Fire emergency evacuation indication method applied to large indoor scene and related device

Abstract

The invention discloses a fire emergency evacuation indication method applied to a large indoor scene and a related device, relates to the field of artificial intelligence, and obtains a personnel evacuation network flow model, a density matrix, a speed matrix and an evacuation time consumption matrix. And inputting the parameters into a pre-constructed evacuation path prediction model, and obtaining first escape paths corresponding to all nodes through the evacuation path prediction model. The personnel can escape based on the first escape path; as time goes by, fire detailed information of nodes where people do not pass through in the first escape path may change, the first escape path is not the most suitable escape path, at the moment, updated parameters need to be obtained and input into an evacuation path prediction model, and a second escape path is obtained through the evacuation path prediction model; in this way, people can escape through the second escape path.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a fire emergency evacuation indication method and related devices applied to large indoor scenes. Background Art

[0002] Compared with other disasters, fires have the characteristics of high frequency, high spread of disasters, and strong smoke hazards, causing great losses to people's lives and property. Large indoor scenes, such as large shopping malls and bank branches, especially bank branches with vaults, business halls, teller offices and other areas, have many flammable items, densely populated areas, and complex building structures, and are high-risk places for fires. Once a fire occurs, people may choose unreasonable escape routes out of panic, delaying evacuation time and causing serious loss of life and property.

[0003] Therefore, an emergency evacuation indication method applicable to large indoor scenes is needed to guide employees and customers in large indoor scenes to escape safely and efficiently, so as to minimize the loss of life and property. Summary of the invention

[0004] In view of the above problems, the present application provides a fire emergency evacuation indication method and related devices applied to large indoor scenes to achieve the purpose of automatically generating escape routes. The specific scheme is as follows:

[0005] The first aspect of the present application provides a fire emergency evacuation indication method applied to a large indoor scene, comprising:

[0006] Obtaining a personnel evacuation network flow model, wherein the personnel evacuation network flow model includes a plurality of nodes and edges between two nodes; the edges between two nodes represent passages between two nodes for personnel to walk; the plurality of nodes include escape exit nodes, ordinary nodes, and stair nodes;

[0007] Obtaining a distance matrix corresponding to the personnel evacuation network flow model; the distance matrix includes distances between all pairs of nodes with edges in the personnel evacuation network flow model;

[0008] If a fire is detected, a density matrix corresponding to the personnel evacuation network flow model is obtained; the density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges;

[0009] Determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs;

[0010] If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, determining that the edge between the two nodes is a congested section;

[0011] Updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model;

[0012] Based on the density matrix, a speed matrix corresponding to the personnel evacuation network flow model is obtained; the speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model;

[0013] Based on the distance matrix and the speed matrix, obtaining an evacuation time consumption matrix corresponding to the personnel evacuation network flow model;

[0014] Inputting the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtaining the first escape paths corresponding to all nodes respectively through the evacuation path prediction model;

[0015] For each of the nodes, sending the first escape path corresponding to the node to a terminal for escape corresponding to the node;

[0016] If it is detected that the density of human traffic corresponding to any of the nodes changes, or the density of human traffic in the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation at any of the nodes changes, an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix are obtained;

[0017] Inputting the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix and the updated evacuation time matrix into the evacuation path prediction model, and obtaining a second escape path through the evacuation path prediction model;

[0018] For each of the nodes, the second escape path corresponding to the node is sent to a terminal for escape corresponding to the node.

[0019] In an optional implementation, the method further includes:

[0020] For each of the nodes, a smoke sensor deployed at the node is used to detect whether the node has changed from a disaster location to a non-disaster location, or from a non-disaster location to a disaster location. If so, it is determined that the fire situation of the node has changed;

[0021] For each of the nodes, a sensor deployed at the node is used to detect a density of human traffic at the node. If the human traffic density changes from being greater than or equal to the first preset density value to being less than the first preset density value, or from being less than the first preset density value to being greater than or equal to the first preset density value, it is determined that the human traffic density of the node has changed.

[0022] For each pairwise edge of the nodes, the density of human traffic located in the passage is detected by using sensors deployed in the passage between the pairs of nodes. If the human traffic density changes from being greater than or equal to a second preset density value to being less than the second preset density value, or from being less than the second preset density value to being greater than or equal to the second preset density value, it is determined that the human traffic density of the passage has changed.

[0023] For each pairwise node edge, the smoke concentration of the passage is detected by a smoke sensor deployed in the passage between the two nodes; if the smoke concentration level corresponding to the smoke concentration changes, it is determined that the moving speed of the personnel in the passage between the two nodes has changed.

[0024] In an optional implementation, the smoke concentration is CO concentration, and for each edge of two nodes, a method for obtaining an updated moving speed corresponding to a path between the two nodes includes:

[0025] By formula The coefficient α is calculated; ρ co is the CO concentration of the pathway between the two nodes;

[0026] The updated moving speed is determined to be the product of the moving speed before updating and the coefficient α.

[0027] In an optional implementation, the evacuation path prediction model is a dual deep Q network model, and the method for training the dual deep Q network includes:

[0028] Obtaining a sample density matrix corresponding to the personnel evacuation network flow model; the sample density matrix includes the flow density of people corresponding to all nodes in the personnel evacuation network flow model and the flow density of people between nodes with edges;

[0029] Based on the sample density matrix, a sample speed matrix corresponding to the personnel evacuation network flow model is obtained; the sample speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model;

[0030] Based on the sample distance matrix and the sample speed matrix, obtaining a sample evacuation time consumption matrix corresponding to the personnel evacuation network flow model;

[0031] Determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs;

[0032] If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, determining that the edge between the two nodes is a congested section;

[0033] Updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model;

[0034] Acquire state information of the agent, wherein the state information includes the time taken to move from the first node to each of the second nodes;

[0035] The personnel evacuation network flow model, the distance matrix, the sample density matrix, the sample speed matrix, the sample evacuation time matrix and the state information are input into the dual deep Q network model, and the predicted Q values ​​corresponding to multiple actions are output through the dual deep Q model; the dual deep Q network model includes a deep Q network model and a target Q network model; the action is the action of the agent moving from the first node to the second node;

[0036] Get a random number;

[0037] If the random number is greater than a preset threshold, randomly selecting a target action from the multiple actions, and the second node in the target action is the target second node;

[0038] If the random number is less than or equal to the preset threshold, a target action corresponding to the maximum expected Q value is selected from the multiple actions; the second node in the target action is the target second node;

[0039] Determine a first reward for the agent to move to the target second node based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken for the agent to reach the target second node from the initial node exceeds a preset prescribed time length;

[0040] Inputting the first reward and the predicted Q value corresponding to the target action into the dual deep Q network model to update the deep Q network model in the dual deep Q network model;

[0041] If the target second node is not the disaster node and is not the escape exit node, and the target time does not exceed the preset specified time, return to the step of obtaining the state information of the intelligent agent;

[0042] If the target second node is the disaster node or the target time exceeds the preset specified time, determine that the agent returns to the initial node, and return to the step of obtaining the state information of the agent;

[0043] If the target second node is the escape exit node, obtain an escape planning path, where the escape planning path is a path for the agent to reach the escape exit node from the initial node;

[0044] Determining a second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path;

[0045] Inputting the second reward and the predicted Q value corresponding to the target action into the deep Q network model to update the deep Q network model;

[0046] Obtaining a target real Q value of an action of moving from a previous node to a next node in each pair of nodes having an edge in the escape planning path;

[0047] If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path increase successively according to the escape order, it is determined that the escape planning path is the optimal path;

[0048] If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path do not increase in sequence according to the escape sequence, it is determined that the agent returns to the initial node, and the step of obtaining the state information of the agent is returned;

[0049] If the update number of the deep Q network model reaches a preset update number, assigning the parameters of the deep Q network model to the target Q network model;

[0050] If the optimal paths corresponding to all nodes have been obtained, it is determined that the dual deep Q network model training is completed.

[0051] In an optional implementation, the first reward step for the agent to move to the target second node based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken for the agent to reach the target second node from the initial node exceeds a preset prescribed time, includes:

[0052] If the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset specified time, the formula , calculate R t1 ; Determine R t1 is the first reward; k1 is the preset reward magnification factor, k1 is a positive number greater than 0, means the shortest distance; refers to the shortest distance between the first node and the escape exit;

[0053] If the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node, a second preset reward value is determined to be the first reward.

[0054] In an optional implementation, the step of determining the second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path includes:

[0055] If the target second node is the escape exit node and the target time is less than or equal to the preset time, the formula , calculate R -1 ; Determine R -1 For the second reward;

[0056] Wherein, k2 is a preset scaling factor, k3 is a preset time coefficient, and t2 is the total time taken by the agent to travel from the initial node to the escape exit node;

[0057] in, , ,in is the three-dimensional coordinate of the initial node; The three-dimensional coordinates of the escape exit node closest to the initial node; is the three-dimensional coordinate of the jth node in the escape planning path; N is the total number of all nodes in the escape planning path; ,in, Refers to the number of dangerous nodes in the escape planning path; Refers to the number of disaster nodes in the escape planning path;

[0058] in, , It refers to the three-dimensional coordinates of the starting node of the i-th congested section; It refers to the three-dimensional coordinates of the terminal node of the i-th congested section.

[0059] In an optional implementation, the target reward is the first reward or the second reward, and the target reward and the predicted Q value corresponding to the target action are input into the dual deep Q network model to update the deep Q network model in the dual deep Q network model, comprising:

[0060] If the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset specified time, the formula , calculate the real Q value of the agent moving from the first node to the target second node;

[0061] in, is the true Q value, is the predicted Q value corresponding to the movement of the agent from the first node to the target second node output by the dual Q network model; lr is the preset learning rate parameter, γ is the preset discount factor, is the preset function;

[0062] If the target second node is the escape exit node and the target time is less than or equal to the preset specified time, determine R -1 is the actual Q value of the agent moving from the first node to the target second node;

[0063] If the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node, the second preset reward value is determined to be the real Q value of the agent moving from the first node to the target second node;

[0064] By formula , update the deep Q network model.

[0065] In an optional implementation, the deep Q network model includes a first CSP module, a first CBS module, a second CSP module, a second CBS module, a third CSP module, and a third CBS module connected in sequence;

[0066] The first CSP module includes: a module carrying an unfold function, a fourth CBS module connected to a first output end of the module carrying the unfold function, a first BottleNeck module connected to a second output end of the module carrying the unfold function, and a second BottleNeck module connected to a third output end of the module carrying the unfold function; a fifth CBS module connected to an output end of the fourth CBS module, an output end of the first BottleNeck module, and an output end of the second BottleNeck module, respectively, and a sixth CBS module connected to the fifth CBS module;

[0067] The first CBS module includes a convolutional layer, a batch normalization layer connected to the convolutional layer, and an activation function layer connected to the batch normalization layer;

[0068] Among them, the first CSP module, the second CSP module and the third CSP module have the same structure; the first CBS module, the second CBS module, the third CBS module, the fourth CBS module, the fifth CBS module and the sixth CBS module have the same structure.

[0069] The second aspect of the present application provides a fire emergency evacuation indication device applied to a large indoor scene, comprising:

[0070] A first acquisition module is used to acquire a personnel evacuation network flow model, wherein the personnel evacuation network flow model includes a plurality of nodes and edges between two nodes; the edges between two nodes represent passages between two nodes for personnel to walk; the plurality of nodes include escape exit nodes, ordinary nodes, and stair nodes;

[0071] A second acquisition module is used to acquire a distance matrix corresponding to the personnel evacuation network flow model; the distance matrix includes distances between all pairs of nodes with edges in the personnel evacuation network flow model;

[0072] A third acquisition module is used to acquire a density matrix corresponding to the personnel evacuation network flow model if a fire is detected; the density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges;

[0073] A first determination module is used to determine a dangerous node and a disaster node from the plurality of nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs;

[0074] A second determination module is used to determine that the edge between the two nodes is a congested section if the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value;

[0075] A first updating module, used for updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model;

[0076] A fourth acquisition module is used to acquire a speed matrix corresponding to the personnel evacuation network flow model based on the density matrix; the speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model;

[0077] A fifth acquisition module, configured to acquire an evacuation time consumption matrix corresponding to the personnel evacuation network flow model based on the distance matrix and the speed matrix;

[0078] A sixth acquisition module, used to input the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtain the first escape paths corresponding to all nodes respectively through the evacuation path prediction model;

[0079] A first sending module, configured to send, for each of the nodes, the first escape path corresponding to the node to a terminal for escape corresponding to the node;

[0080] A seventh acquisition module is used to obtain an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix if it is detected that the human flow density corresponding to any of the nodes changes, or the human flow density of the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation of any of the nodes changes;

[0081] an eighth acquisition module, configured to input the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix, and the updated evacuation time matrix into the evacuation path prediction model, and acquire a second escape path through the evacuation path prediction model;

[0082] The second sending module is used to send, for each of the nodes, the second escape path corresponding to the node to the terminal for escape corresponding to the node.

[0083] The third aspect of the present application provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the fire emergency evacuation indication method applied to large indoor scenes according to the above-mentioned first aspect or any implementation method of the first aspect.

[0084] A fourth aspect of the present application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0085] The memory is used to store computer programs;

[0086] The processor is used to execute the computer program so that the electronic device can implement the fire emergency evacuation indication method applied to a large indoor scene according to the first aspect or any implementation manner of the first aspect.

[0087] In a fifth aspect, the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement the fire emergency evacuation indication method applied to large indoor scenes according to the above-mentioned first aspect or any implementation of the first aspect.

[0088] By means of the above technical scheme, the present application provides a fire emergency evacuation indication method applied to large indoor scenes. The spatial characteristics of the actual evacuation scene can be restored through the personnel evacuation network flow model; the evacuation path of the actual evacuation scene can be restored through the distance matrix and the personnel evacuation network flow model; if a fire is detected, the density matrix can be used to determine which nodes have a larger flow density of people and which passages between two nodes have a larger flow density of people, so as to determine which nodes are dangerous nodes, which nodes are disaster nodes, and which passages are congested sections; the speed matrix is ​​obtained based on the density matrix, and the speed matrix can be used to know which passage the personnel move faster and which passage the personnel move slower during the evacuation process. The evacuation time matrix is ​​obtained based on the distance matrix and the speed matrix; the evacuation time matrix can be used to know which escape exit the personnel takes the shortest time to escape from. The personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix, and the evacuation time matrix are input into the pre-built evacuation path prediction model, and the first escape paths corresponding to all nodes are obtained through the evacuation path prediction model. People can escape based on the first escape path; as time goes by, the fire details of the nodes that people have not passed through in the first escape path may change, causing the first escape path to no longer be the most suitable escape path. At this time, it is necessary to obtain an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix; the personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix, and the updated evacuation time matrix are input into the evacuation path prediction model, and the second escape path is obtained through the evacuation path prediction model; in this way, people can escape through the second escape path. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0090] Figure 1 A schematic diagram of a system architecture provided for this application;

[0091] Figure 2 A schematic diagram of an optional hardware structure of a terminal 100 provided in this application;

[0092] Figure 3 A schematic diagram of the structure of an electronic device 200 provided in this application;

[0093] Figure 4 A flow chart of a fire emergency evacuation indication method applied to a large indoor scene provided in an embodiment of the present application;

[0094] Figure 5 A schematic diagram of the construction process of the personnel evacuation network flow model provided in the embodiment of the present application;

[0095] Figure 6 A schematic diagram of a personnel evacuation network flow model for a bank branch provided in an embodiment of the present application;

[0096] Figure 7 A schematic diagram of part of the distance matrix L of the second layer of bank branches provided in an embodiment of the present application;

[0097] Figure 8 A schematic diagram of a personnel evacuation network flow model updated by step S406 provided in an embodiment of the present application;

[0098] Fig. 9 A schematic diagram comparing a static evacuation path and a dynamic evacuation path provided in an embodiment of the present application;

[0099] Fig.10 A schematic diagram of the structure of a deep Q network model provided in an embodiment of the present application;

[0100] Fig.11 A structural diagram of the first CBS module provided in an embodiment of the present application;

[0101] Fig.12 A schematic diagram of the interaction process between an intelligent agent and a bank branch environment provided in an embodiment of the present application;

[0102] Fig.13 A schematic diagram of the structure of a fire emergency evacuation indication device applied to a large indoor scene provided in an embodiment of the present application;

[0103] Fig.14 A schematic diagram of a fire emergency evacuation indication system applied to a large indoor scene provided in an embodiment of the present application. DETAILED DESCRIPTION

[0104] The following describes the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. The terms used in the implementation method section of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0105] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0106] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0107] See also Figure 1 , Figure 1 A schematic diagram of a system architecture is shown. The system may include multiple terminals 100, electronic devices 200 and multiple sensors 300.

[0108] Among them, multiple sensors 300 can be deployed at corresponding locations in large indoor scenes, for example, in the business hall of a bank branch, the customer waiting area, the teller reception area and other areas where employees and customers may be present.

[0109] Exemplarily, the multiple sensors 300 include but are not limited to: laser radar, camera, smoke sensor, temperature sensor. It is understandable that if the smoke sensor or temperature sensor is deployed in the customer waiting area, the smoke sensor or temperature sensor can detect whether a fire occurs in the customer waiting area.

[0110] For example, if the camera is deployed in a customer waiting area, then when a fire occurs, the density of human traffic in the customer waiting area can be estimated through images captured by the camera.

[0111] For example, when a fire occurs, the smoke density may be high, and the crowd density cannot be obtained through the images taken by the camera. At this time, the crowd density in the customer waiting area can be obtained through the point cloud data collected by the lidar.

[0112] For example, the deployment of multiple sensors at various locations may depend on specific circumstances and is not limited here.

[0113] It is understandable that the information collected by multiple sensors 300 can be uploaded to the electronic device 200, so that the electronic device 200 can execute the fire emergency evacuation indication method applied to large indoor scenes provided in the embodiment of the present application to obtain the first escape path or the second escape path.

[0114] It is understandable that multiple terminals 100 can be deployed in areas that can carry people, and on pathways for people to walk, and the electronic device 200 sends the first escape path and the second escape path to the terminal 100, thereby prompting people how to escape.

[0115] Exemplarily, the terminal 100 may be a mobile phone, a tablet computer, a wearable device, a vehicle-mounted device, an augmented reality (AR) / virtual reality (VR) device, a laptop computer, an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (PDA), etc., and the embodiments of the present application do not impose any restrictions on this.

[0116] Next describe Figure 1 The product form of the mid-terminal 100;

[0117] Figure 2 An optional hardware structure diagram of the terminal 100 is shown.

[0118] refer to Figure 2 As shown, the terminal 100 may include a radio frequency unit 110, a memory 120, an input unit 130, a display unit 140, a camera 150 (optional), an audio circuit 160 (optional), a speaker 161 (optional), a microphone 162 (optional), an earphone jack 163 (optional), a processor 170, an external interface 180, a power supply 190 and other components. Those skilled in the art will appreciate that Figure 2 This is merely an example of a terminal and does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown in the figure, or a combination of certain components, or different components.

[0119] The input unit 130 can be used to receive input digital or character information, and generate key signal input related to the user settings and function control of the terminal. Specifically, the input unit 130 may include a touch screen 131 (optional) and / or other input devices 132. The touch screen 131 can collect the user's touch operations on or near it (such as the user's operation on or near the touch screen using any suitable object such as fingers, joints, stylus, etc.), and drive the corresponding connection device according to a pre-set program. The touch screen can detect the user's touch action on the touch screen, convert the touch action into a touch signal and send it to the processor 170, and can receive and execute the command sent by the processor 170; the touch signal at least includes the touch point coordinate information. The touch screen 131 can provide an input interface and an output interface between the terminal 100 and the user. In addition, the touch screen can be implemented using multiple types such as resistive, capacitive, infrared and surface acoustic wave. In addition to the touch screen 131, the input unit 130 can also include other input devices. Specifically, other input devices 132 may include, but are not limited to, one or more of a physical keyboard, function keys (such as a volume control key, a switch key, etc.), a trackball, a mouse, a joystick, and the like.

[0120] Among them, the input device 132 can receive input data and the like.

[0121] The display unit 140 may be used to display information input by the user or provided to the user, various menus of the terminal 100, interactive interfaces, file display and / or playback of any multimedia file. In the embodiment of the present application, the display unit 140 may be used to display interfaces, processing results, etc.

[0122] The memory 120 can be used to store instructions and data. The memory 120 can mainly include an instruction storage area and a data storage area. The data storage area can store various data, such as multimedia files, texts, etc.; the instruction storage area can store software units such as operating systems, applications, instructions required for at least one function, or their subsets and extensions. It can also include a non-volatile random access memory; provide the processor 170 with hardware, software and data resources including management of computing and processing equipment, and support control software and applications. It is also used for the storage of multimedia files, and the storage of running programs and applications.

[0123] The processor 170 is the control center of the terminal 100. It uses various interfaces and lines to connect various parts of the entire terminal 100. By running or executing instructions stored in the memory 120 and calling data stored in the memory 120, it executes various functions of the terminal 100 and processes data, thereby controlling the terminal device as a whole. Optionally, the processor 170 may include one or more processing units; preferably, the processor 170 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application program, and the modem processor mainly processes wireless communication. It is understandable that the above-mentioned modem processor may not be integrated into the processor 170. In some embodiments, the processor and the memory may be implemented on a single chip, and in some embodiments, they may also be implemented separately on separate chips. The processor 170 may also be used to generate corresponding operation control signals, send them to corresponding components of the computing and processing device, read and process data in the software, especially read and process data and programs in the memory 120, so that each functional module therein performs corresponding functions, thereby controlling the corresponding components to act according to the requirements of the instructions.

[0124] The radio frequency unit 110 (optional) can be used for receiving and sending information or receiving and sending signals during a call, for example, after receiving the downlink information of the base station, it is sent to the processor 170 for processing; in addition, the designed uplink data is sent to the base station. Generally, the RF circuit includes but is not limited to an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier (Low Noise Amplifier, LNA), a duplexer, etc. In addition, the radio frequency unit 110 can also communicate with network devices and other devices through wireless communication. The wireless communication can use any communication standard or protocol, including but not limited to Global System of Mobile communication (Global System of Mobile communication, GSM), General Packet Radio Service (General Packet Radio Service, GPRS), Code Division Multiple Access (Code Division Multiple Access, CDMA), Wideband Code Division Multiple Access (Wideband Code Division Multiple Access, WCDMA), Long Term Evolution (Long Term Evolution, LTE), email, Short Messaging Service (SMS), etc.

[0125] In this embodiment of the present application, the radio frequency unit 110 can send data to the electronic device 200 and receive a processing result sent by the electronic device 200.

[0126] It should be understood that the radio frequency unit 110 is optional and can be replaced by other communication interfaces, such as a network port.

[0127] The terminal 100 also includes a power supply 190 (such as a battery) for supplying power to various components. Preferably, the power supply can be logically connected to the processor 170 through a power management system, so that the power management system can manage functions such as charging, discharging, and power consumption.

[0128] The terminal 100 further includes an external interface 180 , which may be a standard Micro USB interface or a multi-pin connector, and may be used to connect the terminal 100 to communicate with other devices, or to connect a charger to charge the terminal 100 .

[0129] Although not shown, the terminal 100 may also include a flashlight, a wireless fidelity (WiFi) module, a Bluetooth module, sensors with different functions, etc., which are not described in detail here. Some or all of the methods described below may be applied in the following embodiments. Figure 2 In the terminal 100 shown.

[0130] Next describe Figure 1 The product form of the electronic device 200;

[0131] Exemplarily, the electronic device 200 may be an edge computing box or a server deployed in a large indoor scene.

[0132] Figure 3 A schematic diagram of the structure of an electronic device 200 is provided. Figure 3 As shown, the electronic device 200 includes a bus 201, a processor 202, a communication interface 203 and a memory 204. The processor 202, the memory 204 and the communication interface 203 communicate with each other via the bus 201.

[0133] The bus 201 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0134] The processor 202 may be any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0135] The memory 204 may include a volatile memory, such as a random access memory (RAM). The memory 204 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0136] Among them, the memory 204 can be used to store software codes related to the fire emergency evacuation indication method applied to large indoor scenes, the processor 202 can execute the steps of the fire emergency evacuation indication method applied to large indoor scenes of the chip, and can also dispatch other units to realize corresponding functions.

[0137] It should be understood that the above-mentioned terminal 100 and electronic device 200 can be centralized or distributed devices, and the processors in the above-mentioned terminal 100 and electronic device 200 (such as processor 170 and processor 202) can be hardware circuits (such as application specific integrated circuit (ASIC), field-programmable gate array (FPGA), general-purpose processor, digital signal processor (DSP), microprocessor or microcontroller, etc.), or a combination of these hardware circuits. For example, the processor can be a hardware system with an instruction execution function, such as a CPU, DSP, etc., or a hardware system without an instruction execution function, such as an ASIC, FPGA, etc., or a combination of the above-mentioned hardware systems without an instruction execution function and hardware systems with an instruction execution function.

[0138] Every year, fires cause more than tens of thousands of casualties. Large indoor scenes are numerous and widely distributed, with complex environments and large traffic volumes. After a disaster occurs, if there is an efficient evacuation strategy that can guide people to evacuate to a safe area, the losses and casualties caused by the accident can be minimized. Planning a reasonable escape route is a crucial step in these measures, while traditional path planning algorithms only use path length as the only criterion, which may lead people into the area where the fire occurs, potentially causing secondary injuries.

[0139] There are a large number of people in large indoor scenes, which can easily cause excessive density of pedestrian traffic on the passage, resulting in congestion or stagnation, and in serious cases, stampedes may occur. In the process of evacuating people, environmental factors are complex and change frequently, such as a fire suddenly breaking out in an area where there was no fire before, or a sudden increase in the density of pedestrian traffic in an area with a low density of pedestrian traffic. In related technologies, emergency evacuation route planning is a static evacuation route. If the static evacuation route has been planned (assuming it is area A → area B → area C), and environmental factors change, then a certain area in the static evacuation route may not be suitable for people to escape. For example, if a fire breaks out in area B, if the static evacuation route is not updated in time, people may be injured again after escaping from area A to area B.

[0140] In order to solve the above problems, the embodiment of the present application provides a fire emergency evacuation indication method applied to a large indoor scene. The fire emergency evacuation indication method applied to a large indoor scene of the embodiment of the present application is described in detail below with reference to the accompanying drawings.

[0141] Reference Figure 4 , Figure 4 A flow chart of a fire emergency evacuation indication method for a large indoor scene provided in an embodiment of the present application is shown as follows: Figure 4 As shown, a fire emergency evacuation indication method applied to a large indoor scene provided in an embodiment of the present application may include steps S401 to S413, and these steps are described in detail below.

[0142] Step S401: Obtaining a personnel evacuation network flow model.

[0143] The personnel evacuation network flow model includes multiple nodes and edges between two nodes; the edges between two nodes represent the passages between two nodes for people to walk; the multiple nodes include escape exit nodes, ordinary nodes, and stair nodes.

[0144] The following example illustrates the process of constructing a personnel evacuation network flow model.

[0145] like Figure 5As shown, it is a schematic diagram of the construction process of the personnel evacuation network flow model provided in an embodiment of the present application.

[0146] Assumptions Figure 5 The left side is a top view of various objects on a floor. For example, various objects (i.e., areas filled with gray) include but are not limited to: counters, business physical stores, etc. Figure 5 The blank areas shown on the left are all passages that people can pass through; Figure 5 The gray filled area shown on the left is not accessible to people.

[0147] Passages accessible to people can be abstracted into edges. Forks with multiple directions can be abstracted into nodes. Exemplarily, the following formula can be used to characterize the personnel evacuation network flow model.

[0148] ; ; ; Where V represents a node set. For example, set V includes n nodes, namely v1, v2, ..., v n .like Figure 5 As shown on the right, n=12, that is, n is the number of nodes; E represents the set of edges, such as the connection node v a and node v b The line segment is the edge. Figure 5 As shown on the right, set E includes 17 edges.

[0149] It can be understood that the personnel evacuation network flow model restores the spatial characteristics of the actual evacuation scene with nodes and edges. It is suitable for scenes with large evacuation areas and a large number of people to be evacuated.

[0150] In order for those skilled in the art to better understand the significance of the personnel evacuation network flow model, an example is given below to illustrate.

[0151] like Figure 6 , which is a schematic diagram of a personnel evacuation network flow model for a bank branch provided in an embodiment of the present application.

[0152] Exemplarily, the bank branch includes: an office area, a basement, and a vault. The bank branch includes two floors.

[0153] Each floor of a bank branch is 42m long and 14m wide, with 4 fire stairs on each floor. This application considers that all personnel who walk to the three safe escape exits on the first floor of the bank branch are considered safe escapes. The process of constructing the personnel evacuation network flow model of the bank branch is as follows:

[0154] Step 1: Set the business hall, customer waiting area, teller reception area, etc. in the bank's business outlets as nodes, except for restricted areas such as elevator shafts and ventilation system control rooms.

[0155] Step 2: Set the stairs connecting the upper and lower floors of the bank branch as stair nodes.

[0156] Step 3: Set the intersections (i.e. bifurcations) of the paths on each level of the bank's business outlets as nodes;

[0157] Step 4: Set the roads, corridors, aisles and other passages accessible to bank branches as edges.

[0158] It is understandable that the above nodes and edges refer to areas with a certain space in the bank's business outlets, rather than just a point or an edge in a geometric sense. In order to facilitate calculation and research, the geometric center position of the intersection of the passages (i.e. the fork in the road) is defined as a node; the geometric center position of the business hall, the customer waiting area and the counter is defined as a node. According to the above four steps, we can get Figure 6 The personnel evacuation network flow model is shown.

[0159] Figure 6 In the figure, a white filled circle (i.e., a white circular node) is used to represent an escape exit node, such as: escape exit node 5, escape exit node 21, and escape exit node 36; Figure 6 In the figure, the stair nodes are represented by gray-filled rectangles (gray square nodes). Figure 6 In the figure, ordinary nodes are represented by gray-filled circles (gray circular nodes).

[0160] Step S402: Obtain a distance matrix corresponding to the personnel evacuation network flow model.

[0161] The distance matrix includes the distances between all nodes with edges in the personnel evacuation network flow model.

[0162] Still Figure 6 Taking the personnel evacuation network flow model shown as an example, Figure 6 There are 68 nodes in total. In order to more accurately describe the distance relationship between these nodes, the distance matrix L is introduced. The distance matrix is ​​a 68×68 matrix. Since the node numbering starts from 0, the element L(i, j) in the i-th row and j-th column of the matrix represents the actual distance of the path between node i-1 and node j-1. If node i and node j are not connected, the corresponding element value is 0.

[0163] like Figure 7 As shown, it is a schematic diagram of part of the distance matrix L of the second layer of the bank branch provided in the embodiment of the present application. During the actual evacuation process of the bank branch, each passage allows escapees to pass in both directions, so the distance matrix L is a symmetric matrix.

[0164] Step S403: If a fire is detected, a density matrix corresponding to the personnel evacuation network flow model is obtained.

[0165] The density matrix includes the human flow densities corresponding to all nodes in the personnel evacuation network flow model and the human flow densities of paths between two nodes with edges.

[0166] The process of obtaining the density matrix is ​​described below.

[0167] Density matrix D t1 It represents the traffic density of each node and each edge at time t1; if the density matrix is Figure 6 Taking the personnel evacuation network flow model shown as an example, the density matrix D t1 It is also a 68×68 matrix, and the element value unit is person / m 2 The element D in the i-th row and j-th column of the matrix t1 (i, j) represents the traffic density of the passage between node i-1 and node j-1 at time t1. When i=j, it represents the traffic density at node i-1. If node i is not directly connected to node j, the corresponding element value of the density matrix is ​​set to 99. If D t1 If (i, j) is greater than 4, it means that the corresponding passage is impassable. Since the density of pedestrian traffic in each passage of a bank branch is different at different time points, the elements in the density matrix will change over time. The calculation method of the pedestrian traffic density of the element in the i+1th row and j+1th column of the density matrix at time t1 is as follows.

[0168] , , .

[0169] in, represents the path r from node i to node j at time t1 ij The density of traffic on Represents the traffic density at node i at time t1. and represents the path r at time t1 ij and the number of people at node i. These two values ​​can be calculated by the program based on the current location information of each person. and Respectively represent the path r ij and the relative area of ​​node i.

[0170] Step S404: Determine dangerous nodes and disaster nodes from the multiple nodes.

[0171] The dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs.

[0172] For example, whether a fire occurs may be detected by a smoke sensor or a temperature sensor.

[0173] Exemplarily, the first preset density value may be determined based on actual conditions, and is not limited in comparison with the embodiments of the present application.

[0174] Step S405: If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, it is determined that the edge between the two nodes is a congested section.

[0175] Exemplarily, the second preset density value may be determined based on actual conditions, and is not limited in comparison with the embodiments of the present application.

[0176] Step S406: updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model.

[0177] It is understandable that dangerous nodes and disaster nodes can be determined from multiple nodes; dangerous nodes are nodes where the personnel flow density is greater than or equal to the first preset density value, and disaster nodes are nodes where fire occurs. If the flow of people on the path represented by the edge between two nodes is greater than or equal to the second preset density value, the edge between two nodes is determined to be a congested section, otherwise it is a smooth section.

[0178] like Figure 8 , which is a schematic diagram of the personnel evacuation network flow model updated through step S406 provided in an embodiment of the present application.

[0179] Still Figure 6 For example, Figure 8 In the figure, a white filled circle (i.e., a white circular node) is used to represent an escape exit node, such as: escape exit node 5, escape exit node 21, and escape exit node 36; Figure 8 In the figure, disaster nodes are represented by circles filled with grids (grid circular nodes), such as disaster node 7 and disaster node 46. Figure 8 In the figure, a gray-filled rectangle (gray square node) is used to represent a stair node. It can be understood that if the stair node is a fire node, the color of the stair node may change. Figure 6 A gray-filled circle (gray circular node) is used to represent an ordinary node, where no fire occurs and the human flow density is less than the first preset density value.

[0180] Figure 8 The dotted lines are used to represent the congested sections and the sections where fires occur.

[0181] Gray circular nodes, i.e. ordinary nodes, represent regular nodes where people can pass smoothly. Grid circular nodes, i.e. disaster nodes, represent nodes where disasters occur. They can be used as the starting point for evacuation, but cannot be used as the target node for the next step. Disaster nodes do not need to be considered in the static evacuation path planning stage. White circular nodes, i.e. escape exit nodes, represent the escape exits that evacuees need to reach. Gray square nodes, i.e. staircase nodes, represent stairs. There are four stairs on each floor of the building. Edges are divided into two types based on whether they are congested or not. Black solid line edges represent regular sections where people can pass smoothly; dotted line edges represent congested sections, or passages that are prone to congestion, such as passages with an average width of 0.8m to 1.1m. Since congestion is likely to occur when people pass through these passages, if there are better passages to choose from, people should try to avoid passing through these passages.

[0182] Understandably, Figure 8 The dangerous nodes, fire nodes, and congested sections shown in can be used to obtain the density matrix and speed matrix later. Figure 8 The personnel evacuation network flow model shown is the personnel evacuation network flow model after executing step S406.

[0183] Step S407: Based on the density matrix, obtain the speed matrix corresponding to the personnel evacuation network flow model.

[0184] The speed matrix includes the moving speeds of personnel in all paths between any two nodes with edges in the personnel evacuation network flow model.

[0185] The process of obtaining the velocity matrix is ​​described below.

[0186] Speed ​​Matrix It represents the moving speed of people in each passage at time t1. Figure 6 Taking the personnel evacuation network flow model shown in the figure as an example, It is also a 68×68 matrix. Its element value at time t1 is It can be calculated according to the following formula, where: Indicates the moving speed when going down stairs. It represents the moving speed on the horizontal path, and the elements in the set S represent the sequence numbers of the nodes that people need to pass when going down the stairs.

[0187] ;by Figure 6 For example, the set .

[0188] ;

[0189] .

[0190] Step S408: Based on the distance matrix and the speed matrix, an evacuation time consumption matrix corresponding to the personnel evacuation network flow model is obtained.

[0191] The process of obtaining the evacuation time-consuming matrix is ​​described below.

[0192] Evacuation time matrix It represents the time it takes for a person to move from node i to node j at time t1. If the evacuation time matrix So Figure 6 Taking the personnel evacuation network flow model shown as an example, the evacuation time matrix It is also a 68×68 matrix, which is calculated by dividing the corresponding elements of the distance matrix and the velocity matrix, as shown in the formula If the evacuation time matrix If the value of the element is 0, it means that the two corresponding nodes are not directly connected, and the time required for transfer between non-nodes is 0.

[0193] Step S409: input the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtain the first escape paths corresponding to all nodes respectively through the evacuation path prediction model.

[0194] The first escape route is a static evacuation route.

[0195] Static evacuation path means that once an emergency such as a fire occurs, it is in the static evacuation stage before obtaining detailed information about the changed disaster situation. In this stage, it is necessary to plan a static evacuation path for all personnel with their current location as the starting point in the shortest possible time to guide their escape.

[0196] Exemplarily, the evacuation path prediction model may be a machine learning model.

[0197] Exemplarily, the training process of the evacuation path prediction model is as follows: the personnel evacuation network flow model, distance matrix, sample density matrix, sample speed matrix, and sample evacuation time matrix are used as inputs of the machine learning model, and the labeled escape path is used as the training target. The machine learning model is trained to obtain an evacuation path prediction model.

[0198] It can be understood that the process of obtaining the sample density matrix, the sample velocity matrix, and the sample evacuation time matrix is ​​the same as the process of obtaining the density matrix, the velocity matrix, and the evacuation time matrix, and is not limited here.

[0199] Step S410: For each of the nodes, the first escape path corresponding to the node is sent to a terminal for escape corresponding to the node.

[0200] Step S411: If it is detected that the pedestrian flow density corresponding to any of the nodes changes, or the pedestrian flow density of the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation at any of the nodes changes, an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix and an updated evacuation time matrix are obtained.

[0201] It is understandable that if any node is changed from an ordinary node to a disaster node, it is determined that the fire situation has changed. If any node is changed from an ordinary node to a dangerous node or from a dangerous node to an ordinary node, it is determined that the human flow density has changed. If the human flow density of the passage between any two nodes changes from greater than or equal to the second preset density value to less than the second preset density value, or from less than the second preset density value to greater than or equal to the second preset density value, it is determined that the human flow density has changed.

[0202] An updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix can be obtained in real time.

[0203] In the embodiment of the present application, the parameters that affect the element values ​​in the density matrix, speed matrix, or evacuation time matrix are referred to as detailed disaster information (i.e., the density of human flow has changed, or the density of human flow in the passage between two nodes has changed, or the moving speed of people in the passage between two nodes has changed, or the fire situation at the node has changed).

[0204] The process of determining the detailed disaster information in step S411 is described below.

[0205] For each of the nodes, a smoke sensor deployed at the node is used to detect whether the node has changed from a disaster-prone location to a non-disaster-prone location, or from a non-disaster-prone location to a disaster-prone location. If so, it is determined that the fire situation at the node has changed.

[0206] For each of the nodes, the density of human traffic at the node is detected by a sensor deployed at the node. If the human traffic density changes from greater than or equal to the first preset density value to lower than the first preset density value, or changes from lower than the first preset density value to greater than or equal to the first preset density value, it is determined that the human traffic density of the node has changed.

[0207] For each pairwise edge of nodes, the density of pedestrian flow in the passage is detected by sensors deployed in the passage between the two nodes. If the pedestrian flow density changes from greater than or equal to a second preset density value to less than the second preset density value, or changes from less than the second preset density value to greater than or equal to the second preset density value, it is determined that the pedestrian flow density of the passage has changed.

[0208] For each pairwise node edge, smoke concentration in the passage is detected by a smoke sensor deployed in the pairwise node passage; if the smoke concentration level corresponding to the smoke concentration changes, it is determined that the moving speed of people in the passage between the pairwise nodes changes.

[0209] It is understandable that some factors may affect the moving speed of a person, for example, the greater the smoke concentration, the slower the moving speed of a person. The following takes the smoke concentration as CO concentration as an example for explanation. For each edge of each pair of nodes, the method for obtaining the updated moving speed corresponding to the path of the pair of nodes includes the following steps A1 to A2.

[0210] Step A1: By formula The coefficient α is calculated; ρ co is the CO concentration of the pathway between the two nodes.

[0211] Step A2: Determine the updated moving speed as the product of the moving speed before updating and the coefficient α.

[0212] The above formula abstractly expresses the effect of fire smoke on the evacuation speed of personnel.

[0213] Step S412: input the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix and the updated evacuation time matrix into the evacuation path prediction model, and obtain the second escape path through the evacuation path prediction model.

[0214] The second escape path is a dynamic evacuation path. As time goes by, the updated density matrix, the updated speed matrix, and the updated evacuation time matrix are constantly changing, and the second escape path is also constantly changing.

[0215] In the embodiment of the present application, the parameters that affect the element values ​​in the density matrix, speed matrix, or evacuation time matrix are referred to as detailed disaster information (i.e., the density of human flow has changed, or the density of human flow in the passage between two nodes has changed, or the moving speed of people in the passage between two nodes has changed, or the fire situation at the node has changed).

[0216] In the actual emergency evacuation process, the impact of dynamic environmental factors such as fire spread and road congestion on escaping personnel must be considered. During the evacuation process, once environmental factors such as the disaster situation change, if the static evacuation path is still used, there may be huge safety hazards. Therefore, after detecting the detailed information of the changed disaster situation, the evacuation path must be dynamically adjusted according to the real-time environmental information, that is, the dynamic evacuation path, i.e. the second escape path, is obtained.

[0217] In order for those skilled in the art to better understand the static evacuation path and the dynamic evacuation path mentioned in the embodiments of the present application, Fig. 9 Provide explanation.

[0218] like Fig. 9 The figure is a comparative schematic diagram of the static evacuation path and the dynamic evacuation path provided in the embodiment of the present application.

[0219] like Fig. 9 As shown in the figure, when a fire occurs, the personnel are located at the initial node and can obtain a static evacuation path based on the current fire situation; during the evacuation based on the static evacuation path, if the detailed information of the changed disaster situation is obtained, a dynamic evacuation path can be obtained. Personnel can evacuate based on the dynamic evacuation path.

[0220] It is understandable that as time goes by, the detailed information of the disaster situation may be constantly changing, and multiple dynamic evacuation paths, that is, multiple second escape paths, may be obtained.

[0221] Step S413: For each of the nodes, the second escape path corresponding to the node is sent to the terminal for escape corresponding to the node.

[0222] The embodiment of the present application provides a fire emergency evacuation indication method applied to large indoor scenes. The spatial characteristics of the actual evacuation scene can be restored through the personnel evacuation network flow model; the evacuation path of the actual evacuation scene can be restored through the distance matrix and the personnel evacuation network flow model; if a fire is detected, the density matrix can be used to determine which nodes have a larger flow density of people and which passages between two nodes have a larger flow density of people, so as to determine which nodes are dangerous nodes, which nodes are disaster nodes, and which passages are congested sections; the speed matrix is ​​obtained based on the density matrix, and the speed matrix can be used to know which passage the personnel move faster and which passage the personnel move slower during the evacuation process. The evacuation time matrix is ​​obtained based on the distance matrix and the speed matrix; the evacuation time matrix can be used to know which escape exit the personnel takes the shortest time to escape from. The personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix, and the evacuation time matrix are input into the pre-built evacuation path prediction model, and the first escape path corresponding to all nodes is obtained through the evacuation path prediction model. People can escape based on the first escape path; as time goes by, the fire details of the node may change, resulting in the first escape path no longer being the most suitable escape path. At this time, it is necessary to obtain an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix; the personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix, and the updated evacuation time matrix are input into the evacuation path prediction model, and the second escape path is obtained through the evacuation path prediction model; in this way, people can escape through the second escape path.

[0223] It is understandable that if the disaster nodes are not suitable as nodes of the escape route, these disaster nodes will be avoided during route planning to ensure the safety of the evacuation process. As time goes by, if the pedestrian density of the passages between two nodes is higher than the second preset density value, a dynamic pedestrian density control strategy will be implemented according to the congestion of the current passages. For example, during route planning, these passages will be avoided to reduce the congestion of the congested sections by restricting access to the congested sections.

[0224] It is understandable that as time goes by, the node where the personnel are currently located will change. For example, some personnel are located at a disaster node, some are located at a dangerous node, and some are located at an ordinary node. No matter which node the personnel are located at, they will use their current node as the starting point to update the planned escape path in real time and send it to the terminal used for escape, such as on-site auxiliary equipment (such as sports bracelets, voice broadcast systems), so that personnel can escape based on the escape path.

[0225] In a possible implementation, the evacuation path prediction model is a dual-depth Q network model, and the method for training the dual-depth Q network includes the following steps B01 to B23.

[0226] Step B01: Obtain a sample density matrix corresponding to the personnel evacuation network flow model; the sample density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of paths between two nodes with edges.

[0227] It is understandable that different personnel evacuation network flow models correspond to different evacuation path prediction models.

[0228] It can be understood that a fire scene can be simulated and the agents can be used as people who need to be evacuated to obtain a sample density matrix.

[0229] The agent plays the role of evacuees and needs to constantly choose paths to find the optimal escape path. It is the entity that performs actions.

[0230] Step B02: Based on the sample density matrix, a sample speed matrix corresponding to the personnel evacuation network flow model is obtained; the sample speed matrix includes the movement speeds of personnel in all paths between any two nodes with edges in the personnel evacuation network flow model.

[0231] The process of obtaining the sample velocity matrix is ​​the same as the process of obtaining the velocity matrix, which will not be repeated here.

[0232] Step B03: Based on the sample distance matrix and the sample speed matrix, obtain a sample evacuation time consumption matrix corresponding to the personnel evacuation network flow model.

[0233] The process of obtaining the sample sparse time matrix is ​​the same as the process of obtaining the sparse time matrix, which will not be repeated here.

[0234] Step B04: Determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node with a personnel flow density greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs.

[0235] Step B05: If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, it is determined that the edge between the two nodes is a congested section.

[0236] Step B06: updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model.

[0237] Step B07: Acquire state information of the agent, wherein the state information includes the time taken to move from the first node to each of the second nodes.

[0238] The first node is the node where the agent was located before moving.

[0239] The sample evacuation time matrix combines the information of the sample density matrix, the distance matrix, and the sample speed matrix, and can comprehensively reflect the real-time location information of the agent in the evacuation environment. Since the agent can only move to the next node directly connected to it at a time, that is, the agent can move from the first node to any second node. Then the time it takes to move from the first node to the second node is the element value in the sample evacuation time matrix. Assume that the number of second nodes directly connected to the first node is e, and the serial number of the first node is ti, and the serial numbers of the e second nodes are: j1, j2,…, je; Assume that the time it takes for the first node to move to the second node j1 is represented in the sample evacuation time matrix by T ti j1 The time taken to move from the first node to the second node j2 is represented by T in the sample evacuation time matrix. ti j2 Characterize, ..., the time taken to move from the first node to the second node je in the sample evacuation time matrix is ​​T ti je Representation, then the state information of the agent can be obtained as: .

[0240] Step B08: Input the personnel evacuation network flow model, the distance matrix, the sample density matrix, the sample speed matrix, the sample evacuation time matrix and the state information into the dual deep Q network model, and output the predicted Q values ​​corresponding to multiple actions respectively through the dual deep Q model.

[0241] The dual deep Q network model includes a deep Q network model and a target Q network model; the action is the action of the agent moving from the first node to the second node.

[0242] In the DDQN (Double Deep Q-Network) model, the action selected by the agent at each time usually only depends on the action A with the largest predicted Q value output by the DDQN model. t1 ,Right now: The time when the agent is at the first node is time t1. i It refers to the predicted Q value corresponding to the action of moving from the first node to the second node ji.

[0243] However, if the agent selects the action corresponding to the largest predicted Q value every time, although the agent can successfully complete the escape task, it will also lose the opportunity to explore a better escape path, which means that the reinforcement learning falls into a local minimum. To avoid this situation, this application adopts the ε-greedy strategy to determine the next action taken by the agent, that is, step B08 is equal to step B10.

[0244] Step B09: Obtain a random number.

[0245] Step B10: If the random number is greater than a preset threshold, a target action is randomly selected from the multiple actions, and the second node in the target action is the target second node.

[0246] Step B11: If the random number is less than or equal to the preset threshold, a target action corresponding to the maximum expected Q value is selected from the multiple actions; the second node in the target action is the target second node.

[0247] For example, the application sets the exploration rate ε to 5%, that is, the preset threshold is 0.95. When executing the ε-greedy strategy, a random number between 0 and 1 is randomly generated. If the random number is less than or equal to 0.95, the action corresponding to the largest predicted Q2 value is selected; otherwise, a random selection is made from the optional actions.

[0248] Step B12: Based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken by the agent to reach the target second node from the initial node exceeds a preset prescribed time, determine the first reward for the agent to move to the target second node.

[0249] Step B13: Inputting the first reward and the predicted Q value corresponding to the target action into the dual deep Q network model to update the deep Q network model in the dual deep Q network model.

[0250] In order to enable the agent to move towards the escape exit node when choosing an escape path, this application proposes a reward function design based on a distance estimator. To implement the distance estimator, a distance estimation list needs to be initialized. Each element in the list represents the shortest estimated distance from the node to any escape exit node. When the agent moves from the first node to the next target second node, if the distance of the target second node to the escape exit node is closer than before the movement and the target second node is not a disaster node, then the action is regarded as a positive incentive, and the agent will get The first reward R shown t1 ; If the distance between the target second node and the escape exit node is farther than before the move and the target second node is not a disaster node, then the action is regarded as a negative incentive and the agent will get The first reward R shown t1 .

[0251] Among them, k1 is the preset reward amplification factor, which is used to control the optimization efficiency.

[0252] In summary, if the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset specified time, the formula , calculate R t1 ; Determine R t1 This is the first reward.

[0253] If the target second node is not an escape exit node and the target time is greater than the preset time, or the target second node is a disaster node, it means that the path is dangerous, and the second preset reward value is determined to be the first reward. Exemplarily, the second preset reward value is -100.

[0254] Step B14: If the target second node is not the disaster node and is not the escape exit node, and the target time does not exceed the preset specified time, return to step B07.

[0255] Step B15: If the target second node is the disaster node or the target time exceeds the preset prescribed time, determine that the agent returns to the initial node and return to step B07.

[0256] Step B16: If the target second node is the escape exit node, obtain an escape planning path, where the escape planning path is the path for the agent to reach the escape exit node from the initial node.

[0257] Step B17: Determine a second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path.

[0258] In order to enable the intelligent agent to avoid disaster nodes and congested sections and choose relatively smooth and safe sections when choosing an escape route, this application proposes a reward function design based on a path scorer.

[0259] The path scorer is designed to evaluate the quality of the generated escape plan path and provide the agent with corresponding rewards at the end of each round based on the quality of the escape plan path. In terms of scoring criteria, the path scorer mainly considers the length of the escape plan path, the safety of the escape plan path, and the smoothness of the escape plan path. The total path score is calculated as follows:

[0260] .

[0261] in, , , They represent the length score of the escape planning path, the safety score of the escape planning path, and the smoothness score of the escape planning path. During the evacuation process, the shorter the distance of the escape planning path, the shorter the time will be. However, if there are disaster nodes in the escape planning path, it may cause secondary accidents during people's escape. Therefore, the length score and safety score of the escape planning path need to be considered. In addition, the presence of easily blocked sections in the escape planning path may cause congestion, thereby reducing the escape efficiency. Therefore, the smoothness score of the escape planning path also needs to be considered. These three factors are equally important in judging the quality of the escape planning path, so the score weights of these three factors are 1 / 3 each, so as to ensure that the sum of the maximum theoretical scores is 1.

[0262] in, , ,in is the three-dimensional coordinate of the initial node; The three-dimensional coordinates of the escape exit node closest to the initial node; is the three-dimensional coordinate of the jth node in the escape planning path; N is the total number of all nodes in the escape planning path. is the Manhattan distance.

[0263] The length factor of the escape planning path is taken into account. The larger the value, the straighter the escape planning path, the shorter the escape planning path, and the higher the probability of successful evacuation.

[0264] ,in, Refers to the number of dangerous nodes in the escape planning path; It refers to the number of disaster nodes in the escape planning path.

[0265] It describes the relative number of disaster nodes and dangerous nodes passed by the agent during the escape process. The higher the number, the fewer dangerous nodes and disaster nodes the agent passes on the escape planning path, and the safer the escape planning path is.

[0266] in, , It refers to the three-dimensional coordinates of the starting node of the i-th congested section; It refers to the three-dimensional coordinates of the terminal node of the i-th congested section.

[0267] The numerator represents the sum of the lengths of all congested sections, and the denominator represents the length of the entire escape planning path. The larger the value, the smoother the escape planning path and the higher the efficiency of personnel escape.

[0268] Since the entire escape planning path can only be scored after it is generated, at the end of each round, the agent will receive a second reward based on the escape planning path. At the end of the current round, that is, if the target second node is the escape exit node and the target time is less than or equal to the preset specified time, the formula , calculate R -1 ; Determine R -1 This is the second reward.

[0269] Among them, k2 is a preset scaling factor, k3 is a preset time coefficient, and t2 is the total time taken for the agent to travel from the initial node to the escape exit node.

[0270] Among them, K2 is used to prevent the difference between the two rewards obtained after the end of the current round from being too large. k3t2 is a time term, which has two main functions: on the one hand, it draws on the idea of ​​batch normalization to ensure that the proportion of each reward is reasonable; on the other hand, it considers the time index of the escape planning path, so that the intelligent agent is more inclined to choose the escape planning path with a shorter distance and the least time, thereby improving the quality of the escape planning path and speeding up the escape process.

[0271] Step B18: Inputting the second reward and the predicted Q value corresponding to the target action into the deep Q network model to update the deep Q network model.

[0272] Step B19: Obtain the target real Q value of the action of moving from the previous node to the next node in each pair of nodes with edges in the escape planning path.

[0273] The following example illustrates that, assuming the escape planning path is: node A→node B→node C→node D, the target real Q value of the action of moving from the previous node to the next node in each pair of nodes with an edge includes: the target real Q value of the action from node A to node B, the target real Q value of the action from node B to node C, and the target real Q value of the action from node C to node D.

[0274] Step B20: If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path increase successively according to the escape sequence, it is determined that the escape planning path is the optimal path.

[0275] Still taking the above as an example, the target real Q values ​​sorted according to the escape order are: the target real Q value of the action of moving from node A to node B, the target real Q value of the action of moving from node B to node C, and the target real Q value of the action of moving from node C to node D.

[0276] Step B21: If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path do not increase successively according to the escape order, determine that the agent returns to the initial node and return to step B07.

[0277] Step B22: If the update times of the deep Q network model reaches a preset update times, the parameters of the deep Q network model are assigned to the target Q network model.

[0278] For example, the preset number of updates may be determined based on actual conditions and is not limited here. For example, the preset number of updates is 15.

[0279] Step B23: If the optimal paths corresponding to all nodes have been obtained, it is determined that the dual deep Q network model training is completed.

[0280] The following describes a process of inputting the predicted Q value corresponding to the target reward and the target action into the dual deep Q network model to update the deep Q network model in the dual deep Q network model. The process includes the following steps C1 to C4.

[0281] Step C1: If the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset time, the formula , calculate and obtain the real Q value of the agent moving from the first node to the target second node.

[0282] in, is the true Q value, is the predicted Q value corresponding to the movement of the agent from the first node to the target second node output by the dual Q network model; lr is the preset learning rate parameter, γ is the preset discount factor, is a preset function.

[0283] Exemplarily, the preset function can be obtained artificially.

[0284] In the iterative update process of the DDQN model, it is mainly divided into two parts: the update of the true Q value and the update of the internal parameters of the deep Q network model. When updating the true Q value, the DDQN model is updated according to the target reward given by the environment so that the agent can obtain the maximum total reward during the escape process. Since the rewards obtained by future actions cannot be predicted in the current escape process, it is necessary to use the target Q network model to indirectly provide future rewards. In addition, the final reward cannot be obtained directly when the escape exit node is not reached, so it is necessary to introduce a discount factor γ to control the degree of discount of future rewards. The larger the γ, the more attention is paid to the reward obtained after the agent reaches the escape exit node; otherwise, the more attention is paid to the reward of the current single-step action. In order to escape smoothly, it is necessary to pay more attention to the potential reward obtained after reaching the escape exit node rather than the reward for moving one step, so γ should take a larger value. At the same time, it is also necessary to define the learning rate parameter lr, which indicates the degree to which future information covers current information. When lr is larger, it means that future rewards are more valued. In the present invention, these two parameters take fixed values ​​γ=0.95 and lr=0.01 respectively. Finally, the true Q value corresponding to the selected action is updated in a weighted manner .

[0285] Step C2: If the target second node is the escape exit node and the target time is less than or equal to the preset specified time, determine R -1 is the actual Q value of the agent moving from the first node to the target second node.

[0286] Step C3: If the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node, determine that the second preset reward value is the real Q value of the agent moving from the first node to the target second node.

[0287] If the agent reaches the escape exit node or does not reach the escape exit node or moves to the disaster node within the preset time, the current iteration round ends. -1 If the agent does not reach the escape exit node or move to the disaster node within the preset time, the true Q value is the second preset reward value, i.e. -100.

[0288] Step C4: By formula , update the deep Q network model.

[0289] In the DDQN model, the update of the true Q value and the internal parameters of the deep Q network model have different update methods. For the parameter update of the deep Q network model, the error back propagation method can be used to perform several rounds of iterative training on the deep Q network through the minimum mean square loss function, calculate the error between each predicted Q value and the true Q value, and back propagate the error back to the deep Q network model, so that the predicted Q value output by the deep Q network model at a given position is consistent with the true Q value, thereby completing the update of the parameters in the deep Q network model. The specific update method can be achieved by minimizing the loss function renew.

[0290] The structure of the deep Q network model is explained below.

[0291] like Fig.10 , which is a schematic diagram of the structure of the deep Q network model provided in an embodiment of the present application.

[0292] like Fig.10 As shown, the deep Q network model includes a first CSP (Cross Stage Partial) module 1001, a first CBS module 1002, a second CSP module 1003, a second CBS module 1004, a third CSP module 1005, and a third CBS module 1006 which are connected in sequence.

[0293] The structure of the first CSP module is described below. Fig.11 , which is a structural diagram of the first CBS module provided in an embodiment of the present application.

[0294] Among them, the first CSP module includes: a module 1101 carrying an unfold function, a fourth CBS module 1102 connected to the first output end of the module carrying the unfold function, a first BottleNeck module 1103 connected to the second output end of the module carrying the unfold function, and a second BottleNeck module 1104 connected to the third output end of the module carrying the unfold function; a fifth CBS module 1105 respectively connected to the output end of the fourth CBS module, the output end of the first BottleNeck module, and the output end of the second BottleNeck module, and a sixth CBS module 1106 connected to the fifth CBS module.

[0295] The first CBS module includes a convolutional layer, a batch normalization layer (BN2D) connected to the convolutional layer (CONV), and an activation function layer (SWISH) connected to the batch normalization layer.

[0296] Among them, the first CSP module, the second CSP module and the third CSP module have the same structure; the first CBS module, the second CBS module, the third CBS module, the fourth CBS module, the fifth CBS module and the sixth CBS module have the same structure.

[0297] The second BottleNeck module has the same structure as the first BottleNeck module.

[0298] The first BottleNeck module includes three LBS modules connected in sequence. The LBS module includes: a fully connected layer (LINEAR), a batch normalization layer (BN2D) connected to the fully connected layer, and an activation function layer (SWISH) connected to the batch normalization layer.

[0299] The internal structure of the deep Q network model provided in this application includes multiple modules. Among them, the LBS module is mainly used to extract the basic features related to the state sequence of each node in the personnel network flow evacuation model. The CBS module (the first CBS module to the sixth CBS module are collectively referred to as the CNS module) is mainly used to extract cross-node features between the input states corresponding to different nodes, such as the time comparison information of the agent crossing the node during the evacuation process. The BottleNeck module (the first BottleNeck module and the second BottleNeck module are collectively referred to as the BottleNeck module) is also called the bottleneck module. It reduces the computational complexity of the deep Q network model by first reducing the dimension of the feature and then increasing the dimension, and effectively extracts the independent feature information in the input state corresponding to the node, such as evacuation time and predicted Q value. The CSP module (the first CSP module to the third CSP module are collectively referred to as the CSP module) includes multiple BottleNeck modules and CBS modules, which can extract both independent features of nodes and cross-node features between nodes. Finally, these feature vectors are spliced ​​and the number of pathways is changed through the CBS module.

[0300] The structure of the deep Q network model consists of 3 CSP modules and 3 CBS modules. Each CSP module is followed by a CBS module and stacked in this way. The input of the deep Q network model is the state vector corresponding to the node (i.e., the personnel evacuation network flow model, distance matrix, density matrix, speed matrix, and evacuation time matrix). The number of paths in each layer changes in the pattern of 1-64-128-128-64-1. The output of the deep Q network model is a one-dimensional list, in which each element represents the predicted Q value of the agent taking different actions after moving to the first node. These predicted Q values ​​can be used to evaluate which action the agent takes to maximize the benefit after moving to the first node, thereby training the optimal path.

[0301] It is understandable that the DDQN model includes a deep Q network model and a target Q network model. The purpose is to first find the action A corresponding to the largest predicted Q value in the deep Q network model, and then use the action A to calculate the target Q value in the target Q network model. Usually, the deep Q network model and the target Q network model need to maintain weight sharing, so the structural design of the two should also be the same, and the network parameters of the deep Q network model will be regularly synchronized to the target Q network model.

[0302] Figure 6 Taking a bank branch as an example, a fire emergency evacuation indication method applied to a large indoor scene is introduced. First, a bank branch with a business hall, office area, and basement is taken as an actual evacuation scene. The network flow is converted into a personnel evacuation network flow model composed of nodes and edges, and the corresponding model parameters are set. Secondly, based on the deep reinforcement learning DDQN model, the personnel network flow model is abstracted into a distance matrix, a density matrix, an evacuation time matrix, and a speed matrix. By designing state and action elements, reward functions, and DDQN models, static evacuation routes are planned, aiming to provide evacuation path plans as soon as possible after a disaster occurs and guide personnel to escape quickly. Finally, according to the internal sensor detection system of the bank branch, detailed information on the real-time disaster situation is obtained, and full consideration is given to issues such as smoke diffusion, disaster spread, and personnel congestion. Dynamic adjustments are made on the basis of the static evacuation path, and a dynamic evacuation path is planned in real time. On-site auxiliary equipment is used to guide personnel to escape through voice broadcasts and screen prompts. This process is repeated until all personnel have completed the evacuation.

[0303] In an optional implementation, the data center obtains real-time environmental parameters (temperature, smoke, population density, etc.) during the evacuation process through sensors installed inside the bank branch, and stores these parameters in the database.

[0304] The fire emergency evacuation indication method for large indoor scenes provided by the embodiment of the present application is highly versatile and applicable to emergency evacuation indications of all bank branches including bank outlets. Different evacuation scenarios only need to replace the corresponding personnel network flow evacuation model without modifying the algorithm process and system structure. In the process of instructing personnel to evacuate, the present application not only considers the length factor of the escape planning path, but also takes the evacuation time as an indicator, focusing on the overall evacuation efficiency, which is more realistic. The deep reinforcement learning DDQN model is used to obtain the escape planning path, which has a stronger non-steady data fitting ability and a faster convergence speed than the traditional path planning algorithm. The quality of the escape planning path is taken into account by designing a reward function based on a path scorer, which improves the safety during the escape process. In addition, dynamic environmental information such as fire spread, smoke spread and personnel density is also considered, and the escape planning path can be dynamically adjusted in time, which is more in line with the actual emergency evacuation needs of bank outlets.

[0305] The embodiment of the present application uses the DDQN model. The DDQN model is an algorithm in the field of deep reinforcement learning. It does not have a specific training set and needs to continuously interact with the environment to obtain samples. It has a deep Q network model and a target Q network model, and can fit more complex states. In order to apply the DDQN model to the evacuation of personnel in large indoor scenes, it is necessary to define several basic concepts of the DDQN model in combination with the evacuation path planning background of large indoor scenes (taking bank branches as an example).

[0306] Agent: The agent plays the role of evacuees and needs to constantly choose paths to find the optimal escape path. It is the entity that performs actions.

[0307] Environment: The environment refers to the office area, corridor intersections, bank self-service terminal equipment, fire stairs, vaults and other environmental information inside the bank branch. It interacts with the agent by providing feasible nodes and paths.

[0308] State: The state is the information that the agent directly observes from the environment and is the only basis for making decisions. It mainly includes data such as the time required for the current node to reach the next adjacent node, the length of the path, the number of people on the path, and the speed of people moving.

[0309] Action: The expression of the agent's execution behavior. In the bank branch evacuation scenario, it mainly refers to the agent's behavior of moving from the current node to the next adjacent node.

[0310] Reward: The feedback evaluation of the agent performing a certain action during the evacuation process.

[0311] like Fig.12 The figure is a schematic diagram of the interaction process between the intelligent agent and the environment of a bank branch provided in an embodiment of the present application.

[0312] Fig.12 The figure shows the process of interaction between the agent and the bank branch environment. At time t1, the agent is in the current state S t1 Next take action A t1 , get reward R t1 , and enter the next state. Repeating this process of interaction between the agent and the environment will eventually form multiple experience sequences Based on these experience sequences, the agent will learn the optimal path from the current node to the escape exit node.

[0313] A fire emergency evacuation indication method for a large indoor scene provided in an embodiment of the present application is introduced above. A device for executing the fire emergency evacuation indication method for a large indoor scene will be introduced below.

[0314] See also Fig.13 , Fig.13 This is a schematic diagram of the structure of a fire emergency evacuation indication device for use in a large indoor scene provided by an embodiment of the present application. Fig.13 As shown, the fire emergency evacuation indication device applied to large indoor scenes includes:

[0315] The first acquisition module 1201 is used to acquire a personnel evacuation network flow model, wherein the personnel evacuation network flow model includes a plurality of nodes and edges between two nodes; the edges between two nodes represent passages between two nodes for personnel to walk; the plurality of nodes include escape exit nodes, ordinary nodes, and stair nodes;

[0316] The second acquisition module 1202 is used to acquire a distance matrix corresponding to the personnel evacuation network flow model; the distance matrix includes the distances between all nodes with edges in the personnel evacuation network flow model;

[0317] The third acquisition module 1203 is used to acquire a density matrix corresponding to the personnel evacuation network flow model if a fire is detected; the density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges;

[0318] A first determination module 1204 is used to determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs;

[0319] The second determination module 1205 is configured to determine that the edge between the two nodes is a congested section if the pedestrian flow density of the path represented by the edge between the two nodes is greater than or equal to a second preset density value;

[0320] A first updating module 1206 is used to update the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model;

[0321] The fourth acquisition module 1207 is used to acquire a speed matrix corresponding to the personnel evacuation network flow model based on the density matrix; the speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model;

[0322] A fifth acquisition module 1208 is used to acquire an evacuation time consumption matrix corresponding to the personnel evacuation network flow model based on the distance matrix and the speed matrix;

[0323] The sixth acquisition module 1209 is used to input the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtain the first escape paths corresponding to all nodes respectively through the evacuation path prediction model;

[0324] A first sending module 1210 is used to send, for each of the nodes, the first escape path corresponding to the node to a terminal for escape corresponding to the node;

[0325] The seventh acquisition module 1211 is used to obtain an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix if it is detected that the human flow density corresponding to any of the nodes changes, or the human flow density of the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation of any of the nodes changes;

[0326] An eighth acquisition module 1212 is used to input the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix and the updated evacuation time matrix into the evacuation path prediction model, and obtain a second escape path through the evacuation path prediction model;

[0327] The second sending module 1213 is used to send, for each of the nodes, the second escape path corresponding to the node to the terminal for escape corresponding to the node.

[0328] In an optional implementation, the method further includes:

[0329] A third determination module is used for detecting, for each of the nodes, whether the node has changed from a disaster-occurring location to a non-disaster-occurring location, or from a non-disaster-occurring location to a disaster-occurring location, and if so, determining that the fire situation of the node has changed;

[0330] a fourth determination module, configured to detect, for each of the nodes, a density of human traffic at the node through a sensor deployed at the node, and determine that the density of human traffic at the node has changed if the density of human traffic changes from being greater than or equal to the first preset density value to being less than the first preset density value, or from being less than the first preset density value to being greater than or equal to the first preset density value;

[0331] A fifth determination module is used for detecting, for each pairwise edge of the nodes, a density of human traffic located in the passage through sensors deployed in the passage between the nodes, and determining that the density of human traffic in the passage has changed if the density of human traffic changes from being greater than or equal to a second preset density value to being less than the second preset density value, or from being less than the second preset density value to being greater than or equal to the second preset density value;

[0332] The sixth determination module is used to detect the smoke concentration of the passage between the two nodes for each pair of node edges by using smoke sensors deployed in the passage between the two nodes; if the smoke concentration level corresponding to the smoke concentration changes, it is determined that the moving speed of people in the passage between the two nodes has changed.

[0333] In an optional implementation, the smoke concentration is CO concentration, and the seventh acquisition module includes:

[0334] The first calculation unit is used for each pairwise edge of nodes, using the formula:

[0335] The coefficient α is calculated; ρ co is the CO concentration of the pathway between the two nodes;

[0336] The first determination unit is used to determine, for each pairwise edge of each node, that the updated moving speed corresponding to the pairwise edge of the node is the product of the moving speed corresponding to the pairwise edge of the node before updating and the coefficient α.

[0337] In an optional implementation, the evacuation path prediction model is a dual deep Q network model, further comprising:

[0338] A ninth acquisition module, used to acquire a sample density matrix corresponding to the personnel evacuation network flow model; the sample density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges;

[0339] A tenth acquisition module, used for acquiring a sample speed matrix corresponding to the personnel evacuation network flow model based on the sample density matrix; the sample speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model;

[0340] An eleventh acquisition module is used to acquire a sample evacuation time consumption matrix corresponding to the personnel evacuation network flow model based on the sample distance matrix and the sample speed matrix;

[0341] A seventh determination module is used to determine a dangerous node and a disaster node from the plurality of nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs;

[0342] An eighth determination module, configured to determine that the edge between the two nodes is a congested section if the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value;

[0343] A second updating module is used to update the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model;

[0344] A twelfth acquisition module, used to acquire state information of the agent, wherein the state information includes the time taken to move from the first node to each of the second nodes;

[0345] A thirteenth acquisition module is used to input the personnel evacuation network flow model, the distance matrix, the sample density matrix, the sample speed matrix, the sample evacuation time matrix and the state information into the dual deep Q network model, and output the predicted Q values ​​corresponding to multiple actions respectively through the dual deep Q model; the dual deep Q network model includes a deep Q network model and a target Q network model; the action is the action of the agent moving from the first node to the second node;

[0346] A fourteenth acquisition module, used to obtain a random number;

[0347] A first selection module, configured to randomly select a target action from the multiple actions if the random number is greater than a preset threshold, wherein the second node in the target action is the target second node;

[0348] A second selection module is used to select a target action corresponding to the maximum expected Q value from the multiple actions if the random number is less than or equal to the preset threshold; the second node in the target action is the target second node;

[0349] A ninth determination module, configured to determine a first reward for the agent to move to the target second node based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken for the agent to reach the target second node from the initial node exceeds a preset prescribed time length;

[0350] A third updating module, configured to input the predicted Q value corresponding to the first reward and the target action into the dual deep Q network model to update the deep Q network model in the dual deep Q network model;

[0351] A first triggering module, configured to trigger a twelfth obtaining module if the target second node is not the disaster node and is not the escape exit node, and the target time consumption does not exceed the preset prescribed time length;

[0352] A second triggering module is used to determine that the agent returns to the initial node and trigger a twelfth acquisition module if the target second node is the disaster node or the target time exceeds the preset specified time length;

[0353] A fifteenth acquisition module is used to acquire an escape planning path if the target second node is the escape exit node, where the escape planning path is a path for the agent to reach the escape exit node from the initial node;

[0354] a tenth determination module, configured to determine a second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path;

[0355] a fourth updating module, configured to input the second reward and the predicted Q value corresponding to the target action into the deep Q network model to update the deep Q network model;

[0356] A sixteenth acquisition module is used to obtain the target real Q value of the action of moving from the previous node to the next node in the escape planning path between two nodes with edges;

[0357] an eleventh determination module, for determining that the escape planning path is the optimal path if the target real Q values ​​corresponding to the two nodes with edges in the escape planning path increase in sequence according to the escape order;

[0358] A twelfth determination module is used to determine that the agent returns to the initial node and return to the step of obtaining the state information of the agent if the target real Q values ​​corresponding to the two nodes with edges in the escape planning path according to the escape order do not increase in sequence;

[0359] An assignment module, configured to assign the parameters of the deep Q network model to the target Q network model if the update number of the deep Q network model reaches a preset update number;

[0360] The thirteenth determination module is used to determine that the dual deep Q network model training is completed if the optimal paths corresponding to all nodes have been obtained.

[0361] In an optional implementation, the ninth determining module includes:

[0362] The second calculation unit is used to calculate the target second node if it is not the disaster node and the escape exit node and the target time is less than the preset specified time. , calculate R t1 ; Determine R t1 is the first reward; k1 is the preset reward magnification factor, k1 is a positive number greater than 0, means the shortest distance; refers to the shortest distance between the first node and the escape exit;

[0363] The second determining unit is used to determine a second preset reward value as the first reward if the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node.

[0364] In an optional implementation, the tenth determining module includes:

[0365] The third calculation unit is used to calculate the target second node if it is the escape exit node and the target time is less than or equal to the preset time. , calculate R -1 ; Determine R -1 For the second reward;

[0366] Wherein, k2 is a preset scaling factor, k3 is a preset time coefficient, and t2 is the total time taken by the agent to travel from the initial node to the escape exit node;

[0367] in, , ,in is the three-dimensional coordinate of the initial node; The three-dimensional coordinates of the escape exit node closest to the initial node; is the three-dimensional coordinate of the jth node in the escape planning path; N is the total number of all nodes in the escape planning path; ,in, Refers to the number of dangerous nodes in the escape planning path; Refers to the number of disaster nodes in the escape planning path;

[0368] in, , It refers to the three-dimensional coordinates of the starting node of the i-th congested section; It refers to the three-dimensional coordinates of the terminal node of the i-th congested section.

[0369] In an optional implementation, the target reward is the first reward or the second reward, and the third updating module or the fourth updating module includes:

[0370] The fourth calculation unit is used to calculate the target second node if it is not the disaster node and the escape exit node and the target time is less than the preset time. , calculate the real Q value of the agent moving from the first node to the target second node;

[0371] in, is the true Q value, is the predicted Q value corresponding to the movement of the agent from the first node to the target second node output by the dual Q network model; lr is the preset learning rate parameter, γ is the preset discount factor, is the preset function;

[0372] The third determining unit is used to determine R if the target second node is the escape exit node and the target time is less than or equal to the preset specified time. -1 is the actual Q value of the agent moving from the first node to the target second node;

[0373] A fourth determining unit is used to determine that the second preset reward value is a real Q value of the agent moving from the first node to the target second node if the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node;

[0374] Update cells for the formula , update the deep Q network model.

[0375] In an optional implementation, the deep Q network model includes a first CSP module, a first CBS module, a second CSP module, a second CBS module, a third CSP module, and a third CBS module connected in sequence;

[0376] The first CSP module includes: a module carrying an unfold function, a fourth CBS module connected to a first output end of the module carrying the unfold function, a first BottleNeck module connected to a second output end of the module carrying the unfold function, and a second BottleNeck module connected to a third output end of the module carrying the unfold function; a fifth CBS module connected to an output end of the fourth CBS module, an output end of the first BottleNeck module, and an output end of the second BottleNeck module, respectively, and a sixth CBS module connected to the fifth CBS module;

[0377] The first CBS module includes a convolutional layer, a batch normalization layer connected to the convolutional layer, and an activation function layer connected to the batch normalization layer;

[0378] Among them, the first CSP module, the second CSP module and the third CSP module have the same structure; the first CBS module, the second CBS module, the third CBS module, the fourth CBS module, the fifth CBS module and the sixth CBS module have the same structure.

[0379] The embodiment of the present application also provides a fire emergency evacuation indication system applied to a large indoor scene. The system mainly includes three parts. Fig.14 As shown. Fig.14 , which is a schematic diagram of a fire emergency evacuation indication system applied to a large indoor scene provided by an embodiment of the present application.

[0380] Part I: Personnel evacuation model, illustratively, the personnel evacuation model includes: personnel evacuation network flow model, node attributes and edge attributes. Among them, the node attribute refers to whether the node is a dangerous node or a disaster node; the edge attribute refers to whether the edge is a congested section.

[0381] Part II: Setting up a static evacuation strategy. The static evacuation strategy includes but is not limited to: DDQ model environment elements, reward function, dual Q network model, and optimal path determination.

[0382] Exemplarily, the experimental platform refers to a platform where the electronic device is located. Exemplarily, the data center is a storage space for information collected by sensors.

[0383] The third part: actual emergency evacuation system, through the actual emergency evacuation system combined with real-time environmental information, obtain the dynamic evacuation strategy, that is, the dynamic evacuation path.

[0384] Exemplarily, the actual emergency evacuation system includes: a plurality of sensors, on-site auxiliary equipment, for example, a terminal for escape

[0385] An embodiment of the present application also provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any fire emergency evacuation indication method applied to a large indoor scene provided in the embodiment of the present application.

[0386] A computer-readable storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any fire emergency evacuation indication method applied to large indoor scenes provided in an embodiment of the present application.

[0387] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0388] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0389] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0390] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A fire emergency evacuation indication method applied to a large indoor scene, characterized in that: include: Obtaining a personnel evacuation network flow model, wherein the personnel evacuation network flow model includes a plurality of nodes and edges between two nodes; the edges between two nodes represent passages between two nodes for personnel to walk; the plurality of nodes include escape exit nodes, ordinary nodes, and stair nodes; Obtaining a distance matrix corresponding to the personnel evacuation network flow model; the distance matrix includes distances between all pairs of nodes with edges in the personnel evacuation network flow model; If a fire is detected, a density matrix corresponding to the personnel evacuation network flow model is obtained; the density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges; Determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs; If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, determining that the edge between the two nodes is a congested section; Updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model; Based on the density matrix, a speed matrix corresponding to the personnel evacuation network flow model is obtained; the speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model; Based on the distance matrix and the speed matrix, obtaining an evacuation time consumption matrix corresponding to the personnel evacuation network flow model; Inputting the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtaining the first escape paths corresponding to all nodes respectively through the evacuation path prediction model; For each of the nodes, sending the first escape path corresponding to the node to a terminal for escape corresponding to the node; If it is detected that the density of human traffic corresponding to any of the nodes changes, or the density of human traffic in the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation at any of the nodes changes, an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix are obtained; Inputting the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix and the updated evacuation time matrix into the evacuation path prediction model, and obtaining a second escape path through the evacuation path prediction model; For each of the nodes, the second escape path corresponding to the node is sent to a terminal for escape corresponding to the node.

2. The fire emergency evacuation indication method according to claim 1, characterized in that: Also includes: For each of the nodes, a smoke sensor deployed at the node is used to detect whether the node has changed from a disaster location to a non-disaster location, or from a non-disaster location to a disaster location. If so, it is determined that the fire situation of the node has changed; For each of the nodes, a sensor deployed at the node is used to detect a density of human traffic at the node. If the human traffic density changes from being greater than or equal to the first preset density value to being less than the first preset density value, or from being less than the first preset density value to being greater than or equal to the first preset density value, it is determined that the human traffic density of the node has changed. For each pairwise edge of the nodes, the density of human traffic located in the passage is detected by using sensors deployed in the passage between the pairs of nodes. If the human traffic density changes from being greater than or equal to a second preset density value to being less than the second preset density value, or from being less than the second preset density value to being greater than or equal to the second preset density value, it is determined that the human traffic density of the passage has changed. For each pairwise node edge, the smoke concentration of the passage is detected by a smoke sensor deployed in the passage between the two nodes; if the smoke concentration level corresponding to the smoke concentration changes, it is determined that the moving speed of the personnel in the passage between the two nodes has changed.

3. The fire emergency evacuation indication method according to claim 2, characterized in that: The smoke concentration is CO concentration, and for each edge of two nodes, a method for obtaining an updated moving speed corresponding to a path between the two nodes includes: By formula The coefficient α is calculated; ρ co is the CO concentration of the pathway between the two nodes; The updated moving speed is determined to be the product of the moving speed before updating and the coefficient α.

4. The fire emergency evacuation indication method according to any one of claims 1 to 3, characterized in that: The evacuation path prediction model is a dual-depth Q network model, and the method for training the dual-depth Q network includes: Obtaining a sample density matrix corresponding to the personnel evacuation network flow model; the sample density matrix includes the flow density of people corresponding to all nodes in the personnel evacuation network flow model and the flow density of people between nodes with edges; Based on the sample density matrix, a sample speed matrix corresponding to the personnel evacuation network flow model is obtained; the sample speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model; Based on the sample distance matrix and the sample speed matrix, obtaining a sample evacuation time consumption matrix corresponding to the personnel evacuation network flow model; Determine a dangerous node and a disaster node from the multiple nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs; If the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value, determining that the edge between the two nodes is a congested section; Updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model; Acquire state information of the agent, wherein the state information includes the time taken to move from the first node to each of the second nodes; The personnel evacuation network flow model, the distance matrix, the sample density matrix, the sample speed matrix, the sample evacuation time matrix and the state information are input into the dual deep Q network model, and the predicted Q values ​​corresponding to multiple actions are output through the dual deep Q model; the dual deep Q network model includes a deep Q network model and a target Q network model; the action is the action of the agent moving from the first node to the second node; Get a random number; If the random number is greater than a preset threshold, randomly selecting a target action from the multiple actions, and the second node in the target action is the target second node; If the random number is less than or equal to the preset threshold, a target action corresponding to the maximum expected Q value is selected from the multiple actions; the second node in the target action is the target second node; Determine a first reward for the agent to move to the target second node based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken for the agent to reach the target second node from the initial node exceeds a preset prescribed time length; Inputting the first reward and the predicted Q value corresponding to the target action into the dual deep Q network model to update the deep Q network model in the dual deep Q network model; If the target second node is not the disaster node and is not the escape exit node, and the target time does not exceed the preset specified time, return to the step of obtaining the state information of the intelligent agent; If the target second node is the disaster node or the target time exceeds the preset specified time, determine that the agent returns to the initial node, and return to the step of obtaining the state information of the agent; If the target second node is the escape exit node, obtain an escape planning path, where the escape planning path is a path for the agent to reach the escape exit node from the initial node; Determining a second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path; Inputting the second reward and the predicted Q value corresponding to the target action into the deep Q network model to update the deep Q network model; Obtaining a target real Q value of an action of moving from a previous node to a next node in each pair of nodes having an edge in the escape planning path; If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path increase successively according to the escape order, it is determined that the escape planning path is the optimal path; If the target real Q values ​​corresponding to the two nodes with edges in the escape planning path do not increase in sequence according to the escape sequence, it is determined that the agent returns to the initial node, and the step of obtaining the state information of the agent is returned; If the update number of the deep Q network model reaches a preset update number, assigning the parameters of the deep Q network model to the target Q network model; If the optimal paths corresponding to all nodes have been obtained, it is determined that the dual deep Q network model training is completed.

5. The fire emergency evacuation indication method according to claim 4, characterized in that: The first reward step for determining the movement of the agent to the target second node based on the shortest distance between the target second node and the escape exit node, whether the target second node is the disaster node, whether the target second node is the dangerous node, whether the target second node is the escape exit node, and whether the target time taken for the agent to reach the target second node from the initial node exceeds a preset prescribed time, includes: If the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset specified time, the formula , calculate R t1 ; Determine R t1 is the first reward; k1 is the preset reward magnification factor, k1 is a positive number greater than 0, means the shortest distance; refers to the shortest distance between the first node and the escape exit; If the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node, a second preset reward value is determined to be the first reward.

6. The fire emergency evacuation indication method according to claim 5, characterized in that: The step of determining the second reward based on the number of disaster nodes in the escape planning path, the number of dangerous nodes in the escape planning path, the number of congested sections in the escape planning path, and the length of the escape planning path includes: If the target second node is the escape exit node and the target time is less than or equal to the preset time, the formula , calculate R -1 ; Determine R -1 For the second reward; Wherein, k2 is a preset scaling factor, k3 is a preset time coefficient, and t2 is the total time taken by the agent to travel from the initial node to the escape exit node; in, , ,in is the three-dimensional coordinate of the initial node; The three-dimensional coordinates of the escape exit node closest to the initial node; is the three-dimensional coordinate of the jth node in the escape planning path; N is the total number of all nodes in the escape planning path; ,in, Refers to the number of dangerous nodes in the escape planning path; Refers to the number of disaster nodes in the escape planning path; in, , It refers to the three-dimensional coordinates of the starting node of the i-th congested section; It refers to the three-dimensional coordinates of the terminal node of the i-th congested section.

7. The fire emergency evacuation indication method according to claim 6, characterized in that: The target reward is the first reward or the second reward, and the target reward and the predicted Q value corresponding to the target action are input into the dual deep Q network model to update the deep Q network model in the dual deep Q network model, comprising: If the target second node is not the disaster node and is not the escape exit node and the target time is less than the preset specified time, the formula , calculate the real Q value of the agent moving from the first node to the target second node; in, is the true Q value, is the predicted Q value corresponding to the movement of the agent from the first node to the target second node output by the dual Q network model; lr is the preset learning rate parameter, γ is the preset discount factor, is the preset function; If the target second node is the escape exit node and the target time is less than or equal to the preset specified time, determine R -1 is the actual Q value of the agent moving from the first node to the target second node; If the target second node is not an escape exit node and the target time is greater than the preset prescribed time, or the target second node is the disaster node, the second preset reward value is determined to be the real Q value of the agent moving from the first node to the target second node; By formula , update the deep Q network model.

8. The fire emergency evacuation indication method according to claim 4, characterized in that: The deep Q network model includes a first CSP module, a first CBS module, a second CSP module, a second CBS module, a third CSP module, and a third CBS module connected in sequence; The first CSP module includes: a module carrying an unfold function, a fourth CBS module connected to a first output end of the module carrying the unfold function, a first BottleNeck module connected to a second output end of the module carrying the unfold function, and a second BottleNeck module connected to a third output end of the module carrying the unfold function; a fifth CBS module connected to an output end of the fourth CBS module, an output end of the first BottleNeck module, and an output end of the second BottleNeck module, respectively, and a sixth CBS module connected to the fifth CBS module; The first CBS module includes a convolutional layer, a batch normalization layer connected to the convolutional layer, and an activation function layer connected to the batch normalization layer; Among them, the first CSP module, the second CSP module and the third CSP module have the same structure; the first CBS module, the second CBS module, the third CBS module, the fourth CBS module, the fifth CBS module and the sixth CBS module have the same structure.

9. A fire emergency evacuation indication device applied to a large indoor scene, characterized in that: include: A first acquisition module is used to acquire a personnel evacuation network flow model, wherein the personnel evacuation network flow model includes a plurality of nodes and edges between two nodes; the edges between two nodes represent passages between two nodes for personnel to walk; the plurality of nodes include escape exit nodes, ordinary nodes, and stair nodes; A second acquisition module is used to acquire a distance matrix corresponding to the personnel evacuation network flow model; the distance matrix includes distances between all pairs of nodes with edges in the personnel evacuation network flow model; A third acquisition module is used to acquire a density matrix corresponding to the personnel evacuation network flow model if a fire is detected; the density matrix includes the human flow density corresponding to all nodes in the personnel evacuation network flow model and the human flow density of the paths between two nodes with edges; A first determination module is used to determine a dangerous node and a disaster node from the plurality of nodes; the dangerous node is a node where the personnel flow density is greater than or equal to a first preset density value, and the disaster node is a node where a fire occurs; A second determination module is used to determine that the edge between the two nodes is a congested section if the pedestrian flow density of the passage represented by the edge between the two nodes is greater than or equal to a second preset density value; A first updating module, used for updating the dangerous nodes, the disaster nodes, and the congested road sections to the personnel evacuation network flow model; A fourth acquisition module is used to acquire a speed matrix corresponding to the personnel evacuation network flow model based on the density matrix; the speed matrix includes the movement speeds of personnel in all paths between two nodes with edges in the personnel evacuation network flow model; A fifth acquisition module, configured to acquire an evacuation time consumption matrix corresponding to the personnel evacuation network flow model based on the distance matrix and the speed matrix; A sixth acquisition module, used to input the personnel evacuation network flow model, the distance matrix, the density matrix, the speed matrix and the evacuation time matrix into a pre-built evacuation path prediction model, and obtain the first escape paths corresponding to all nodes respectively through the evacuation path prediction model; A first sending module, configured to send, for each of the nodes, the first escape path corresponding to the node to a terminal for escape corresponding to the node; A seventh acquisition module is used to obtain an updated personnel evacuation network flow model, an updated density matrix, an updated speed matrix, and an updated evacuation time matrix if it is detected that the human flow density corresponding to any of the nodes changes, or the human flow density of the passage between any two nodes changes, or the moving speed of personnel in the passage between any two nodes changes, or the fire situation of any of the nodes changes; an eighth acquisition module, configured to input the updated personnel evacuation network flow model, the distance matrix, the updated density matrix, the updated speed matrix, and the updated evacuation time matrix into the evacuation path prediction model, and acquire a second escape path through the evacuation path prediction model; The second sending module is used to send, for each of the nodes, the second escape path corresponding to the node to the terminal for escape corresponding to the node.

10. A computer program product, characterized in that It includes computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the fire emergency evacuation indication method applied to a large indoor scene as described in any one of claims 1 to 8.

Images