Distributed tunnel fire safety risk avoidance method and device

Abstract

A distributed tunnel fire safety risk avoidance method and device. The method comprises: acquiring section sensor data and tunnel section image data of a tunnel section (S110); performing fire identification on the tunnel section on the basis of the section sensor data and the tunnel section image data, and determining fire identification information (S120); when the fire identification information is a section fire location and a section fire hazard degree, considering there to be a fire in the tunnel section, and acquiring safety node information of the tunnel section (S130); performing section path planning on the basis of the safety node information and the fire identification information of the tunnel section, determining at least one piece of safety path information, and uploading the safety path information to a tunnel safety center (S140); and receiving a target evacuation route issued by the tunnel safety center, and performing risk avoidance prompting on the basis of the target evacuation route (S150).

Description

Distributed tunnel fire safety avoidance methods and devices

[0001] This application claims priority to Chinese Patent Application No. 202411459081.1, filed with the Chinese Patent Office on October 18, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of tunnel safety technology, such as a distributed tunnel fire safety avoidance method and apparatus. Background Technology

[0003] Tunnels, as passageways built in mountains, underground, and underwater, can improve transportation capacity. However, tunnels typically connect an exit and an entrance. In the event of a fire in a tunnel, although the tunnel fire alarm system and fire extinguishing system can be used to alert and control the fire, and orderly evacuation can also be carried out along the tunnel, the lack of data sharing between the fire alarm and fire extinguishing systems prevents the full utilization of data resources from each system. When planning evacuation routes for trapped personnel, it is impossible to effectively avoid various disaster situations in the tunnel, adapt to complex disaster changes, and meet actual needs. Summary of the Invention

[0004] This application provides a distributed tunnel fire safety avoidance method and device to solve the problem that tunnel safety systems are unable to provide optimal emergency evacuation and escape plans.

[0005] According to one aspect of this application, a distributed tunnel fire safety avoidance method is provided, applied to interval edge computing nodes, including:

[0006] Acquire section sensor data and tunnel section image data within the tunnel section; wherein, the section sensor data includes section temperature data, section humidity data, and section smoke data;

[0007] Fire identification is performed in the tunnel section based on interval sensor data and tunnel interval image data to determine fire identification information;

[0008] If the fire identification information is the location of the fire in the tunnel section and the degree of fire hazard in the tunnel section, then it is considered that there is a fire in the tunnel section, and the safety node information of the tunnel section is obtained.

[0009] Based on the safety node information and fire identification information of the tunnel section, the section path is planned to determine at least one safe path and the safe path information is uploaded to the tunnel safety center;

[0010] Receive the target evacuation route issued by the tunnel safety center and provide hazard avoidance prompts based on the target evacuation route.

[0011] According to another aspect of this application, a distributed tunnel fire safety and avoidance device is provided, applied to an inter-area edge computing node, including:

[0012] The data acquisition module is configured to acquire section sensor data and tunnel section image data within the tunnel section; wherein, the section sensor data includes section temperature data, section humidity data, and section smoke data;

[0013] The data fusion processing module is configured to perform fire identification in the tunnel section based on the interval sensor data and tunnel interval image data, and determine the fire identification information.

[0014] The fire identification module is configured to recognize the presence of a fire in the tunnel section and acquire the safety node information of the tunnel section if the fire identification information includes the location and severity of the fire in the section.

[0015] The section path planning module is set to plan the section path based on the safety node information and fire identification information of the tunnel section, determine at least one safe path information, and upload the safe path information to the tunnel safety center.

[0016] The section evacuation module is configured to receive the target evacuation route issued by the tunnel safety center and provide evacuation prompts based on the target evacuation route.

[0017] According to another aspect of this application, a distributed tunnel fire safety avoidance method is provided, applied to a tunnel safety center, including:

[0018] Receive at least one secure path information uploaded by at least one interval edge computing node;

[0019] Path planning is performed based on at least one safe path information to determine a safe evacuation route;

[0020] Obtain the location information of the trapped personnel, select a route from at least one safe evacuation route based on the location information of the trapped personnel, and determine the target evacuation route;

[0021] The target evacuation route is distributed to each interval edge computing node.

[0022] According to another aspect of this application, a distributed tunnel fire safety avoidance device is provided, applied in a tunnel safety center, including:

[0023] The data receiving module is configured to receive at least one secure path information uploaded by at least one interval edge computing node;

[0024] The edge computing node linkage module is configured to perform path planning based on at least one safe path information to determine a safe evacuation route.

[0025] The intelligent decision-making module is configured to obtain the location information of the trapped personnel, select a route from at least one safe evacuation route based on the location information of the trapped personnel, and determine the target evacuation route.

[0026] The data communication module is configured to send the target evacuation route to each interval edge computing node. Attached Figure Description

[0027] Figure 1 is a flowchart of a distributed tunnel fire safety avoidance method provided in an embodiment of this application;

[0028] Figure 2 is a flowchart of another distributed tunnel fire safety avoidance method provided in an embodiment of this application;

[0029] Figure 3 is a schematic diagram of a distributed tunnel fire safety and avoidance device provided in an embodiment of this application;

[0030] Figure 4 is a schematic diagram of another distributed tunnel fire safety and avoidance device provided in an embodiment of this application. Detailed Implementation

[0031] Figure 1 is a flowchart of a distributed tunnel fire safety avoidance method provided in an embodiment of this application. This embodiment is applicable to situations where controlled personnel in a tunnel are guided to safe routes during a tunnel fire. This method can be executed by a distributed tunnel fire safety avoidance device, which can be implemented in hardware and / or software and can be configured in an interval edge computing node. As shown in Figure 1, the method includes:

[0032] S110. Acquire the section sensor data and tunnel section image data of the tunnel section.

[0033] The tunnel section can be any of the road segments that make up the tunnel. It should be noted that the tunnel section can be divided according to the edge computing nodes of the section, and the maximum shooting range of the camera connected to the edge computing node and / or the data sensing range of the section sensor is taken as the tunnel section.

[0034] In some embodiments, the interval edge computing node is set in the tunnel. The interval edge computing node can communicate with the camera and the interval sensor and receive the data transmitted by the camera and the interval sensor. The interval edge computing node has data computing and processing capabilities, and can perform calculations on the data and also has the ability to process images and identify and analyze the image content.

[0035] In some embodiments, the section sensor can be a sensor installed inside the tunnel, capable of collecting data within the tunnel; the section sensor can consist of a temperature sensor, a humidity sensor, and a smoke concentration sensor. The section sensor data includes section temperature data, section humidity data, and section smoke data; the temperature of the tunnel section is collected by the temperature sensor in the section sensor to obtain section temperature data, the humidity of the tunnel section is collected by the humidity data in the section sensor to obtain section humidity data, and the smoke concentration of the tunnel section is collected by the smoke concentration sensor in the section sensor to obtain section smoke data.

[0036] In some embodiments, when monitoring temperature changes in a tunnel, the temperature sensor identifies the temperature change signal within the tunnel section, converts the signal into a digital signal, combines it with time information to form section temperature data, and sends the section temperature data to the edge computing node of the section. For example, the temperature sensor may be an infrared temperature sensor.

[0037] In some embodiments, the humidity sensor identifies the humidity of the tunnel section based on the water molecules in the air within the tunnel section, obtains the humidity signal of the tunnel section, performs digital conversion on the signal, combines it with time information to form the section humidity data, and sends the section humidity data to the edge computing node of the section.

[0038] In some embodiments, the smoke concentration sensor can detect the concentration of smoke particles within a tunnel section, identify the smoke concentration within the tunnel section, output a smoke concentration signal for the tunnel section, digitally convert the signal, combine it with time information to form section smoke data, and send the section smoke data to the edge computing node of the section. For example, the smoke concentration sensor can be a photoelectric smoke sensor or an ionization smoke sensor.

[0039] In some embodiments, at least one camera is installed in the tunnel section, which is capable of capturing and covering the entire area of ​​the tunnel section. The camera collects video streams of the tunnel section and converts the video streams of the tunnel section into image data of the tunnel section.

[0040] For example, the interval edge computing node acquires interval sensor data collected by sensors in the tunnel interval and tunnel interval image data collected by cameras.

[0041] S120. Based on the interval sensor data and tunnel interval image data, fire identification is performed in the tunnel interval to determine the fire identification information.

[0042] The fire identification information can be the identification result information of the tunnel section fire identified by the interval edge computing node. It should be noted that the interval edge computing node can judge the fire situation in the tunnel section based on interval sensor data and tunnel section image data, and obtain fire identification information. The fire identification information can be whether a fire has occurred or not. When a fire has occurred, the interval edge computing node will also locate the fire position in the tunnel section and determine the degree of fire hazard based on the fire intensity and fire source, thereby generating the interval fire location and interval fire hazard level, and using the interval fire location and interval fire hazard level as fire identification information; when no fire has occurred, the interval edge computing node directly uses the absence of a fire as fire identification information.

[0043] In some embodiments, the interval edge computing node performs fire identification based on interval sensor data and tunnel interval image data to determine whether a fire has occurred in the tunnel interval. If a fire has occurred, the location of the ignition point is located to obtain the interval fire location, and the degree of fire hazard in the interval is determined based on the fire intensity and fire source. The interval fire location and the degree of fire hazard in the interval are combined to form fire identification information. If no fire has occurred in the tunnel interval, the absence of a fire is used as fire identification information.

[0044] In other embodiments of this application, the step of identifying fires in the tunnel section based on interval sensor data and tunnel section image data, and determining fire identification information, includes:

[0045] Data fusion of interval sensor data is performed using a pre-established Bayesian network model to determine the first data fusion information;

[0046] Image processing of tunnel section image data is performed using a convolutional neural network model to determine the second data fusion information;

[0047] Fire identification information is determined based on the first data fusion information and the second data fusion information, wherein the fire identification information includes the location of the fire in the area and the degree of fire hazard in the area.

[0048] The Bayesian network model can be pre-set for data fusion of interval temperature data, interval humidity data, and interval smoke data. It should be noted that the Bayesian network model describes the dependencies between interval temperature data, interval humidity data, and interval smoke data in the interval sensor data, and then analyzes the impact of fire on interval temperature, interval humidity, and interval smoke concentration. The interval temperature data, interval humidity data, and interval smoke data are fused to identify whether a fire has occurred in the tunnel interval.

[0049] The first data fusion information can be the conditional probability distribution information obtained by fusing interval temperature data, interval humidity data, and interval smoke data. It should be noted that for the interval temperature data, interval humidity data, and interval smoke data, the data are synchronized according to their corresponding timestamps, integrating them onto the same time base. If there are missing data in the interval temperature data, interval humidity data, and interval smoke data, interpolation is used to fill the gaps, integrating the interval temperature data, interval humidity data, and interval smoke data into a unified dataset. At the same timestamp, corresponding interval temperature data, interval humidity data, and interval smoke data exist. Using a Bayesian network model, fire, interval temperature data, interval humidity data, and interval smoke data are defined as model variables, and their dependencies are determined, ultimately obtaining the conditional probability distribution information of the fused interval temperature data, interval humidity data, and interval smoke data.

[0050] The convolutional neural network (CNN) model can be a pre-trained image recognition model based on a CNN. In some embodiments, the CNN model comprises an input layer, a first convolutional network, a second convolutional network, a feature fusion network, and an output layer. The first convolutional network includes a first convolutional layer, an activation layer, a pooling layer, a fully connected layer, and an output layer, and is used to identify smoke concentration in the image. The second convolutional network includes a second convolutional layer, an activation layer, a pooling layer, a fully connected layer, and an output layer, and is used to identify illumination in the image. The feature fusion network includes a feature fusion layer. The fully connected layer uses historical images of the tunnel section as training data, assigning illumination labels for different illumination levels and smoke concentration labels for different smoke concentrations to the historical images. This constructs a model training set and a model test set. The CNN model is trained using the training set and tested using the test set. Backpropagation is performed on the CNN model based on the test error and loss function, ultimately completing the training of the CNN model.

[0051] In some embodiments, the convolutional neural network can perform smoke concentration recognition and illuminance recognition based on tunnel section image data. For smoke concentration recognition, an edge detection algorithm can be used to extract smoke edges in the image, and the smoke concentration feature vector is obtained by combining the smoke edges, local contrast changes, and smoke color changes in the image. For illuminance recognition, luminance extraction can be used to extract the luminance features of the image, and the flame region in the image can be extracted and identified by threshold segmentation. The fire region is segmented from the image, and the local luminance of the flame region is calculated to identify the local luminance distribution of the flame region, thus obtaining an image luminance feature vector. The smoke concentration feature vector and the image luminance feature vector are concatenated to obtain a multimodal feature vector. The multimodal feature vector is then used by a fully connected network to calculate the classification probability and identify the fire location, resulting in second data fusion information. The second data fusion information consists of the fire location and fire probability information of the tunnel section with the highest probability of fire occurrence in the tunnel section image data.

[0052] In some embodiments, after obtaining the first data fusion information and the second data fusion information, the first data fusion information and the second data fusion information are judged by fire prediction judgment indicators pre-set at the interval edge computing nodes. If both the first data fusion information and the second data fusion information are within the fire index interval where a fire occurs, then a fire is considered to have occurred in the tunnel interval. The location of the fire in the tunnel interval in the second data fusion information is taken as the location of the fire in the interval. Based on the location of the first data fusion information and the second data fusion information within the fire index interval, the degree of fire hazard in the interval is judged. For example, the fire index interval is divided into five equal intervals: the first interval, the second interval, the third interval, the fourth interval, and the fifth interval. Each interval corresponds to a degree of fire hazard. The fire hazard degrees corresponding to the first interval, the second interval, the third interval, the fourth interval, and the fifth interval are respectively: minor small-scale fire, mild medium-scale fire, moderate large-scale fire, severe large-scale fire, and extremely large-scale fire.

[0053] For example, data fusion is performed on the inter-tunnel sensor data using a pre-established Bayesian network model to determine the first data fusion information. Then, image processing is performed on the tunnel inter-tunnel image data using a convolutional neural network model to determine the second data fusion information. Finally, fire identification information is determined based on both the first and second data fusion information. This approach, based on the joint analysis of data fusion and heterogeneous data, can identify the location and severity of fires within the tunnel inter-tunnel, effectively improving the accuracy and efficiency of tunnel fire identification.

[0054] S130. If the fire identification information is the location of the fire in the tunnel section and the degree of fire hazard in the tunnel section, then it is considered that there is a fire in the tunnel section, and the safety node information of the tunnel section is obtained.

[0055] The safety node information includes safety node access information, safety equipment location information, and safety equipment status information. Safety node access information can indicate whether safety equipment is passable within the tunnel section. For example, the access information for a safety node can be the direction of passage through a fire door. Safety equipment location information can indicate the location of safety equipment within the tunnel section. For example, fire doors and fire extinguishers are marked on the map corresponding to the tunnel section, and their corresponding locations are displayed. Safety equipment status information can indicate the state of the safety equipment. For example, the state of a fire door can be open or closed; the state of a fire extinguisher can be usable or unusable; and the state of a fire hydrant can be that it has water and is usable or unusable.

[0056] In some embodiments, the tunnel safety equipment is equipped with a corresponding communication unit, which can upload the safety node information of the safety equipment to the inter-section edge computing node through the communication unit.

[0057] In some embodiments, if the fire identification information of the tunnel section in the interval edge computing node is the location of the fire in the interval and the degree of fire hazard in the interval, then the interval edge computing node should take safety precautions against the fire in the interval and obtain the safety node information of the tunnel section.

[0058] S140. Based on the safety node information and fire identification information of the tunnel section, perform section path planning, determine at least one safe path information, and upload the safe path information to the tunnel safety center.

[0059] The safe path information can be information on safe escape routes to avoid fire within tunnel sections. It should be noted that, due to the varying locations and severity of fires within different tunnel sections, at least one safe path can be planned for the same fire location and severity within the same tunnel section.

[0060] In some embodiments, when planning safe path information, the location and status information of safety equipment in the safety node information are identified by referring to the fire location and fire hazard level in the fire identification information.

[0061] In other embodiments of this application, the step of performing section path planning based on safety node information and fire identification information within the tunnel section to determine at least one safe path information includes:

[0062] Based on safety node information and the location of the fire in the tunnel section, safety nodes are screened within the tunnel section to determine at least one optional safety node; based on the optional safety nodes and their access information, path planning is performed to determine optional safety paths; based on the degree of fire hazard in the section and tunnel section image data, the optional safety paths are dynamically evaluated to determine at least one safe evacuation path; and based on each safe evacuation path, corresponding safety path information is determined.

[0063] Optional safety nodes can be safety nodes within the tunnel section that are unaffected by fire; safety nodes can be safety equipment within the tunnel section that is passable or usable, such as fire doors and fire extinguishers. It should be noted that the number of optional safety nodes within the tunnel section is affected by the location of the fire. During the selection process for safety nodes, by identifying the location of the fire within the section, it is determined whether the safety equipment within the tunnel section is located at the location of the fire. If the safety equipment is not located at the location of the fire, it can be considered an optional safety node; if the safety equipment is located at the location of the fire, it can be considered unusable.

[0064] The optional safe path can be a safe path that allows escape from the fire location within the section via a safe node. It should be noted that when the section edge computing node identifies an optional safe node, it identifies the passage information of each optional safe node, determines the passage status of the optional safe node, and performs path planning to obtain the optional safe path.

[0065] In some embodiments, the path planning method in this application embodiment may be as follows: select a passable optional safety node based on the safety node access information of the optional safety node, take the passable optional safety node as the planning starting point, connect the next passable optional safety node according to the direction of the passable optional safety node being away from the fire location in the section, use the connected passable optional safety node as the path planning transfer point, based on the path planning transfer point, connect the next optional safety node with the direction of the passable node being away from the fire location in the section, update the path planning transfer point to the newly connected passable optional safety node, and repeat the process of finding and connecting passable optional safety nodes until leaving the tunnel section to which the edge calculation node belongs, thereby obtaining at least one optional safety path.

[0066] The safe evacuation route is one that will not be affected by the fire within a preset evacuation time. It should be noted that fires in tunnel sections have the potential to spread and smoke to diffuse. Furthermore, the spread of fire and smoke can lead to factors such as increased air temperature and decreased oxygen levels within the tunnel. Therefore, when determining the available safe routes, it is necessary to assess the fire hazard level within the tunnel section, combined with tunnel image data, to evaluate the fire spread and smoke diffusion of the available safe routes. The preset evacuation time is used as the evaluation criterion to identify the safe routes that will not be affected by fire spread and smoke diffusion within the preset evacuation time, and these are then designated as safe evacuation routes. The preset evacuation time can be a fixed time pre-set by the calculation nodes at the tunnel edge and flexibly adjusted according to the number of people trapped in the tunnel section.

[0067] In some embodiments, after obtaining the degree of fire hazard corresponding to the location of the fire in the section, the edge computing node of the section continuously monitors the location of the fire in the section through tunnel section image data, continuously identifies the smoke concentration and location of the fire in the tunnel section through a convolutional neural model, and then calculates the rate of change of the location of the fire in the section and the change of smoke concentration. Based on the rate of change of the location of the fire in the section and the change of smoke concentration, the optional safe routes are evaluated, and a safe evacuation route that is not affected by the fire within a preset evacuation time is identified; wherein, the number of safe evacuation routes is at least one.

[0068] In other embodiments of this application, the corresponding safe evacuation path information will be determined based on each safe evacuation path, including:

[0069] Obtain the section identifier of the tunnel section, and combine each safe evacuation path and the section identifier into messages to determine the safe path information corresponding to the safe evacuation path.

[0070] The interval identifier can be the identification information of the tunnel interval. For example, the interval identifier can be information such as the identification number and identification code of the tunnel interval.

[0071] In some embodiments, the interval identifier can be used to identify interval edge computing nodes.

[0072] For example, after obtaining at least one safe evacuation path, the interval edge computing node combines each safe evacuation path with the interval identifier to obtain the safe path information corresponding to each safe evacuation path, and uploads all the safe path information to the tunnel safety center.

[0073] S150: Receive the target evacuation route issued by the tunnel safety center, and provide hazard avoidance prompts based on the target evacuation route.

[0074] The target evacuation route can be an evacuation route that guides trapped personnel out of the tunnel section. It should be noted that the target evacuation route is the optimal evacuation route for trapped personnel from their current location through various tunnel sections. The target evacuation route is the route by which trapped personnel escape from their current tunnel position.

[0075] In some embodiments, corresponding rescue safety devices are set up in the tunnel section. The controlled personnel can perform evacuation and rescue operations according to the rescue safety devices. The rescue safety devices will upload their own location, i.e. the location of the trapped personnel, to the section edge computing node and the tunnel safety center. The tunnel safety center determines the target evacuation route based on the location of the controlled personnel.

[0076] For example, after receiving the target evacuation route, the interval edge computing node identifies the evacuation route corresponding to the current tunnel interval within the target evacuation route. Based on the evacuation route corresponding to the current tunnel interval, it generates a hazard avoidance prompt. This prompt is then displayed on the broadcast and display modules of the rescue safety equipment via the interval edge computing node to guide trapped personnel to evacuate the current tunnel interval in an orderly manner. The hazard avoidance prompt can be either a voice prompt indicating the evacuation route or a display showing the evacuation route.

[0077] The technical solution of this application embodiment sets up edge computing nodes in the tunnel section, and connects sensors and cameras in the tunnel section through the edge computing nodes. By acquiring sensor data and image data from the sensors, the edge computing nodes can identify the fire situation in the tunnel section and calculate safe routes. The calculated safe routes are sent to the tunnel safety center, and the safe routes of each edge computing node are linked for planning. The safest evacuation path can be planned, and safety guidance can be provided to trapped personnel. This solves the technical problem in related technologies that tunnel safety systems are unable to provide the best emergency evacuation and escape plan. It can integrate the data of edge nodes to obtain the best emergency evacuation and escape plan to adapt to the complex disaster changes in the tunnel, and improve the flexibility and adaptability of the system.

[0078] Figure 2 is a flowchart illustrating another distributed tunnel fire safety and avoidance method provided in this embodiment of the application. This embodiment is applicable to situations where controlled personnel in a tunnel are guided to safe routes during a tunnel fire. This method can be executed by a distributed tunnel fire safety and avoidance device, which can be configured in a tunnel safety center. As shown in Figure 2, the method includes:

[0079] S210. Receive at least one secure path information uploaded by at least one interval edge computing node.

[0080] For example, the tunnel safety center receives at least one safe path information uploaded by each section edge computing node, and performs message parsing on each safe path information of each section edge computing node to obtain the section identifier and safe evacuation path of each section edge computing node.

[0081] S220. Based on at least one safe path information, perform path planning to determine a safe evacuation route.

[0082] The safe evacuation route can be an evacuation route away from the tunnel section where the fire occurred. It should be noted that the tunnel associated with the tunnel safety is a complete tunnel, which is divided into multiple tunnel sections. Each tunnel section has a corresponding section identifier and section edge computing center. The safe evacuation route is obtained by connecting the safe path information uploaded by the section edge computing center of each tunnel section according to the location of the tunnel section.

[0083] For example, the tunnel safety center receives multiple safe evacuation paths uploaded by all the edge computing centers within the tunnel section, integrates these multiple safe evacuation paths, and connects the safe evacuation paths corresponding to each tunnel section into a single safe evacuation route.

[0084] In other embodiments of this application, the step of planning a route based on at least one safe path information to determine a safe evacuation route includes:

[0085] Each safe path information is parsed and identified sequentially to determine the safe edge node of the safe evacuation path corresponding to each interval identifier;

[0086] The safety nodes at the edge of the safe evacuation path corresponding to each section are used as the path planning information for the tunnel section. Path planning is performed based on the path planning information of each tunnel section to determine at least one safe evacuation route.

[0087] Among them, the section edge safety node can be a safety node located at the edge of the tunnel section. It should be noted that the position of the section edge safety node is located adjacent to other tunnel sections. Trapped personnel can leave or enter the current tunnel section through the section edge safety node.

[0088] In some embodiments, when connecting the safe evacuation paths of each tunnel section, it is necessary to obtain the safe nodes adjacent to each tunnel section and other tunnel sections. By identifying the location of the passable optional safe nodes in each safe path information, the location of each passable optional safe node is obtained. By comparing with the section edge of the tunnel section, the section edge safe nodes in each tunnel section are obtained.

[0089] The route planning information is composed of the safety nodes at the edge of each safe evacuation path corresponding to each tunnel section. By connecting the safety nodes at the edge of adjacent tunnel sections according to the location of each tunnel section, an evacuation path connecting all tunnel sections of the entire tunnel can be obtained, which is then used as a safe evacuation route.

[0090] For example, by analyzing and identifying the location of the passable optional safe nodes in the safe evacuation path of each safe path information, the safe edge safe node of each safe evacuation path in the tunnel section is determined. Then, the safe edge safe node of the safe evacuation path corresponding to each section identifier is used as the path planning information of the tunnel section. By using the positional relationship of each tunnel section, the safe edge safe nodes of each adjacent tunnel section are used for path planning to obtain at least one safe evacuation route.

[0091] In other embodiments of this application, the step of determining at least one safe evacuation route based on the path planning information of each tunnel section includes:

[0092] Select an initial planning section from at least one tunnel section;

[0093] Based on the section identifiers and route planning information of the initial planning section, route planning is carried out to determine at least one safe evacuation route.

[0094] The initial planning section can be used as the starting tunnel section for planning safe evacuation routes. It should be noted that the planning of safe evacuation routes is carried out gradually based on the adjacent tunnel sections. Thus, any tunnel section can be selected as the starting planning section. Usually, in order to improve planning efficiency, the tunnel section where the fire occurred is selected as the starting planning section. During the planning process, after the initial planning section is completed, it is necessary to update the initial planning section and plan a new starting planning section.

[0095] For example, starting from the initial zone, adjacent tunnel zones are selected based on the zone identifier of the initial planning zone, and path connection planning is performed based on the zone edge safety nodes of the adjacent tunnel zones and the zone edge safety nodes of the initial tunnel zone to obtain at least one first safe evacuation route.

[0096] In other embodiments of this application, the step of determining at least one safe evacuation route based on the section identifier and route planning information of the initial planning section includes:

[0097] Select at least one adjacent tunnel section as the path planning section based on the section identifier of the initial planning section. Perform path planning based on the path planning information of the initial planning section and the path planning section to determine at least one planned evacuation route.

[0098] If there are adjacent tunnel intervals in the path planning interval, update the path planning interval to the starting planning interval, update the adjacent tunnel interval to the path planning interval, and return to execute the path planning operation based on the path planning information of the starting planning interval and the path planning interval until there are no adjacent tunnel intervals in the path planning interval.

[0099] Each planned evacuation route will be designated as a safe evacuation route.

[0100] The path planning interval can be used for tunnel sections in path planning. It should be noted that the path planning interval is the tunnel section adjacent to the initial planning interval.

[0101] The planned evacuation route can be an evacuation route that is currently in the path planning process. It should be noted that the planned evacuation route can be route information from the tunnel safety center during the path planning process. The tunnel safety center will determine the planned evacuation route as the safe evacuation route once the path planning is completed.

[0102] For example, at least one adjacent tunnel interval is selected as the path planning interval based on the interval identifier of the initial planning interval. The safety nodes of the safe evacuation path of the initial planning interval and the safety nodes of the safe evacuation path of each path planning interval are connected to obtain at least one planned evacuation route. Based on the interval identifier of each path planning interval, it is determined whether there are adjacent tunnel intervals in each path planning interval. If there is an unconnected adjacent tunnel interval in a path planning interval, it means that the path planning is not completed. The path planning interval is updated to the initial planning interval, and the adjacent tunnel intervals of the path planning interval are updated to the path planning interval. The operation of path planning based on the path planning information of the initial planning interval and the path planning interval is returned to be executed. If there are no adjacent tunnel intervals in the path planning interval, it means that the path planning is completed, and each planned evacuation route is determined as a safe evacuation route.

[0103] S230. Obtain the location information of the trapped personnel, select a route from at least one safe evacuation route based on the location information of the trapped personnel, and determine the target evacuation route.

[0104] Among them, the personnel location information can be the location information of the rescue safety equipment used by trapped personnel to carry out evacuation and rescue operations.

[0105] For example, after obtaining the location information of trapped personnel in tunnel safety, the tunnel section corresponding to the controlled personnel is determined based on the location information of the trapped personnel, and the nearest passable optional safety node is identified. Based on the passable optional safety node, the optimal path is matched in the safe evacuation route, and the nearest safe evacuation route is selected as the target evacuation route.

[0106] In some embodiments, the process of optimal path matching in a safe evacuation route based on passable alternative safe nodes is as follows: take the passable alternative safe nodes as the starting safe nodes, take all safe evacuation routes including the starting safe nodes as candidate safe evacuation routes, calculate the route length of each candidate safe evacuation route, and then take the safe evacuation route with the shortest route among the candidate safe evacuation routes as the target evacuation route.

[0107] In some embodiments, after obtaining the target evacuation route, the tunnel safety center monitors the on / off status of each optional safety node in the target evacuation route. When an optional safety node is detected to be closed, it is opened remotely to ensure the smooth passage of the target evacuation route.

[0108] S240. Distribute the target evacuation route to each interval edge calculation node.

[0109] For example, the target evacuation route is distributed to each interval edge computing node.

[0110] The technical solution of this application embodiment sets up edge computing nodes in the tunnel section, and connects sensors and cameras in the tunnel section through the edge computing nodes. By acquiring sensor data and image data from the sensors, the edge computing nodes can identify the fire situation in the tunnel section and calculate safe routes. The calculated safe routes are sent to the tunnel safety center, and the safe routes of each edge computing node are linked for planning. The safest evacuation path can be planned, and safety guidance can be provided to trapped personnel. This solves the technical problem in related technologies that tunnel safety systems are unable to provide the best emergency evacuation and escape plan. It can integrate the data of edge nodes to obtain the best emergency evacuation and escape plan to adapt to the complex disaster changes in the tunnel, and improve the flexibility and adaptability of the system.

[0111] Figure 3 is a schematic diagram of a distributed tunnel fire safety and avoidance device provided in an embodiment of this application. As shown in Figure 3, the device includes: a data acquisition module 310, a data fusion processing module 320, a fire identification module 330, an interval path planning module 340, and an interval evacuation module 350; wherein,

[0112] The data acquisition module 310 is configured to acquire section sensor data and tunnel section image data of the tunnel section; wherein, the section sensor data includes section temperature data, section humidity data and section smoke data;

[0113] The data fusion processing module 320 is configured to perform fire identification in the tunnel section based on the interval sensor data and tunnel interval image data, and determine the fire identification information.

[0114] The fire identification module 330 is configured to consider that there is a fire in the tunnel section when the fire identification information is the location of the fire in the section and the degree of fire hazard in the section, and to obtain the safety node information of the tunnel section.

[0115] The section path planning module 340 is configured to perform section path planning based on the safety node information and fire identification information of the tunnel section, determine at least one safe path information, and upload the safe path information to the tunnel safety center.

[0116] The section evacuation module 350 is configured to receive the target evacuation route issued by the tunnel safety center and provide evacuation prompts based on the target evacuation route.

[0117] The technical solution of this application embodiment sets up edge computing nodes in the tunnel section, and connects sensors and cameras in the tunnel section through the edge computing nodes. By acquiring sensor data and image data from the sensors, the edge computing nodes can identify the fire situation in the tunnel section and calculate safe routes. The calculated safe routes are sent to the tunnel safety center, and the safe routes of each edge computing node are linked for planning. The safest evacuation path can be planned, and safety guidance can be provided to trapped personnel. This solves the technical problem in related technologies that tunnel safety systems are unable to provide the best emergency evacuation and escape plan. It can integrate the data of edge nodes to obtain the best emergency evacuation and escape plan to adapt to the complex disaster changes in the tunnel, and improve the flexibility and adaptability of the system.

[0118] In some embodiments, the data fusion processing module is configured as follows:

[0119] Data fusion of interval sensor data is performed using a pre-established Bayesian network model to determine the first data fusion information;

[0120] Image processing of tunnel section image data is performed using a convolutional neural network model to determine the second data fusion information;

[0121] Fire identification information is determined based on the first data fusion information and the second data fusion information, wherein the fire identification information includes the location of the fire in the area and the degree of fire hazard in the area.

[0122] In some embodiments, the interval path planning module is configured as follows:

[0123] Based on the safety node information and the location of the fire in the tunnel section, safety nodes are screened within the tunnel section to determine at least one optional safety node;

[0124] Path planning is performed based on available safe nodes and safe node access information to determine available safe paths;

[0125] Based on the degree of fire hazard in the section and the tunnel section image data, the available safe routes are dynamically evaluated to determine at least one safe evacuation route;

[0126] Determine the corresponding safe path information based on each safe evacuation route.

[0127] In some embodiments, the interval path planning module is configured as follows:

[0128] Obtain the section identifier of the tunnel section, and combine each safe evacuation path and the section identifier into messages to determine the safe path information corresponding to the safe evacuation path.

[0129] The distributed tunnel fire safety and avoidance device provided in this application embodiment can execute the distributed tunnel fire safety and avoidance method provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects of the execution method.

[0130] Figure 4 is a schematic diagram of another distributed tunnel fire safety and avoidance device provided in an embodiment of this application. As shown in Figure 4, the device includes: a data receiving module 410, an edge computing node linkage module 420, an intelligent decision-making module 430, and a data communication module 440; wherein,

[0131] The data receiving module 410 is configured to receive at least one secure path information uploaded by at least one interval edge computing node;

[0132] The edge computing node linkage module 420 is configured to perform path planning based on at least one safe path information to determine a safe evacuation route.

[0133] The intelligent decision-making module 430 is configured to obtain the location information of the trapped personnel, select a route from at least one safe evacuation route based on the location information of the trapped personnel, and determine the target evacuation route.

[0134] The data communication module 440 is configured to send the target evacuation route to each interval edge computing node.

[0135] The technical solution of this application embodiment sets up edge computing nodes in the tunnel section, and connects sensors and cameras in the tunnel section through the edge computing nodes. By acquiring sensor data and image data from the sensors, the edge computing nodes can identify the fire situation in the tunnel section and calculate safe routes. The calculated safe routes are sent to the tunnel safety center, and the safe routes of each edge computing node are linked for planning. The safest evacuation path can be planned, and safety guidance can be provided to trapped personnel. This solves the technical problem in related technologies that tunnel safety systems are unable to provide the best emergency evacuation and escape plan. It can integrate the data of edge nodes to obtain the best emergency evacuation and escape plan to adapt to the complex disaster changes in the tunnel, and improve the flexibility and adaptability of the system.

[0136] In some embodiments, the edge computing node linkage module is configured as follows:

[0137] Each safe path information is parsed and identified sequentially to determine the safe edge node of the safe evacuation path corresponding to each interval identifier;

[0138] The safety nodes at the edge of the safe evacuation path corresponding to each section are used as the path planning information for the tunnel section. Path planning is performed based on the path planning information of each tunnel section to determine at least one safe evacuation route.

[0139] In some embodiments, the edge computing node linkage module is configured as follows:

[0140] Select an initial planning section from at least one tunnel section;

[0141] Based on the section identifiers and route planning information of the initial planning section, route planning is carried out to determine at least one safe evacuation route.

[0142] In some embodiments, the edge computing node linkage module is configured as follows:

[0143] Select at least one adjacent tunnel section as the path planning section based on the section identifier of the initial planning section. Perform path planning based on the path planning information of the initial planning section and the path planning section to determine at least one planned evacuation route.

[0144] If there are adjacent tunnel intervals in the path planning interval, update the path planning interval to the starting planning interval, update the adjacent tunnel interval to the path planning interval, and return to execute the path planning operation based on the path planning information of the starting planning interval and the path planning interval until there are no adjacent tunnel intervals in the path planning interval.

[0145] Each planned evacuation route will be designated as a safe evacuation route.

[0146] The distributed tunnel fire safety and avoidance device provided in this application embodiment can execute the distributed tunnel fire safety and avoidance method provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects of the execution method.

Claims

1. A distributed tunnel fire safety avoidance method, applied to inter-area edge computing nodes, including; Obtain interval sensor data and tunnel interval image data of a tunnel interval; wherein The interval sensor data includes interval temperature data, interval humidity data, and interval smoke data; Fire identification is performed on the tunnel section based on the interval sensor data and the tunnel section image data to determine fire identification information; If the fire identification information is the location of a fire in a tunnel section and the degree of fire hazard in that section, then it is considered that there is a fire in the tunnel section, and the safety node information of the tunnel section is obtained. Based on the safety node information and fire identification information of the tunnel section, section path planning is performed to determine at least one safe path information, and the safe path information is uploaded to the tunnel safety center; Receive the target evacuation route issued by the tunnel safety center and provide hazard avoidance prompts based on the target evacuation route.

2. The method of claim 1, wherein, The step of identifying fires in the tunnel section based on the interval sensor data and the tunnel section image data, and determining fire identification information, includes: The data from the interval sensors are fused using a pre-established Bayesian network model to determine the first data fusion information; Image processing is performed on the tunnel section image data using a convolutional neural network model to determine the second data fusion information; Fire identification information is determined based on the first data fusion information and the second data fusion information, wherein the fire identification information includes the location of the fire in the area and the degree of fire hazard in the area.

3. The method of claim 2, wherein, The safety node information includes safety node access information, safety equipment location information, and safety equipment status information. The step of performing interval path planning based on the safety node information and fire identification information within the tunnel section to determine at least one safety path includes: Based on the safety node information and the location of the fire in the tunnel section, safety nodes are screened within the tunnel section to determine at least one selectable safety node; Based on the available safe nodes and their access information, a path is planned to determine the available safe paths. The optional safe routes are dynamically evaluated based on the fire hazard level of the section and the tunnel section image data to determine at least one safe evacuation route; The corresponding safe evacuation path information is determined based on each of the safe evacuation paths.

4. The method of claim 3, wherein, The step of determining the corresponding safe path information based on each safe evacuation path includes: Obtain the interval identifier of the tunnel section, and combine each safe evacuation path and the interval identifier into messages to determine the safe path information corresponding to the safe evacuation path.

5. A distributed tunnel fire safety avoidance method, applied to a tunnel safety center, including: Receive at least one secure path information uploaded by at least one interval edge computing node; Based on at least one of the aforementioned safe path information, route planning is performed to determine a safe evacuation route; Obtain the location information of the trapped personnel, select a route from at least one of the safe evacuation routes based on the location information of the trapped personnel, and determine the target evacuation route; The target evacuation route is distributed to each of the interval edge computing nodes.

6. The method of claim 5, wherein, The step of planning a route based on at least one of the safe path information to determine a safe evacuation route includes: Each piece of safe path information is parsed and identified sequentially to determine the safe edge node of the safe evacuation path corresponding to each interval identifier; The safety node at the edge of the safe evacuation path corresponding to each of the interval identifiers is used as the path planning information of the tunnel interval. Path planning is performed based on the path planning information of each tunnel interval to determine at least one safe evacuation route.

7. The method of claim 6, wherein, The step of determining at least one safe evacuation route based on the path planning information of each tunnel section includes: Select an initial planning section from at least one of the tunnel sections; Based on the interval identifier and the route planning information of the initial planning interval, route planning is performed to determine at least one safe evacuation route.

8. The method of claim 7, wherein, The step of determining at least one safe evacuation route by performing route planning based on the interval identifier of the initial planning interval and the route planning information includes: Based on the interval identifier of the initial planning interval, at least one adjacent tunnel interval is selected as the path planning interval. Based on the path planning information of the initial planning interval and the path planning interval, path planning is performed to determine at least one planned evacuation route. If the adjacent tunnel interval exists in the path planning interval, update the path planning interval to the starting planning interval, update the adjacent tunnel interval to the path planning interval, and return to perform the path planning operation based on the path planning information of the starting planning interval and the path planning interval until the adjacent tunnel interval does not exist in the path planning interval. Each of the planned evacuation routes is designated as a safe evacuation route.

9. A distributed tunnel fire safety and avoidance device, applied to inter-area edge computing nodes, comprising: The data acquisition module is configured to acquire interval sensor data and tunnel interval image data of a tunnel interval; wherein The interval sensor data includes interval temperature data, interval humidity data, and interval smoke data; The data fusion processing module is configured to perform fire identification on the tunnel section based on the interval sensor data and the tunnel section image data, and determine fire identification information. The fire identification module is configured to determine that a fire exists in the tunnel section when the fire identification information is the location of the fire in the section and the degree of fire hazard in the section, and to obtain the safety node information of the tunnel section. The section path planning module is configured to perform section path planning based on the safety node information and fire identification information of the tunnel section, determine at least one safe path information, and upload the safe path information to the tunnel safety center; The section evacuation module is configured to receive the target evacuation route issued by the tunnel safety center and provide hazard avoidance prompts based on the target evacuation route.

10. A distributed tunnel fire safety and hazard avoidance device, applied in a tunnel safety center, comprising: The data receiving module is configured to receive at least one secure path information uploaded by at least one interval edge computing node; The edge computing node linkage module is configured to perform path planning based on at least one of the aforementioned safe path information to determine a safe evacuation route. The intelligent decision-making module is configured to acquire the location information of the trapped personnel, select a route from at least one of the safe evacuation routes based on the location information of the trapped personnel, and determine the target evacuation route. The data communication module is configured to send the target evacuation route to each of the interval edge computing nodes.

Images