: the built-in communication module of the fire alarm (such as Wi-Fi, cellular network or dedicated emergency communication network) generates a fire alarm signal based on the red area (i.e. high-risk area), yellow area (i.e. medium-risk area) and green area (i.e. low-risk area) in the confident fire situation in the target area. The fire alarm signal includes the area location, alarm time and risk level, etc., and then the fire alarm sends the fire alarm signal to the fire command center through wireless communication technology.

[0101] As a preferred embodiment, after the fire command center receives the fire alarm signal, it immediately initiates emergency response measures for the red area (i.e., high-risk area) in the confident fire situation in the target area, such as dispatching additional fire brigades, cutting off flammable materials, evacuating people, etc.; for the yellow area (i.e., medium-risk area) in the confident fire situation in the target area, it is necessary to prepare emergency plans, increase monitoring and prevention measures, and ensure rapid response; for the green area (i.e., low-risk area) in the confident fire situation in the target area, no special intervention is required, but continuous monitoring should be carried out to prevent the sudden spread of fire; through the above method, the fire command center can intuitively understand the risk distribution of fire spread in the target area, better formulate corresponding prevention and control and evacuation measures, so as to quickly respond and dispatch fire resources, thereby realizing fire emergency management.

[0102] In addition, in another aspect of the present application, in some embodiments, the present application provides a fire emergency management system, which includes a fire prediction unit, referring to Figure 4 , which is a schematic diagram of the structure of a fire prediction unit according to some embodiments of the present application. The fire prediction unit 400 includes: a collection module 401, a processing module 402, and an execution module 403, which are described as follows:

[0103] Acquisition module 401, in this application, acquisition module 401 is mainly used to monitor fire in the target area and acquire fire images of the target area;

[0104] Processing module 402, in this application, is mainly used to segment the fire image to obtain multiple flame regions in the target area, determine the concentration gradient of the smoke in each flame region based on the color distribution of the smoke in each flame region, and determine the smoke spread characteristics in the target area based on the concentration gradients of all the smoke;

[0105] The processing module 402 in the present application is further configured to obtain fire environment data during the occurrence of a fire in the target area, and determine the heat transfer relationship generated by smoke flow in the target area based on a pre-built fire evolution model combined with the fire environment data;

[0106] The processing module 402 in the present application is further configured to predict the thermal runaway state of the fire in the target area by using the heat transfer relationship and the smoke spread characteristics, thereby obtaining a reliable fire situation in the target area;

[0107] Execution module 403, in this application, execution module 403 is mainly used to control the fire alarm to send a fire alarm to the fire command center based on the confident fire situation in the target area.

[0108] The above describes in detail the examples of the fire emergency management system and fire prediction method provided by the embodiments of the present application. It can be understood that in order to realize the above functions, the corresponding device includes hardware structures and / or software modules corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0109] In some embodiments, the present application also provides a computer device, which includes a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned fire prediction method.

[0110] In some embodiments, reference Figure 5 The dotted line in the figure indicates that the unit or module is optional. The figure is a schematic diagram of the structure of the computer device that implements the fire prediction method of the present application. The fire prediction method in the above embodiment can be Figure 5The computer device 500 is implemented as shown, and the computer device 500 includes at least one processor 501, a memory 502 and at least one communication unit 505. The computer device 500 can be a terminal device, a server or a chip.

[0111] The processor 501 may be a general-purpose processor or a special-purpose processor. For example, the processor 501 may be a central processing unit (CPU), which may be used to control the computer device 500, execute software programs, and process data from the software programs. The computer device 500 may also include a communication unit 505 for inputting (receiving) and outputting (transmitting) signals.

[0112] For example, the computer device 500 may be a chip, the communication unit 505 may be an input and / or output circuit of the chip, or the communication unit 505 may be a communication interface of the chip, and the chip may be a component of a terminal device, a network device, or other device.

[0113] For another example, the computer device 500 may be a terminal device or a server, and the communication unit 505 may be a transceiver of the terminal device or the server, or the communication unit 505 may be a transceiver circuit of the terminal device or the server.

[0114] The computer device 500 may include one or more memories 502, on which a program 504 is stored. The program 504 can be executed by the processor 501 to generate instructions 503, so that the processor 501 executes the method described in the above method embodiment according to the instructions 503. Optionally, data (such as a target audit model) can also be stored in the memory 502. Optionally, the processor 501 can also read data stored in the memory 502. The data can be stored at the same storage address as the program 504, or at a different storage address from the program 504.

[0115] The processor 501 and the memory 502 may be provided separately or integrated together, for example, integrated on a system on chip (SOC) of a terminal device.

[0116] It should be understood that each step of the above method embodiment can be completed by a hardware-based logic circuit or software-based instructions in the processor 501. The processor 501 can be a CPU, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, such as discrete gates, transistor logic devices, or discrete hardware components.

[0117] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0118] For example, in some embodiments, the present application also provides a computer-readable storage medium, in which instructions or codes are stored. When the instructions or codes are run on a computer, the computer implements the above-mentioned fire prediction method when executing.

[0119] In summary, in the fire emergency management system and fire prediction method disclosed in the embodiments of the present application, fire monitoring is performed on the target area, and a fire image of the target area is collected; the fire image is segmented to obtain multiple flame areas in the target area, and the concentration gradient of the smoke in each flame area is determined based on the color distribution of the smoke in each flame area, and the smoke spread characteristics in the target area are determined based on the concentration gradient of all smoke; the fire environment data during the fire in the target area is obtained, and the heat transfer relationship generated when the smoke flows in the target area is determined based on a pre-built fire evolution model combined with the fire environment data; the thermal runaway state of the fire in the target area is predicted by the heat transfer relationship and the smoke spread characteristics, and then a confident fire situation in the target area is obtained; the fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area; and a confident warning of fire spread can be achieved in complex fire scenarios.

[0120] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0121] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A fire prediction method for use in a fire emergency management system for fire emergency management, characterized in that: The method comprises the following steps: Conduct fire monitoring in the target area and collect fire images in the target area; performing region segmentation on the fire image to obtain multiple flame regions in a target region, determining a concentration gradient of smoke in each flame region based on a color distribution of smoke in each flame region, and determining a smoke spread characteristic in the target region based on the concentration gradients of all smoke, wherein the smoke spread characteristic represents a degree of smoke spread in the target region; Obtaining fire environment data during a fire in a target area, and determining a heat transfer relationship generated by smoke flow within the target area based on a pre-established fire evolution model combined with the fire environment data, wherein the heat transfer relationship represents the intensity of the change in heat transferred by the smoke flow at corresponding locations within the target area; Predicting the thermal runaway state of the fire in the target area by using the heat transfer relationship and the smoke spread characteristics, thereby obtaining a confident fire situation in the target area; The fire alarm sends a fire alarm to the fire command center based on the confident fire situation in the target area; The thermal runaway state of the fire in the target area is predicted by using the heat transfer relationship and the smoke spread characteristics, and the confident fire situation in the target area is obtained, specifically including: Determine the heat energy release at each location in the target area based on the distribution of combustibles in the target area; Predicting a thermal runaway state of a fire in a target area based on the smoke spread characteristics, the respective heat energy release amounts, and the heat transfer relationship; generating a confident fire situation in a target area based on the thermal runaway state; The step of determining the concentration gradient of the smoke in each flame region based on the color distribution of the smoke in each flame region specifically includes: Performing color distribution analysis on the smoke in each flame area to obtain a change trend of the smoke in each flame area; The concentration gradient of smoke in the target area is determined by all the changing trends.

2. The method according to claim 1, wherein The fire image is segmented to obtain multiple flame regions in the target area, specifically including: Binarizing the fire image to obtain a binary image; A plurality of flame regions in a target area are segmented from the fire image based on the binary image.

3. The method according to claim 1, wherein The color distribution analysis of the smoke in each flame area is performed to obtain the change trend of the smoke in each flame area, which specifically includes: Convert each flame area from RGB color space to HSV color space; Extracting the color distribution of smoke in each flame region based on the HSV color space of each flame region; The change trend of the smoke in each flame area is determined by the color distribution of the smoke in each flame area.

4. The method according to claim 1, wherein Determining the heat transfer relationship generated by smoke flow in the target area based on the pre-built fire evolution model and the fire environment data specifically includes: performing time alignment on the fire environment data to obtain a time aligned sequence; determining the heat distribution of smoke flow during fire spread in the target area based on a pre-built fire evolution model and the time-aligned sequence; The heat transfer relationship generated when the smoke flows in the target area is determined by the heat distribution of the smoke flow.

5. The method according to claim 1, wherein Generating a confident fire situation in a target area based on the thermal runaway state specifically includes: Initialize a prediction model; Using the thermal runaway state as an input to the prediction model; Outputting the risk probability of fire spread at each location in the target area through the prediction model; Generate a confident fire situation in the target area based on the risk probability of all fires spreading.

6. A fire emergency management system, which uses the method according to any one of claims 1 to 5 to perform fire prediction, the fire emergency management system comprising a fire prediction unit, characterized in that: The fire prediction unit includes: An acquisition module is used to monitor fires in a target area and acquire fire images of the target area; a processing module configured to segment the fire image into multiple flame regions in a target area, determine a concentration gradient of the smoke in each flame region based on a color distribution of the smoke in each flame region, and determine a smoke spread characteristic in the target area based on the concentration gradients of all the smoke; The processing module is further configured to obtain fire environment data during the occurrence of a fire in the target area, and determine the heat transfer relationship generated by smoke flow in the target area based on a pre-built fire evolution model combined with the fire environment data; The processing module is further configured to predict a thermal runaway state of a fire in a target area by using the heat transfer relationship and the smoke spread characteristics, thereby obtaining a reliable fire situation in the target area; The execution module is used to control the fire alarm to send a fire alarm to the fire command center based on the confident fire situation in the target area.

7. A computer device, characterized in that: The computer device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the fire prediction method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions or codes, and when the instructions or codes are executed on a computer, the computer implements the fire prediction method according to any one of claims 1 to 5.

Images