Urban fire control method and device based on fire scene simulation

Abstract

The invention provides an urban fire control method and device based on fire scene simulation, and the method comprises the steps: responding to a received instruction signal for carrying out the fire control of a target region where a fire occurs in a city, and collecting the scene data of the target region through an unmanned plane cluster; based on the scene data, constructing a fire scene twinborn simulation model at least comprising a temperature field and a smoke diffusion vector field in the target area; inputting the fire scene twinborn simulation model into a fire behavior prediction model, and performing fire behavior deduction on a target area based on a temperature field and a smoke diffusion vector field to obtain dangerous area information and evacuation channel information; and based on the dangerous area information and the evacuation channel information, generating a rescue path of the target area in the fire scene twinborn simulation model to determine a fire control strategy of the target area, and performing fire control on the target area according to the fire control strategy. Through the method, the real-time performance, the accuracy, the comprehensiveness and the stability of urban fire control are improved.

Description

Technical Field

[0001] The present application relates to the technical field of urban fire control and management, and in particular to an urban fire control and management method and device based on fire scene simulation. Background Art

[0002] With the acceleration of urbanization, the number of high-rise buildings and complex structures in cities has gradually increased, and the suddenness and destructiveness of urban fires have become increasingly severe. Traditional fire control methods rely on fixed sensors and manual inspections, and have problems such as delayed response and single data dimensions. It is difficult to provide real-time and comprehensive situational awareness when making rescue decisions.

[0003] Fire monitoring based on a single sensor cannot accurately obtain information such as the three-dimensional structure, temperature distribution, and smoke diffusion dynamics of the fire scene; in addition, the high temperature, thick smoke, and electromagnetic interference in the fire environment can easily cause problems such as data transmission interruption. The traditional physical models used in fire control have high computational complexity and are difficult to meet the needs of real-time control, which reduces the real-time and accuracy of urban fire control, and thus reduces the comprehensiveness and stability of urban fire control. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method and device for urban fire control based on fire scene simulation, which collects scene data of the target area where a fire occurs in the city through a cluster of drones, and constructs a fire scene twin simulation model corresponding to the target area based on the scene data to dynamically simulate the fire evolution process, and use the fire prediction model field to deduce the fire in the target area, determine the dangerous area information and evacuation channel information, and then generate a rescue path for the target area in the fire scene twin simulation model to determine the fire control strategy of the target area, and control the fire in the target area according to the fire control strategy, thereby improving the real-time and accuracy of urban fire control, and thus improving the comprehensiveness and stability of urban fire control.

[0005] The present invention provides a fire scene simulation-based urban fire control method, comprising:

[0006] In response to receiving a command signal for fire control in a target area where a fire occurs in a city, using a preset drone cluster to collect scene data of the target area;

[0007] Based on the scene data, a fire scene twin simulation model corresponding to the target area is constructed; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field;

[0008] Inputting the fire scene twin simulation model into a preset fire intensity prediction model, using the fire intensity prediction model to deduce the fire intensity of the target area based on the temperature field and the smoke diffusion vector field, and obtaining the dangerous area information and evacuation channel information output by the fire prediction model;

[0009] Based on the dangerous area information and the evacuation channel information, a rescue path for the target area is generated in the fire scene twin simulation model to determine a fire control strategy for the target area, and fire control is performed on the target area according to the fire control strategy.

[0010] Furthermore, in response to receiving a command signal for performing fire control on a target area where a fire occurs in a city, collecting scene data of the target area using a preset drone cluster includes:

[0011] In response to receiving a command signal for fire control in a target area of a city where a fire has occurred, time calibration is performed on data acquisition devices disposed in a preset drone cluster based on a preset time protocol;

[0012] In response to the data acquisition devices being calibrated to a unified time, controlling the drone cluster to a hovering state, and calculating a rigid transformation matrix corresponding to the data acquisition devices based on a calibration plate preset in the target area;

[0013] Based on the rigid transformation matrix, registering the acquisition accuracy of the data acquisition device by reprojection to determine that the multimodal data space of the data acquisition device is in an aligned state;

[0014] In response to determining that the multimodal data space is in an aligned state, scene data of the target area is collected using a preset drone cluster.

[0015] Furthermore, the scene data includes at least point cloud data, multispectral image data, gas data and environmental data;

[0016] The step of constructing a fire scene twin simulation model corresponding to the target area based on the scene data includes:

[0017] Performing denoising processing on the point cloud data to obtain target point cloud data, and performing radiometric normalization processing on the multispectral image data to obtain target multispectral image data;

[0018] Constructing a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data;

[0019] Based on the environmental data, a digital twin simulation is performed on the fire scene simulation model to construct a fire scene twin simulation model corresponding to the target area, so as to map the target area to the fire scene simulation model in real time.

[0020] Furthermore, constructing a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data includes:

[0021] Projecting the target multispectral image data onto the point cloud data to obtain a point cloud spectral image, and filling texture missing areas in the point cloud spectral image using a bilinear interpolation method to obtain a target point cloud spectral image;

[0022] Extracting the radiation temperature of the target area in the target point cloud spectral image based on the thermal infrared band in the target multispectral image data, and constructing the temperature field corresponding to the target area based on the radiation temperature and the point cloud spatial distribution in the target point cloud spectral image;

[0023] constructing a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data;

[0024] A fire scene simulation model corresponding to the target area is constructed based on the target point cloud spectral image, the temperature field and the smoke diffusion vector field.

[0025] Furthermore, constructing the smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data includes:

[0026] Based on the thermal infrared bands in the target multispectral image data, fire points in the target multispectral image data are identified to determine fire point cloud data corresponding to the target area, and based on the fire point cloud data, the combustion area volume and flame height distribution information of the target area are calculated;

[0027] Based on the target multispectral image data, the pixel displacement field of the target area is calculated using the optical flow method, and the plane displacement in the pixel displacement field is converted into a three-dimensional wind speed vector to obtain the target pixel displacement field;

[0028] A smoke diffusion vector field corresponding to the target area is constructed based on the combustion area volume, the flame height distribution information, the target pixel displacement field and the gas data.

[0029] Furthermore, based on the dangerous area information and the evacuation channel information, generating a rescue path for the target area in the fire scene twin simulation model to determine a fire control strategy for the target area, and performing fire control on the target area according to the fire control strategy, includes:

[0030] Based on the dangerous area information and the urban building data corresponding to the target area, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model, and the dangerous level corresponding to each dangerous area in the target area is determined;

[0031] Calculating a rescue path cost based on the evacuation channel information and the fire scene information in the fire scene twin simulation model;

[0032] generating a rescue path for the target area in the fire scene twin simulation model based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the cost value, and the danger level;

[0033] Based on the rescue path, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0034] Furthermore, determining a fire control strategy for the target area based on the rescue path, and performing fire control on the target area according to the fire control strategy, includes:

[0035] In response to the rescue path being determined, receiving an intervention plan determined by an external person based on the rescue path and the fire scene twin simulation model;

[0036] Performing rescue simulation on the intervention plan in the fire scene twin simulation model to obtain rescue simulation results;

[0037] Based on the rescue path and the rescue simulation result, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0038] The present application also provides an urban fire control device based on fire scene simulation, the control device comprising:

[0039] a data acquisition module, configured to, in response to receiving a command signal for fire control in a target area where a fire has occurred in the city, collect scene data of the target area using a preset drone cluster;

[0040] A model construction module is used to construct a fire scene twin simulation model corresponding to the target area based on the scene data; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field;

[0041] a fire prediction module, configured to input the fire scene twin simulation model into a preset fire prediction model, utilize the fire prediction model to perform fire prediction on the target area based on the temperature field and the smoke diffusion vector field, and obtain the dangerous area information and evacuation channel information output by the fire prediction model;

[0042] A fire control module is used to generate a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information, so as to determine the fire control strategy for the target area and perform fire control on the target area according to the fire control strategy.

[0043] Furthermore, when the data acquisition module is used to collect scene data of a target area where a fire occurs in a city in response to receiving a command signal for fire control, the data acquisition module is used to:

[0044] In response to receiving a command signal for fire control in a target area of a city where a fire has occurred, time calibration is performed on data acquisition devices disposed in a preset drone cluster based on a preset time protocol;

[0045] In response to the data acquisition devices being calibrated to a unified time, controlling the drone cluster to a hovering state, and calculating a rigid transformation matrix corresponding to the data acquisition devices based on a calibration plate preset in the target area;

[0046] Based on the rigid transformation matrix, registering the acquisition accuracy of the data acquisition device by reprojection to determine that the multimodal data space of the data acquisition device is in an aligned state;

[0047] In response to determining that the multimodal data space is in an aligned state, scene data of the target area is collected using a preset drone cluster.

[0048] Furthermore, when the model construction module is used to construct the fire scene twin simulation model corresponding to the target area based on the scene data, the model construction module is used to:

[0049] De-noising the point cloud data to obtain target point cloud data, and performing radiometric normalization on the multispectral image data to obtain target multispectral image data;

[0050] Constructing a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data;

[0051] Based on the environmental data, a digital twin simulation is performed on the fire scene simulation model to construct a fire scene twin simulation model corresponding to the target area, so as to map the target area to the fire scene simulation model in real time.

[0052] Furthermore, when the model construction module is used to construct a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data, the model construction module is used to:

[0053] Projecting the target multispectral image data onto the point cloud data to obtain a point cloud spectral image, and filling texture missing areas in the point cloud spectral image using a bilinear interpolation method to obtain a target point cloud spectral image;

[0054] Extracting the radiation temperature of the target area in the target point cloud spectral image based on the thermal infrared band in the target multispectral image data, and constructing the temperature field corresponding to the target area based on the radiation temperature and the point cloud spatial distribution in the target point cloud spectral image;

[0055] constructing a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data;

[0056] A fire scene simulation model corresponding to the target area is constructed based on the target point cloud spectral image, the temperature field and the smoke diffusion vector field.

[0057] Furthermore, when the model building module is used to build the smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data, the model building module is used to:

[0058] Based on the thermal infrared bands in the target multispectral image data, fire points in the target multispectral image data are identified to determine fire point cloud data corresponding to the target area, and based on the fire point cloud data, the combustion area volume and flame height distribution information of the target area are calculated;

[0059] Based on the target multispectral image data, the pixel displacement field of the target area is calculated using the optical flow method, and the plane displacement in the pixel displacement field is converted into a three-dimensional wind speed vector to obtain the target pixel displacement field;

[0060] A smoke diffusion vector field corresponding to the target area is constructed based on the combustion area volume, the flame height distribution information, the target pixel displacement field and the gas data.

[0061] Furthermore, when the fire control module is used to generate a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information, so as to determine a fire control strategy for the target area, and perform fire control on the target area according to the fire control strategy, the fire control module is used to:

[0062] Based on the dangerous area information and the urban building data corresponding to the target area, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model, and the dangerous level corresponding to each dangerous area in the target area is determined;

[0063] Calculating a rescue path cost based on the evacuation channel information and the fire scene information in the fire scene twin simulation model;

[0064] generating a rescue path for the target area in the fire scene twin simulation model based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the cost value, and the danger level;

[0065] Based on the rescue path, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0066] Furthermore, when the fire control module is used to determine the fire control strategy of the target area based on the rescue path and perform fire control on the target area according to the fire control strategy, the fire control module is used to:

[0067] In response to the rescue path being determined, receiving an intervention plan determined by an external person based on the rescue path and the fire scene twin simulation model;

[0068] Performing rescue simulation on the intervention plan in the fire scene twin simulation model to obtain rescue simulation results;

[0069] Based on the rescue path and the rescue simulation result, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0070] An embodiment of the present application also provides an electronic device, comprising: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the urban fire control method based on fire scene simulation as described above are performed.

[0071] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the urban fire control method based on fire scene simulation as described above are executed.

[0072] The embodiments of the present application provide a method and device for urban fire control based on fire scene simulation, the method comprising: in response to receiving a command signal for fire control of a target area where a fire occurs in a city, using a preset drone cluster to collect scene data of the target area; based on the scene data, constructing a fire scene twin simulation model corresponding to the target area; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field; inputting the fire scene twin simulation model into a preset fire prediction model, using the fire prediction model to deduce the fire in the target area based on the temperature field and the smoke diffusion vector field, and obtaining the dangerous area information and evacuation channel information output by the fire prediction model; based on the dangerous area information and the evacuation channel information, generating a rescue path for the target area in the fire scene twin simulation model to determine a fire control strategy for the target area, and performing fire control on the target area according to the fire control strategy.

[0073] Compared with the existing fire control method that relies on fixed sensors and manual inspections, the scene data of the target area where the fire occurs in the city is collected by drone clusters. Based on the scene data, a fire scene twin simulation model corresponding to the target area is constructed to dynamically simulate the fire evolution process. The fire prediction model field is used to deduce the fire in the target area, determine the dangerous area information and evacuation channel information, and then generate a rescue path for the target area in the fire scene twin simulation model to determine the fire control strategy for the target area. The fire is then controlled in the target area according to the fire control strategy, which improves the real-time and accuracy of urban fire control, and thus improves the comprehensiveness and stability of urban fire control.

[0074] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0076] Figure 1A flow chart of a method for urban fire control based on fire scene simulation provided in an embodiment of the present application;

[0077] Figure 2 A schematic diagram of the structure of an urban fire control device based on fire scene simulation provided in an embodiment of the present application;

[0078] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for which protection is claimed, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work falls within the scope of protection of the present application.

[0080] Research has found that traditional fire control methods rely on fixed sensors and manual inspections, and have problems such as delayed response and single data dimensions, making it difficult to provide real-time and comprehensive situational awareness when making rescue decisions.

[0081] Among them, fire monitoring based on a single sensor cannot accurately obtain information such as the three-dimensional structure, temperature distribution, and smoke diffusion dynamics of the fire scene; in addition, the high temperature, thick smoke, and electromagnetic interference in the fire environment can easily cause problems such as data transmission interruption. The traditional physical models used in fire control have high computational complexity and are difficult to meet the needs of real-time control, which reduces the real-time and accuracy of urban fire control, and thus reduces the comprehensiveness and stability of urban fire control.

[0082] Based on this, an embodiment of the present application provides an urban fire control method based on fire scene simulation, which collects scene data of the target area where a fire occurs in the city through a cluster of drones, and constructs a fire scene twin simulation model corresponding to the target area based on the scene data to dynamically simulate the fire evolution process. The fire prediction model field is used to deduce the fire in the target area, determine the dangerous area information and evacuation channel information, and then generate a rescue path for the target area in the fire scene twin simulation model to determine the fire control strategy for the target area, and control the fire in the target area according to the fire control strategy, thereby improving the real-time and accuracy of urban fire control, and thus improving the comprehensiveness and stability of urban fire control.

[0083] See also Figure 1 , Figure 1 This is a flow chart of a method for urban fire control based on fire scene simulation provided by an embodiment of the present application. Figure 1 As shown in , the urban fire control method based on fire scene simulation provided by the embodiment of the present application includes:

[0084] S101: In response to receiving a command signal for performing fire control on a target area where a fire occurs in a city, using a preset drone cluster to collect scene data of the target area.

[0085] In an embodiment of the present application, a drone cluster is pre-constructed, wherein the drones in the drone cluster are selected from medium and large rotor drones with high load capacity to ensure that they can carry multiple sensors; a multi-channel multi-spectral camera (for example, a 16-channel multi-spectral camera in the 400-1700nm band) is installed in the belly gimbal of the drone; a distributed gas detection unit (for example, CO, VOCs, PM2.5 / 10 sensors) is deployed at the end of the drone's arm; a solid-state laser radar with more than 32 lines is used on the top of the drone's fuselage, and the scanning angle of the solid-state laser radar is ensured to be greater than or equal to 270°; the drone cluster also includes an edge computing module.

[0086] Furthermore, a fixing device is added next to the solid-state laser radar, and the 5G communication module and millimeter-wave self-organizing network module are added to the top of the drone's fuselage. At the same time, a broadband RF transceiver is installed in the communication module, and a heat dissipation kit is installed near the core communication module to ensure the reliability of the equipment in extreme fire situations.

[0087] In an embodiment of the present application, the scene data includes but is not limited to point cloud data, multispectral image data, gas data and environmental data.

[0088] Among them, the multispectral image data at least includes RGB image channels and thermal infrared bands (far / near); the gas data at least includes gas type, gas concentration and smoke information; the environmental data at least includes temperature and humidity data, trapped personnel information and building information.

[0089] In one embodiment of the present application, during specific implementation, step S101 may include:

[0090] S1011. In response to receiving a command signal for fire control in a target area where a fire has occurred in a city, time calibration is performed on data acquisition equipment set in a preset drone cluster based on a preset time protocol.

[0091] In this step, when a command signal is received for fire control in a target area where a fire occurs in a city, the sensor clocks of the data acquisition devices in the drone cluster are accurately synchronized based on the preset time protocol and network hybrid communication unit for time calibration.

[0092] S1012. In response to the data acquisition device being calibrated to time uniformity, the drone cluster is controlled to a hovering state, and based on a calibration plate preset in the target area, a rigid transformation matrix corresponding to the data acquisition device is calculated.

[0093] In this step, after the data acquisition equipment is calibrated to a unified time, the drone cluster is controlled to be in a hovering state, and in the hovering state of the drone cluster, a checkerboard calibration plate of known size preset in the target area is used to collect multi-angle data, and based on the point cloud plane fitting and image corner points in the multi-angle data, the rigid transformation matrix corresponding to the data acquisition equipment is calculated.

[0094] S1013 . Based on the rigid transformation matrix, align the acquisition accuracy of the data acquisition device by reprojection to determine that the multimodal data space of the data acquisition device is in an aligned state.

[0095] In the embodiment of the present application, the collection accuracy of the data collection equipment is accurate to less than 5 cm.

[0096] S1014: In response to determining that the multimodal data space is in an aligned state, using a preset drone cluster to collect scene data of the target area.

[0097] In this step, when the multimodal data space is determined to be in an aligned state, the multi-channel multispectral camera, solid-state lidar, distributed gas detection unit and edge computing module in the preset drone cluster are used to collect point cloud data, multispectral image data, gas data and environmental data of the target area.

[0098] S102: Based on the scene data, construct a fire scene twin simulation model corresponding to the target area.

[0099] The fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field.

[0100] In one embodiment of the present application, during specific implementation, step S102 may include:

[0101] S1021 , performing denoising processing on the point cloud data to obtain target point cloud data, and performing radiometric normalization processing on the multispectral image data to obtain target multispectral image data.

[0102] In this step, the statistical outlier filtering method is used to filter and denoise the point cloud data to remove flying points in the point cloud data, and the multispectral image data is radiometrically normalized based on a preset radiometric calibration plate to obtain the target multispectral image data.

[0103] S1022: Construct a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data, and the gas data.

[0104] In one embodiment of the present application, during specific implementation, step S1022 may include:

[0105] S10221. Project the target multispectral image data onto the point cloud data to obtain a point cloud spectral image, and fill in texture missing areas in the point cloud spectral image using a bilinear interpolation method to obtain a target point cloud spectral image.

[0106] In this step, according to the position coordinates of each point in the point cloud data, the closest pixel position is found in the corresponding multispectral image, and the multispectral value (reflectance value of each band) at the pixel position is assigned to the corresponding point in the point cloud to obtain a point cloud spectral image including three-dimensional coordinate information and spectral information.

[0107] Furthermore, for each missing pixel in the point cloud spectral image, the four closest valid pixels are found, and the weights are calculated based on the distance between these pixels and the missing pixel. The closer the pixel, the greater the impact on the final result. These weights and the corresponding pixel values are then used to calculate the estimated value of the missing pixel to fill the texture missing area in the point cloud spectral image and obtain the target point cloud spectral image.

[0108] S10222. Based on the thermal infrared band in the target multispectral image data, extract the radiation temperature of the target area in the target point cloud spectral image, and construct a temperature field corresponding to the target area based on the radiation temperature and the point cloud spatial distribution in the target point cloud spectral image.

[0109] In this step, during specific implementation, first, the multispectral thermal infrared band in the target multispectral image data is radiated and calibrated, and the DN value is converted into a radiant brightness value; then, based on the blackbody radiation formula, the radiation temperature of each pixel in the target multispectral image data is calculated; thereafter, the calibrated rigid transformation matrix is used to accurately map the thermal infrared image pixels to the point cloud data in the target multispectral image data, and a three-dimensional spatial correspondence between the image pixels and the point cloud is established; finally, the point cloud is voxelized, and the temperature value is interpolated and calculated within each voxel using the inverse distance weighted algorithm to construct the temperature field corresponding to the target area.

[0110] S10223. Construct a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data.

[0111] In one embodiment of the present application, during specific implementation, step S10223 may include:

[0112] S102231. Based on the thermal infrared band in the target multispectral image data, identify the fire points in the target multispectral image data to determine the fire point cloud data corresponding to the target area, and calculate the combustion area volume and flame height distribution information of the target area based on the fire point cloud data.

[0113] In this step, the fire point probability index of the target area is constructed using the NIR and SWIR bands, and the point clouds with a fire point probability index greater than 0.7 are regarded as valid fire points to determine the fire point cloud data corresponding to the target area.

[0114] Furthermore, the camera calibration matrix of the solid-state lidar of the drone cluster is used to map the point cloud data of the fire points corresponding to the target area to the point cloud data of the target area, and the burning area volume (voxel statistics) and flame height distribution information (spatial vertical axis statistics) of the target area are calculated.

[0115] S102232. Based on the target multispectral image data, use the optical flow method to calculate the pixel displacement field of the target area, and convert the plane displacement in the pixel displacement field into a three-dimensional wind speed vector to obtain the target pixel displacement field.

[0116] In this step, the optical flow method is used to calculate the pixel displacement field of the continuous multi-frame multispectral images in the target multispectral image data, and the plane displacement in the pixel displacement field is converted into a three-dimensional wind speed vector through point cloud projection to obtain the target pixel displacement field.

[0117] S102233. Construct a smoke diffusion vector field corresponding to the target area based on the volume of the combustion area, the flame height distribution information, the target pixel displacement field and the gas data.

[0118] In this step, based on the volume of the burning area, the flame height distribution information, and the target pixel displacement field, combined with the gas data collected by the gas sensors carried by the drone cluster, a concentration-wind speed correlation model is established to construct the smoke diffusion vector field corresponding to the target area.

[0119] Furthermore, in the smoke diffusion vector field, the measured temperature of the infrared thermal imager is compared with the output of the smoke diffusion vector field, and the smoke diffusion direction is verified using tracer gas.

[0120] S10224. Construct a fire scene simulation model corresponding to the target area based on the target point cloud spectral image, the temperature field, and the smoke diffusion vector field.

[0121] In this step, the target point cloud spectral image, temperature field and smoke diffusion vector field are simulated by model integration to construct a three-dimensional fire scene simulation model corresponding to the target area.

[0122] S1023. Based on the environmental data, perform digital twin simulation on the fire scene simulation model to construct a fire scene twin simulation model corresponding to the target area, so as to map the target area to the fire scene simulation model in real time.

[0123] In this step, based on the environmental data, the target point cloud data and the target multispectral image data are aligned at the sub-pixel level through the cross-modal registration technology of feature point matching, and the actually measured environmental data such as temperature and smoke concentration are mapped to the corresponding positions of the fire scene simulation model in real time; the physical parameters of the fire scene simulation model are visualized through material shader technology (for example, red-blue gradient colors represent the temperature distribution of 200-1200℃), so as to map the target area to the fire scene simulation model in real time and construct a fire scene twin simulation model corresponding to the target area.

[0124] Furthermore, millisecond-level data synchronization between the physical world of the target area and the fire scene simulation model can be achieved through 5G / LAN.

[0125] S103. Input the fire scene twin simulation model into a preset fire prediction model, and use the fire prediction model to deduce the fire in the target area based on the temperature field and the smoke diffusion vector field to obtain the dangerous area information and evacuation channel information output by the fire prediction model.

[0126] In an embodiment of the present application, a fire prediction model is set up through the following steps: pre-collecting the real-time temperature field collected by the drone, combining the smoke diffusion vector and local environmental parameters as the physical model input, and combining the historical fire scene evolution time series data and multispectral image segmentation to obtain the real-time fire scene boundary contour as the deep learning input.

[0127] Furthermore, a unified spatiotemporal coordinate system is established, the spatial grid resolution of the spatiotemporal coordinate system is unified, and the time is aligned. The historical fire boundary is used as the input layer of the LSTM model of the fire prediction model, and the future fire expansion probability map is output. The continuous frame fire images generated by multispectral images are used as the input of the GAN model of the fire prediction model, and the enhanced extreme fire evolution scene is output. The prediction ability of rare events such as deflagration and flying fire is improved, the dangerous areas are marked, and potential evacuation routes are inferred to avoid high temperature and high smoke areas, and the fire prediction model is trained.

[0128] Among them, the fire prediction model includes the LSTM time series prediction model and the GAN model. The GAN model enhances the details and imposes physical constraints on the prediction results.

[0129] S104. Based on the dangerous area information and the evacuation channel information, a rescue path for the target area is generated in the fire scene twin simulation model to determine a fire control strategy for the target area, and fire control is performed on the target area according to the fire control strategy.

[0130] In an embodiment of the present application, the dangerous areas in the target area are graded based on the dangerous area information and the evacuation channel information, and a dynamic optimal rescue path is generated. Based on the continuous reporting of the location of trapped persons by drones, the drone data is linked with the city fire protection system. In addition, external personnel determine the corresponding intervention plan in the fire scene twin simulation model based on the rescue path and environmental data, and then determine the fire control strategy for the target area.

[0131] In one embodiment of the present application, during specific implementation, step S104 may include:

[0132] S1041. Based on the dangerous area information and the urban building data corresponding to the target area, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model, and the dangerous level corresponding to each dangerous area in the target area is determined.

[0133] In this step, based on the dangerous area information and the urban building data corresponding to the target area, combined with the fire point probability index threshold and smoke diffusion field in the fire scene twin simulation model, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model to determine the corresponding dangerous level of each dangerous area in the target area.

[0134] S1042. Calculate the cost value of the rescue path based on the evacuation channel information and the fire scene information in the fire scene twin simulation model.

[0135] Here, the fire scene information includes but is not limited to the path distance, temperature risk coefficient, smoke concentration and channel width value corresponding to each path in the target area.

[0136] In an embodiment of the present application, the cost value of the rescue path is calculated using a preset cost function based on the path distance, temperature risk coefficient, smoke concentration and channel width value. The expression of the cost function is shown below.

[0137] Cost(P)=ax+βy+γz+δm.

[0138] Among them, represents the cost value corresponding to each path segment in the target area; x, y, z and m represent the path distance, temperature risk coefficient, smoke concentration and channel width respectively; α represents the weight coefficient corresponding to the path distance; β represents the weight coefficient corresponding to the temperature risk coefficient; γ represents the weight coefficient corresponding to the smoke concentration; δ represents the weight coefficient corresponding to the channel width.

[0139] Here, for each path section in the high-temperature area of the target area, the weight coefficient corresponding to the set temperature risk coefficient is higher; for each path section corresponding to the narrow channel in the target area, the weight coefficient corresponding to the set channel width value is higher; for each path section corresponding to the headwind smoke in the high-temperature area of the target area, the weight coefficient corresponding to the set smoke concentration is higher.

[0140] S1043. Based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the cost value and the danger level, generate a rescue path for the target area in the fire scene twin simulation model.

[0141] In this step, based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the coordinates of the trapped persons are clustered to generate priority rescue points. Based on the cost value corresponding to each path segment and the danger level corresponding to each dangerous area, a rescue path for the target area is generated in the fire scene twin simulation model to balance the shortest time and maximum survival rate.

[0142] S1044: Determine a fire control strategy for the target area based on the rescue path, and perform fire control on the target area according to the fire control strategy.

[0143] In one embodiment of the present application, during specific implementation, step S1044 may include:

[0144] S10441. In response to the rescue path being determined, receive an intervention plan determined by an external person based on the rescue path and the fire scene twin simulation model.

[0145] In this step, after the rescue path is determined in the fire scene twin simulation model, the external personnel command personnel to interact with the fire scene twin simulation model in the AR / VR environment (such as virtual water injection and obstacle removal, etc.), simulate the effects of different intervention strategies, and determine the intervention plan and send it to the management and control system corresponding to the fire scene twin simulation model.

[0146] The intervention plan at least includes water source deployment, obstacle removal and evacuation guidance.

[0147] S10442. Perform rescue simulation on the intervention plan in the fire scene twin simulation model to obtain a rescue simulation result.

[0148] In this step, each determined intervention plan is subjected to rescue simulation in the fire scene twin simulation model to obtain the rescue simulation results of each intervention plan.

[0149] S10443. Determine a fire control strategy for the target area based on the rescue path and the rescue simulation result, and perform fire control on the target area according to the fire control strategy.

[0150] In this step, based on the rescue path and each intervention plan, combined with the rescue simulation results of each intervention plan, the fire control strategy of the target area is determined, and then the urban fire protection system is linked to carry out fire control in the target area.

[0151] The urban fire control method based on fire scene simulation provided in the embodiment of the present application collects scene data of the target area where the fire occurs in the city through a drone cluster, and constructs a fire scene twin simulation model corresponding to the target area based on the scene data to dynamically simulate the fire evolution process, and uses the fire prediction model field to deduce the fire in the target area, determine the dangerous area information and evacuation channel information, and then generate a rescue path for the target area in the fire scene twin simulation model to determine the fire control strategy for the target area, and control the fire in the target area according to the fire control strategy, thereby improving the real-time and accuracy of urban fire control, and thereby improving the comprehensiveness and stability of urban fire control.

[0152] See also Figure 2 , Figure 2 This is a schematic diagram of the structure of a city fire control device based on fire scene simulation provided by an embodiment of the present application. Figure 2 As shown in , the control device 200 includes:

[0153] The data collection module 210 is configured to, in response to receiving a command signal for fire control in a target area where a fire occurs in a city, collect scene data of the target area using a preset drone cluster;

[0154] A model building module 220 is configured to build a fire scene twin simulation model corresponding to the target area based on the scene data; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field;

[0155] The fire prediction module 230 is configured to input the fire scene twin simulation model into a preset fire prediction model, use the fire prediction model to perform fire prediction on the target area based on the temperature field and the smoke diffusion vector field, and obtain the dangerous area information and evacuation route information output by the fire prediction model;

[0156] The fire control module 240 is used to generate a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information, so as to determine the fire control strategy for the target area and perform fire control on the target area according to the fire control strategy.

[0157] Furthermore, when the data acquisition module 210 is used to collect scene data of a target area where a fire occurs in a city in response to receiving a command signal for fire control, the data acquisition module 210 is used to:

[0158] In response to receiving a command signal for fire control in a target area of a city where a fire has occurred, time calibration is performed on data acquisition devices disposed in a preset drone cluster based on a preset time protocol;

[0159] In response to the data acquisition devices being calibrated to a unified time, controlling the drone cluster to a hovering state, and calculating a rigid transformation matrix corresponding to the data acquisition devices based on a calibration plate preset in the target area;

[0160] Based on the rigid transformation matrix, registering the acquisition accuracy of the data acquisition device by reprojection to determine that the multimodal data space of the data acquisition device is in an aligned state;

[0161] In response to determining that the multimodal data space is in an aligned state, scene data of the target area is collected using a preset drone cluster.

[0162] Furthermore, when the model building module 220 is used to build the fire scene twin simulation model corresponding to the target area based on the scene data, the model building module 220 is used to:

[0163] De-noising the point cloud data to obtain target point cloud data, and performing radiometric normalization on the multispectral image data to obtain target multispectral image data;

[0164] Constructing a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data;

[0165] Based on the environmental data, a digital twin simulation is performed on the fire scene simulation model to construct a fire scene twin simulation model corresponding to the target area, so as to map the target area to the fire scene simulation model in real time.

[0166] Furthermore, when the model building module 220 is used to build a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data, the model building module 220 is used to:

[0167] Projecting the target multispectral image data onto the point cloud data to obtain a point cloud spectral image, and filling texture missing areas in the point cloud spectral image using a bilinear interpolation method to obtain a target point cloud spectral image;

[0168] Extracting the radiation temperature of the target area in the target point cloud spectral image based on the thermal infrared band in the target multispectral image data, and constructing the temperature field corresponding to the target area based on the radiation temperature and the point cloud spatial distribution in the target point cloud spectral image;

[0169] constructing a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data;

[0170] A fire scene simulation model corresponding to the target area is constructed based on the target point cloud spectral image, the temperature field and the smoke diffusion vector field.

[0171] Furthermore, when the model building module 220 is used to build the smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data, the model building module 220 is used to:

[0172] Based on the thermal infrared bands in the target multispectral image data, fire points in the target multispectral image data are identified to determine fire point cloud data corresponding to the target area, and based on the fire point cloud data, the combustion area volume and flame height distribution information of the target area are calculated;

[0173] Based on the target multispectral image data, the pixel displacement field of the target area is calculated using the optical flow method, and the plane displacement in the pixel displacement field is converted into a three-dimensional wind speed vector to obtain the target pixel displacement field;

[0174] A smoke diffusion vector field corresponding to the target area is constructed based on the combustion area volume, the flame height distribution information, the target pixel displacement field and the gas data.

[0175] Furthermore, when the fire control module 240 is used to generate a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information, so as to determine a fire control strategy for the target area, and perform fire control on the target area according to the fire control strategy, the fire control module 240 is used to:

[0176] Based on the dangerous area information and the urban building data corresponding to the target area, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model, and the dangerous level corresponding to each dangerous area in the target area is determined;

[0177] Calculating a rescue path cost based on the evacuation channel information and the fire scene information in the fire scene twin simulation model;

[0178] generating a rescue path for the target area in the fire scene twin simulation model based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the cost value, and the danger level;

[0179] Based on the rescue path, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0180] Furthermore, when the fire control module 240 is used to determine the fire control strategy of the target area based on the rescue path and perform fire control on the target area according to the fire control strategy, the fire control module 240 is used to:

[0181] In response to the rescue path being determined, receiving an intervention plan determined by an external person based on the rescue path and the fire scene twin simulation model;

[0182] Performing rescue simulation on the intervention plan in the fire scene twin simulation model to obtain rescue simulation results;

[0183] Based on the rescue path and the rescue simulation result, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

[0184] The urban fire control device based on fire scene simulation provided in the embodiment of the present application collects scene data of the target area where a fire occurs in the city through a drone cluster, and constructs a fire scene twin simulation model corresponding to the target area based on the scene data to dynamically simulate the fire evolution process, and uses the fire prediction model field to deduce the fire in the target area, determine the dangerous area information and evacuation channel information, and then generate a rescue path for the target area in the fire scene twin simulation model to determine the fire control strategy for the target area, and control the fire in the target area according to the fire control strategy, thereby improving the real-time and accuracy of urban fire control, and thereby improving the comprehensiveness and stability of urban fire control.

[0185] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 3 As shown in FIG, the electronic device 300 includes a processor 310 , a memory 320 and a bus 330 .

[0186] The memory 320 stores machine-readable instructions executable by the processor 310. When the electronic device 300 is running, the processor 310 communicates with the memory 320 via the bus 330. When the machine-readable instructions are executed by the processor 310, the above-mentioned Figure 1 The steps of the urban fire control method based on fire scene simulation in the method embodiment shown are specifically implemented in accordance with the method embodiment and will not be described in detail here.

[0187] The embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The steps of the urban fire control method based on fire scene simulation in the method embodiment shown are specifically implemented in accordance with the method embodiment and will not be described in detail here.

[0188] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0189] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. There may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0190] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0191] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0192] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0193] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-mentioned embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. These modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for urban fire control based on fire scene simulation, characterized in that: The method comprises: In response to receiving a command signal for fire control in a target area where a fire occurs in a city, using a preset drone cluster to collect scene data of the target area; Based on the scene data, a fire scene twin simulation model corresponding to the target area is constructed; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field; Inputting the fire scene twin simulation model into a preset fire intensity prediction model, using the fire intensity prediction model to deduce the fire intensity of the target area based on the temperature field and the smoke diffusion vector field, and obtaining the dangerous area information and evacuation channel information output by the fire prediction model; Based on the dangerous area information and the evacuation channel information, a rescue path for the target area is generated in the fire scene twin simulation model to determine a fire control strategy for the target area, and fire control is performed on the target area according to the fire control strategy.

2. The method according to claim 1, characterized in that In response to receiving a command signal for performing fire control on a target area where a fire occurs in a city, collecting scene data of the target area using a preset drone cluster includes: In response to receiving a command signal for fire control in a target area of a city where a fire has occurred, time calibration is performed on data acquisition devices disposed in a preset drone cluster based on a preset time protocol; In response to the data acquisition devices being calibrated to a unified time, controlling the drone cluster to a hovering state, and calculating a rigid transformation matrix corresponding to the data acquisition devices based on a calibration plate preset in the target area; Based on the rigid transformation matrix, registering the acquisition accuracy of the data acquisition device by reprojection to determine that the multimodal data space of the data acquisition device is in an aligned state; In response to determining that the multimodal data space is in an aligned state, scene data of the target area is collected using a preset drone cluster.

3. The method according to claim 1, characterized in that The scene data includes at least point cloud data, multispectral image data, gas data and environmental data; The step of constructing a fire scene twin simulation model corresponding to the target area based on the scene data includes: Performing denoising processing on the point cloud data to obtain target point cloud data, and performing radiometric normalization processing on the multispectral image data to obtain target multispectral image data; Constructing a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data and the gas data; Based on the environmental data, a digital twin simulation is performed on the fire scene simulation model to construct a fire scene twin simulation model corresponding to the target area, so as to map the target area to the fire scene simulation model in real time.

4. The method according to claim 3, characterized in that The constructing of a fire scene simulation model corresponding to the target area based on the target point cloud data, the target multispectral image data, and the gas data includes: Projecting the target multispectral image data onto the point cloud data to obtain a point cloud spectral image, and filling texture missing areas in the point cloud spectral image using a bilinear interpolation method to obtain a target point cloud spectral image; Extracting the radiation temperature of the target area in the target point cloud spectral image based on the thermal infrared band in the target multispectral image data, and constructing the temperature field corresponding to the target area based on the radiation temperature and the point cloud spatial distribution in the target point cloud spectral image; constructing a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data; A fire scene simulation model corresponding to the target area is constructed based on the target point cloud spectral image, the temperature field and the smoke diffusion vector field.

5. The method according to claim 4, characterized in that The step of constructing a smoke diffusion vector field corresponding to the target area based on the target point cloud spectral image and the gas data includes: Based on the thermal infrared bands in the target multispectral image data, fire points in the target multispectral image data are identified to determine fire point cloud data corresponding to the target area, and based on the fire point cloud data, the combustion area volume and flame height distribution information of the target area are calculated; Based on the target multispectral image data, the pixel displacement field of the target area is calculated using the optical flow method, and the plane displacement in the pixel displacement field is converted into a three-dimensional wind speed vector to obtain the target pixel displacement field; A smoke diffusion vector field corresponding to the target area is constructed based on the combustion area volume, the flame height distribution information, the target pixel displacement field and the gas data.

6. The method according to claim 1, characterized in that The generating of a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information to determine a fire control strategy for the target area, and performing fire control on the target area according to the fire control strategy, including: Based on the dangerous area information and the urban building data corresponding to the target area, the dangerous areas in the target area are divided into dangerous levels in the fire scene twin simulation model, and the dangerous level corresponding to each dangerous area in the target area is determined; Calculating a rescue path cost based on the evacuation channel information and the fire scene information in the fire scene twin simulation model; generating a rescue path for the target area in the fire scene twin simulation model based on the clustered point cloud data of trapped persons in the fire scene twin simulation model, the cost value, and the danger level; Based on the rescue path, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

7. The method according to claim 6, characterized in that The determining of a fire control strategy for the target area based on the rescue path, and performing fire control on the target area according to the fire control strategy, includes: In response to the rescue path being determined, receiving an intervention plan determined by an external person based on the rescue path and the fire scene twin simulation model; Performing rescue simulation on the intervention plan in the fire scene twin simulation model to obtain rescue simulation results; Based on the rescue path and the rescue simulation result, a fire control strategy for the target area is determined, and fire control is performed on the target area according to the fire control strategy.

8. An urban fire control device based on fire scene simulation, characterized in that: The control device includes: a data acquisition module, configured to, in response to receiving a command signal for fire control in a target area where a fire has occurred in the city, collect scene data of the target area using a preset drone cluster; A model construction module is used to construct a fire scene twin simulation model corresponding to the target area based on the scene data; wherein the fire scene twin simulation model includes at least a temperature field and a smoke diffusion vector field; a fire prediction module, configured to input the fire scene twin simulation model into a preset fire prediction model, utilize the fire prediction model to perform fire prediction on the target area based on the temperature field and the smoke diffusion vector field, and obtain the dangerous area information and evacuation channel information output by the fire prediction model; A fire control module is used to generate a rescue path for the target area in the fire scene twin simulation model based on the dangerous area information and the evacuation channel information, so as to determine the fire control strategy for the target area and perform fire control on the target area according to the fire control strategy.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus. When the processor is running, the machine-readable instructions execute the steps of the urban fire control method based on fire scene simulation as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the urban fire management and control method based on fire scene simulation according to any one of claims 1 to 7 are executed.

Images