UE5-based ship fire evacuation simulation method

Abstract

The invention relates to a UE5-based ship fire evacuation simulation method. The method comprises the following steps: S1, constructing a mobile virtual agent in a virtual ship cabin scene; s2, setting an evacuation area matrix and an evacuation area connection condition matrix of a virtual ship cabin; s3, constructing a mobile virtual agent fire evacuation model based on the evacuation area matrix and the evacuation area connection condition matrix; and S4, solving the evacuation model to obtain an evacuation path, and controlling the mobile virtual agent to perform evacuation simulation according to the evacuation path. Compared with the prior art, the method has the advantages of optimizing a fire evacuation route and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of fire evacuation simulation, and in particular to a ship fire evacuation simulation method based on UE5. Background Art

[0002] How to effectively prevent and respond to fire and smoke accidents has become an urgent problem to be solved. There are various computer evacuation models to choose from, and the most commonly used by fire safety practitioners is the agent-based continuous model. For the effectiveness of any evacuation model results, it is important that the input values are applicable to the simulated scenario. Currently, the main methods for extracting pedestrian input values are on-site or laboratory simulation studies. Real experiments have intuitiveness and authenticity, and can directly observe and measure experimental phenomena, with intuitive results and high authenticity. This helps researchers deeply understand the experimental principles, discover potential problems, and verify theoretical predictions, but the required cost is high, and some complex experiments may also require special experimental conditions and equipment, further increasing the cost. Real experiments are often restricted by time, location, and resources, and some experiments may involve hazardous substances or operations, presenting certain safety risks. Researchers need to take strict safety measures to ensure the safety and reliability of the experimental process. Laboratory simulation experiments, although low-cost and not restricted by time, location, and resources, can be carried out anytime and anywhere. However, compared with real experiments, they lack intuitiveness and still need to be verified and confirmed through real experiments. This increases the complexity and time cost of the research.

[0003] In addition, existing fire evacuation simulations do not consider the possible injuries that people may suffer when passing through different evacuation areas during the actual evacuation process, resulting in the generated evacuation routes not being optimal. Summary of the Invention

[0004] The purpose of the present invention is to provide a ship fire evacuation simulation method based on UE5 for optimizing fire evacuation routes.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A ship fire evacuation simulation method based on UE5, the method comprising the following steps:

[0007] S1. Construct a moving virtual agent in the virtual ship cabin scene;

[0008] S2. Set the evacuation area matrix and the evacuation area connection situation matrix of the virtual ship cabin;

[0009] S3. Construct a moving virtual agent fire evacuation model based on the evacuation area matrix and the evacuation area connection situation matrix;

[0010] S4. Solve the evacuation model to obtain the evacuation path, and control the mobile virtual agent to perform evacuation simulation according to the evacuation path.

[0011] Further, the evacuation area matrix represents a vector composed of each evacuation area of the virtual ship's cabin, and each evacuation area corresponds to a node.

[0012] Further, the element in the i-th row and j-th column of the evacuation area connection situation matrix represents the connection situation between evacuation area i and evacuation area j.

[0013] Further, the evacuation model takes minimizing evacuation damage as the objective function.

[0014] Further, the objective function is:

[0015]

[0016] where D is the evacuation damage, i is the i-th evacuation area, N is the total number of evacuation areas in the ship's cabin fire environment, T ij corresponding edge ν ij time attribute of, is the time attribute of the corresponding edge ν ij x is the damage attribute of, ij is the usage of edge ν ij usage.

[0017] Further, the time attribute of the corresponding edge ν ij represents the time required for the mobile virtual agent to pass through or leave area i and reach the next area j, that is, the residence time in area i.

[0018] Further, the damage attribute of the corresponding edge ν ij represents the damage value of the fire smoke suffered by the mobile virtual agent when passing through evacuation area i due to standing upright, bending down, and temperature.

[0019] Further, when the usage x ij = 1, it means that the evacuation path includes edge v ij , when the usage x ij = 0, it means that the evacuation path does not include edge v ij .

[0020] Further, the constraint conditions of the evacuation model include threshold constraint, connectivity constraint, and acyclic constraint.

[0021] Further, the threshold constraint is:

[0022] if, x ij = 1

[0023] where, Indicate the lethal values of CO, CO2, and temperature at the standing and bending heights in each evacuation area of the evacuation path for each evacuee;

[0024] The connectivity constraint is:

[0025] x ij ≤y ij

[0026] where y ij indicates the connection situation between evacuation area i and evacuation area j. y ij = 1 means that evacuation area i and evacuation area j are connected, and y ij = 0 means that evacuation area i and evacuation area j are not connected;

[0027] The non-circular constraint is:

[0028]

[0029] where x inter,j and x j,inter represent the intermediate nodes of the final evacuation path.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] By constructing a model of the evacuation path, the present invention considers the possible injuries that personnel may suffer in different evacuation areas, takes minimizing the total injury as the objective function, and at the same time considers the connectivity of each evacuation area, the path will not form a loop, and ensures that the injury values of standing, bending, and temperature in each evacuation area of the evacuation path of each virtual agent are lower than the threshold, effectively optimizing the fire evacuation route and making the route more practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a schematic diagram of the virtual reality scene of the ship corridor without the influence of fire smoke according to the present invention;

[0033] Figure 2 is a schematic diagram of the virtual reality scene of the ship corridor under the condition of limited fire smoke vision according to the present invention;

[0034] Figure 3 is a schematic diagram of the virtual reality scene of the ship corridor with different smoke passing heights according to the present invention;

[0035] Figure 4 is a schematic diagram of the scene with static obstacles imported according to the present invention;

[0036] Figure 5 is a diagram of the NPC grid and animation output results according to the present invention;

[0037] Figure 6Blueprint construction diagram for the trigger event of the present invention;

[0038] Figure 7 Blueprint construction diagram for the participants of the present invention to maintain a fixed interpersonal distance from NPCs;

[0039] Figure 8 Schematic diagram of the code flow of the present invention;

[0040] Figure 9 Composition diagram of the receiving control processing module;

[0041] Figure 10 Speed data diagram of the barrier-free experiment record;

[0042] Figure 11 Trajectory data diagram of the experiment with obstacles record;

[0043] Figure 12 Head deflection angle data diagram of the experiment with obstacles record. Detailed implementation mode

[0044] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and the detailed implementation mode and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.

[0045] The present invention proposes a ship fire evacuation simulation method based on UE5, and the method includes the following steps:

[0046] S1. Construct a mobile virtual agent in the virtual ship cabin scene;

[0047] S2. Set the evacuation area matrix and the evacuation area connection situation matrix of the virtual ship cabin;

[0048] S3. Construct a mobile virtual agent fire evacuation model based on the evacuation area matrix and the evacuation area connection situation matrix;

[0049] S4. Solve the evacuation model to obtain the evacuation path, and control the mobile virtual agent to perform evacuation simulation according to the evacuation path.

[0050] The schematic diagram of the code flow of the present invention is as Figure 8As shown. The present invention applies the above method to an actual system. The system includes a blueprint scripting system, a Niagara particle system, a motion input module, a spatial positioning module, a control processing module, a scene display module, and a motion data extraction module. The above virtual evacuation experiment platform and the virtual reality device together control the movement of specific personnel in UE5. Multiple virtual reality evacuation scenarios are realized by changing the environmental visibility, smoke layer height, interpersonal distance, and virtual personnel speed in the evacuation scene. The motion data extraction module provides a blueprint for data collection on the UE5 virtual reality experiment platform, which can extract the motion data of participants. It should be further noted that in the specific implementation process, the motion input module is a head-mounted display, which is responsible for inputting the motion information of the participant to the control processing module. The participant can map the movement in the real world to the virtual reality platform through the above device. The spatial positioning module performs spatial positioning based on 6DoF (six degrees of freedom) tracking. It involves six degrees of freedom, namely three translational degrees of freedom and three rotational degrees of freedom. Using external positioning sensors (such as laser trackers, motion capture systems, etc.), built-in sensors (such as inertial measurement units IMUs, cameras, etc.), and sensor fusion technologies, etc., it can achieve precise positioning and attitude tracking of devices or objects in three-dimensional space, providing a richer and more realistic interaction experience for various application scenarios. The control processing module is implemented by a computer, such as Figure 9As shown. The computer can obtain the position information of the head-mounted display through the motion input module, and at the same time efficiently transmit the continuously updated virtual scene information to the scene display module to ensure that the participants can see the synchronously changing experimental environment. The scene display module includes a head-mounted virtual reality device and a computer monitor that receive and display the scene pictures from the control processing module to achieve the immersion and realism of the virtual reality scene; Virtual Agent: A virtual person capable of running trigger events, performing set movements in the virtual reality scene to increase the authenticity of the scene; The virtual ship fire evacuation experimental platform is equipped with a blueprint system and a Niagara particle system: The blueprint system is a complete scripting system that can provide visual elements such as Events, Functions, Variables, and Nodes. By connecting these elements, complex scene logic and interactivity can be created more efficiently; The Niagara particle system is an important feature in the UE (Unreal Engine). It replaces the previous Cascade particle system and provides a more powerful, flexible, and superior-performance particle effect creation tool. The Niagara particle system allows developers to create and manipulate various particle effects, such as flames, smoke, water flows, explosions, rain, snow, etc. These effects are not only realistic but also highly customizable, with a variety of settings that can be adjusted to create unique effects. Generally speaking, the Niagara particle system, with its powerful functions, high flexibility, and superior performance, has become an important tool for development and visual effects production. It can help developers achieve various exquisite games and visual effects, bringing a more vivid experience. The motion data extraction module mainly updates and stores the VR device data from UE5 by combining the blueprint system and C++ programming. The collected data includes the spatio-temporal information of the participants; after receiving the collected data from the environmental information detection module, it processes the collected data. Specifically, based on the spatio-temporal data extracted from the participants, it analyzes the movement trajectory of the participants and calculates their movement speed to realize the research on the movement characteristics of evacuees under different environmental conditions.

[0051] The specific process of the embodiment of the present application is described below:

[0052] Scenes without the influence of fire smoke (such as in the appendix Figure 1 ): A virtual reality platform for ship corridor evacuation scenarios in case of fire is constructed in a three-dimensional open-world map. In this platform, the corridor size refers to the MSC.1 / Circ.1533 circular letter issued by the IMO. There is a virtual protagonist in the scene, and its movement can be controlled by the participant walking in the real world while wearing a head-mounted virtual reality device.

[0053] Scenes with limited visibility due to fire smoke (such as in the appendix Figure 2As shown in the figure: Based on the scenario without the influence of fire smoke, a virtual reality evacuation platform without obstacles in multiple visibility environments is constructed. During the virtual evacuation experiment, the participants can only see a certain range in the moving direction of the virtual protagonist controlled by themselves, and the visibility of the environment is set by UE5. Under the control of the participants, the position and direction of the virtual protagonist change, and the visible area also changes accordingly. At the same time, the Exponential Height Fog in the constructed virtual reality platform is placed in the built scene, and by setting the corresponding fog concentration and visibility range, the construction of multiple scenes with limited vision is realized.

[0054] Scenes with different smoke layer heights (such as attached Figure 3 ):Based on the scenario without the influence of fire smoke, a virtual reality evacuation experiment platform with multiple smoke layer heights is constructed, using the exact same ship corridor environment as in the scenario without the influence of fire smoke. Three postures, namely standing upright, bending down, and crawling, are selected, and the corresponding heights are used as the characteristic heights of the smoke layer. At the same time, the Local Height Fog in the constructed virtual reality platform is set, and through coordinate transformation and fog distribution, the corresponding height, fog height deviation, and radial attenuation are set to achieve the setting of the smoke layer height.

[0055] For the import of static obstacles: Mainly by clicking on the "Shapes" option in the Place ACTOR menu of UE5, select the shape of the obstacle, such as "Cylinder". After selecting the cylinder, the parameters of the cylinder, such as Width, Depth, Height, Radius, etc., can be adjusted in the parameter setting panel on the right. After adjusting the parameters, drag the cylinder to the set position in the scene (such as attached Figure 4 ).

[0056] Virtual agent module (such as attached Figure 5 ):Here, the movement and triggering of the virtual agent are mainly introduced. The moving speed of the virtual agent refers to the value of the moving speed of ship passengers in the MSC.1 / Circ.1533 circular letter issued by IMO. Based on the UE5 platform, the virtual agent module is set through blueprint script components. Specifically, create a skeleton and walking animation in the blueprint interface, and then add a character movement component and set it according to the specified speed magnitude and direction to achieve the movement of the virtual agent. Then, draw an Event Graph, which means that when the participant touches this trigger, the movement of the virtual agent will be activated, and the setting of the passenger countercurrent event can be completed. Attached Figure 6It represents triggering the blueprint construction process of an event. By setting the triggering area, virtual agents are searched for in the scene, thereby triggering the virtual agents to complete the specified event. Two variables, Initial Speed and Current Speed, are introduced in the speed setting of the virtual agent to store the speeds of the virtual agent before and after the change with the speed decay factor. The Set Variable node is used to update the value of the Current Speed variable. Here, we introduce a speed change considering the coupled influence of multiple factors to simulate the characteristics of the speed change of evacuating people in a fire environment. Five floating-point variables, "crowd density", "Tilt angle", "Walked Distance", "Visibility", and "coupling", are created in the blueprint of the virtual agent. First, a blueprint function is created that accepts a floating-point input (tilt angle) and returns a floating-point output (tilt speed output). Inside this function, mathematical operations are used to implement the speed rules at different ship tilt angles. The formula is as follows:

[0057] ν = 0.12(−0.15i + 5.6)

[0058] where ν is the speed of the virtual agent changing with the ship tilt angle; i is the tilt angle, °. It is used to calculate the speed of walking on a flat ground.

[0059] Considering the situation where there are stairs on the ship, the calculation formula on the stairs is as follows:

[0060] ν = 0.05(−0.15i + 5.6)

[0061] In the movement event, the movement direction and distance of the virtual agent are obtained and accumulated to "WalkedDistance". A blueprint function is created that accepts a floating-point input (walking distance) and returns a floating-point output (walking decay factor). Inside this function, mathematical operations are used to implement the speed decay rule. The formula is as follows:

[0062] R tired = 0.27x - 0.003

[0063] where R tired is the speed decay ratio of the virtual agent with the walking distance; x is the movement distance, km.

[0064] Secondly, a blueprint function is created that accepts a floating-point input (visibility) and returns a floating-point output (visibility decay factor). Inside this function, mathematical operations are used to implement the speed decay rule. The formula is as follows:

[0065] R visibility = 0.004V2 +0.014V + 0.68

[0066] where R visibility is the evacuation speed decay ratio; V is the visibility, in m.

[0067] Finally, create a blueprint function that takes the above two floating-point inputs and returns a floating-point output (the decayed speed). The calculation formula for the fire evacuation speed considering the coupled effects of ship heel angle, motion fatigue, and visibility is proposed in the way of multiplying multiple factors as follows:

[0068] V coupling = v(1 - R tired )(1 - R visibility )

[0069] where V coupling is the fire evacuation speed under the coupled influence of human fatigue and visibility, in m / s.

[0070] Call the speed decay function in Event Tick and use Coupling as the input. Assign the speed value returned by the function to "CurrentSpeed". Finally, use "CurrentSpeed" as the movement speed in the movement logic of the virtual agent.

[0071] By adding a blueprint script component to the moving virtual agent to construct a scenario where the interpersonal distance from the virtual agent is fixed, the setting of the ship passenger following behavior event can be completed. Attached Figure 7 shows the blueprint construction process for the participant to maintain a fixed interpersonal distance from the NPC. First, call Get VR Pawn to obtain the VR Pawn. Using the obtained VR Pawn reference, call its Get Actor Location to obtain its world coordinates. Similarly, the world coordinates of the virtual agent need to be obtained. This coordinate will be used to compare with the coordinates of the VR Pawn, and Kismet functions will be used on its coordinates to obtain the vector difference between them. The length (magnitude) of the difference vector represents the interpersonal distance between the VR Pawn and the virtual agent. Use Vector Length to calculate this length to enable the participant to move forward and backward from the NPC at the set interpersonal distance at the start of the experiment, thereby enabling the virtual agent to follow the participant according to the set interpersonal distance. At the same time, draw the Event Graph chart for the trigger event with the virtual protagonist to enable the participant to move forward and backward from the NPC at the set interpersonal distance at the start of the experiment, completing the design of the passenger following behavior event.

[0072] Based on this inventive concept, a computer program product for collecting virtual reality pedestrian movement data is also provided. When this computer program product runs, it causes the computer to execute the above-mentioned Figure 8 method. The following mainly introduces this computer product, and the main steps are as follows:

[0073] (1) Edit the blueprint class asset VR Pawn in UE5 to output the spatial position data of the virtual protagonist at the set time interval.

[0074] (2) Create a C++ project in Microsoft Visual Studio. This C++ code is an implementation of a class in UE, covering header files, constructors, BeginPlay functions, Tick functions, and LogLocationToCSV functions, which are used for debugging, analyzing player behavior, and creating an in-game logging system to record the spatio-temporal information of virtual reality participants.

[0075] (3) Create a C++ project in Microsoft Visual Studio. This section of C++ code is a class declaration in UE. It formats the spatio-temporal information of virtual reality participants into CSV format and writes it to a specified file by overriding the BeginPlay and Tick functions of the Actor class and using the LogLocationToCSV function.

[0076] Based on the spatio-temporal distribution data of participants in the virtual motion scenario, combined with the camera to obtain the walking postures and speeds of the experimenters under different visibility and smoke layer heights. As shown in the appendix Figure 10 、 Figure 11 、 Figure 12 .

[0077] At the same time, on the basis of fully considering key factors such as CO, CO2, and temperature distribution, a mathematical model for minimizing evacuation injuries is constructed. Pyrosim is used to simulate these factors to capture the spatial distribution of smoke and heat. Based on the simulation data, the constructed ship fire injury model is used to optimize the evacuation.

[0078] First, the evacuation goal needs to be clarified, and then the corresponding evacuation mathematical model is established. In the context of ship fires, according to statistical analysis, trapped people are often injured by the toxic smoke generated by the fire due to their upright evacuation or bent-down evacuation. Such injuries include poisoning, burns, getting into danger due to difficulty in environmental identification, and wrong decisions and chaos caused by panic. Based on these characteristics, the upright and bent-down evacuation postures, smoke temperature, and regional residence time are selected as the key factors affecting evacuation injuries, and the harmful gases CO and CO2 are selected.

[0079] The matrix method is adopted to describe each evacuation area in the ship's cabin and the connection status among them.

[0080] (1) Evacuation area matrix E

[0081] Suppose there are N evacuation areas in the ship's cabin, that is, there are N nodes in the network.

[0082] E = [E1 E2…EN]

[0083] (2) Evacuation area connection status matrix Y

[0084] In the actual ship's cabin, each evacuation area has its own connection status with other evacuation matrices. Use y ij to represent the connection status between area i and area j. y ij = 1 indicates that area i and area j are connected, and y ij = 0 indicates that area i and area j are not connected. In addition, since the connection of area i with itself is meaningless, so y ii = 0.

[0085]

[0086] (3) Matrix X of the usage status of the edges in the network model in the final path

[0087] Use x ij to represent the usage status of the edge in the final path. x ij = 1 indicates that the final evacuation path contains the edge v ij ; x ij = 0 indicates that the final evacuation path does not contain the edge v ij , or the edge v ij does not exist (area i and area j are not connected).

[0088] In order to clarify the evacuation goal and find the corresponding final evacuation path, we constructed a mathematical model aiming at minimizing the evacuation harm, and selected harmful gases CO and CO2, as well as the flue gas temperature and the regional residence time as the factors affecting the evacuation harm. The evacuation model can be expressed as the following formula:

[0089]

[0090] Among them: D is the evacuation harm. i is the i-th evacuation area, and N is the total number of evacuation areas in the ship's cabin fire environment. T ij The time attribute of the corresponding edge ν ij , that is: the time required for the participant to pass through or leave the evacuation area i and reach the next area j, that is, the residence time of area i. is the corresponding edge ν ijThe damage attributes, i.e., the damage values of fire smoke CO, CO2, and temperature suffered by evacuees when passing through evacuation area i in unit time due to standing upright and bending down. x ij is the usage situation of the edge in the network model. When ν ij = 1, it means that the edge ν is included in the final evacuation path ij , when ν ij = 0, it means that the edge ν is not included in the final evacuation path ij , or the edge ν ij does not exist.

[0091] The constraint conditions are as follows:

[0092] if x ij = 1

[0093] It means that the damage values of standing upright, bending down, and temperature in each evacuation area of the evacuation path of each evacuee should be lower than the lethal value.

[0094] x ij ≤ y ij

[0095] It means that the connectivity of each evacuation area in the evacuation path should be ensured: when y ij = 1, x ij = 0 or 1, when y ij = 0, x ij = 0.

[0096]

[0097] The flag ensures that the evacuation path will not form a loop.

[0098] The beneficial effects of the present invention are as follows:

[0099] 1. Based on the present invention, a virtual reality evacuation experiment platform can be constructed under various conditions such as the structure of typical ship areas, different visibility settings, and the influence of interference factors during the evacuation process.

[0100] 2. The present invention adopts relatively advanced computer technologies, software and hardware, including graphics rendering, physical engine, programming and scripting, network communication, virtual reality streaming / local area network technology, head-mounted virtual reality devices, etc., and can realize high-quality and high-fidelity evacuation scenarios.

[0101] 3. Compared with previous virtual reality experimental platforms, the evacuation speed calculation method of general virtual agents often only considers a single factor. In contrast, the virtual agent in the present invention uses a multi-factor coupled speed calculation model, comprehensively considering multiple factors such as the density of the evacuated crowd, ship tilt, motion fatigue, and visibility. These factors can all have a significant impact on the evacuation speed during a fire evacuation. Therefore, this virtual agent model can more comprehensively reflect the real situation of fire evacuation.

[0102] 4. Based on the obtained participant data, optimize the fire evacuation of personnel on ships in combination with Pyrosim. Existing research on personnel evacuation routes generally uses evacuation time, distance, and crowding degree as evacuation goals, but these goals do not fully consider the actual needs of the evacuated group. Therefore, it is proposed to use evacuation injury as the evacuation goal, aiming to reduce the actual injury degree of trapped personnel during the evacuation process.

[0103] 1. Based on the present invention, a virtual reality evacuation experimental platform can be constructed under various conditions such as the structure of typical ship areas, different visibility settings, and the influence of interference factors during the evacuation process.

[0104] 2. The present invention adopts relatively advanced computer technologies, software, and hardware, including graphics rendering, physics engines, programming and scripting, network communication, virtual reality streaming / LAN technology, head-mounted virtual reality devices, etc., to achieve high-quality and highly realistic evacuation scenarios.

[0105] 3. Compared with previous virtual reality experimental platforms, the evacuation speed calculation method of general virtual agents often only considers a single factor. In contrast, the virtual agent in the present invention uses a multi-factor coupled speed calculation model, comprehensively considering multiple factors such as the density of the evacuated crowd, ship tilt, motion fatigue, and visibility. These factors can all have a significant impact on the evacuation speed during a fire evacuation. Therefore, this virtual agent model can more comprehensively reflect the real situation of fire evacuation.

[0106] 4. Based on the obtained participant data, optimize the fire evacuation of personnel on ships in combination with Pyrosim. Existing research on personnel evacuation routes generally uses evacuation time, distance, and crowding degree as evacuation goals, but these goals do not fully consider the actual needs of the evacuated group. Therefore, it is proposed to use evacuation injury as the evacuation goal, aiming to reduce the actual injury degree of trapped personnel during the evacuation process.

[0107] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art in this technical field based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art shall fall within the protection scope determined by the claims.

Claims

1. A ship fire evacuation simulation method based on UE5, characterized in that: The method comprises the following steps: S1. Build a mobile virtual agent in a virtual ship cabin scene; S2, setting the evacuation area matrix and the evacuation area connection matrix of the virtual ship cabin; S3, constructing a mobile virtual agent fire evacuation model based on the evacuation area matrix and the evacuation area connection matrix; S4. Solve the evacuation model to obtain the evacuation path, and control the mobile virtual agent to perform evacuation simulation according to the evacuation path.

2. A ship fire evacuation simulation method based on UE5 according to claim 1, characterized in that: The evacuation area matrix represents a vector composed of various evacuation areas of the virtual ship cabin, and each evacuation area corresponds to a node.

3. A ship fire evacuation simulation method based on UE5 according to claim 2, characterized in that: The element in the ith row and jth column of the evacuation area connection status matrix represents the connection status between evacuation area i and evacuation area j.

4. A ship fire evacuation simulation method based on UE5 according to claim 3, characterized in that: The evacuation model takes minimizing evacuation damage as its objective function.

5. A ship fire evacuation simulation method based on UE5 according to claim 4, characterized in that: The objective function is: Where D is the evacuation damage, i is the i-th evacuation area, N is the total number of evacuation areas in the ship cabin fire environment, T ij The corresponding edge ν ij The time attribute, is the corresponding edge ν ij Damage attribute, x ij For edge ij usage.

6. A ship fire evacuation simulation method based on UE5 according to claim 5, characterized in that: The corresponding edge ν ij The time attribute of represents the time required for the mobile virtual agent to pass through or leave area i to reach the next area j, that is, the residence time in area i.

7. A ship fire evacuation simulation method based on UE5 according to claim 5, characterized in that: Corresponding edge ν ij The damage attribute represents the damage value of fire smoke suffered by the mobile virtual agent due to standing upright, bending over and temperature when passing through the evacuation area i.

8. The ship fire evacuation simulation method based on UE5 according to claim 5, characterized in that: Use x ij =1, indicating that the evacuation path includes edge v ij , usage x ij = 0, indicating that the evacuation path does not include edge v ij .

9. A ship fire evacuation simulation method based on UE5 according to claim 5, characterized in that: The constraint conditions of the evacuation model include threshold constraints, connectivity constraints and non-annular constraints.

10. A ship fire evacuation simulation method based on UE5 according to claim 9, characterized in that: The threshold constraints are: in, Indicates the lethal values ​​of CO, CO2 and temperature at the upright and bent heights in each evacuation area in the evacuation route of each evacuee; The connectivity constraints are: x ij ≤y ij Among them, y ij represents the connection between evacuation area i and evacuation area j, y ij =1 means that evacuation area i and evacuation area j are connected, y ij =0 means that evacuation area i and evacuation area j are not connected; The non-circular constraints are: Among them, x inter,j and x j,inter Represents the intermediate node of the final evacuation path.

Images