Virtual reality evacuation experiment method, system and device based on multi-person online cooperation

Abstract

The invention relates to the technical field of virtual reality, and discloses a virtual reality evacuation experiment method, system and device based on multi-person online coordination. The method comprises the following steps: building a virtual reality disaster scene based on a scene template in a preset scene model library; after the virtual reality disaster scene is triggered to run, acquiring evacuation experiment data generated by the tested object based on the virtual reality disaster scene displayed by the tested device; through each intelligent agent model-based NPC in the virtual reality disaster scene, based on evacuation experiment data and scene data of the virtual reality disaster scene, determining a motion and control strategy of the NPC so as to display each NPC in the virtual reality disaster scene displayed by the tested equipment according to the motion and control strategy, and interacting with the tested object to perform an evacuation experiment; the problems of low efficiency, poor effect, high cost and uncontrollable risk caused by performing an evacuation experiment in a real scene can be avoided; the experiment efficiency and effect are improved, and the experiment cost and risk are reduced.

Description

Technical Field

[0001] The present disclosure relates to the field of virtual reality technology, and in particular to a virtual reality evacuation experiment method, system and device based on multi-person online collaboration. Background Art

[0002] In order to improve the public's ability and efficiency in responding to disaster events, it is generally necessary to organize personnel to conduct disaster evacuation experiments in some locations, such as earthquake disaster evacuation experiments, fire evacuation experiments, etc.

[0003] Traditional disaster evacuation experimental methods include: pre-planning the assembly location and evacuation route at a designated location; simulating the occurrence of a disaster at the designated location through alarms, etc., and each subject takes shelter and evacuates according to the pre-planned evacuation route.

[0004] However, the disaster situations simulated by traditional experimental methods in real scenarios are not realistic enough, and the experimental process requires a lot of manpower and material resources, resulting in poor results and efficiency of evacuation experiments. Summary of the invention

[0005] In view of this, the present disclosure proposes a virtual reality evacuation experiment method, system and device based on multi-person online collaboration, which can avoid the problems of low efficiency, poor effect, high cost and uncontrollable risk caused by conducting evacuation experiments in real scenarios; because the disaster scene constructed by the virtual reality disaster scene is more realistic and there is no need to configure experiment-related facilities in the real scene, it can improve the efficiency and effect of the evacuation experiment and reduce the experimental cost and risk.

[0006] According to one aspect of the present disclosure, a virtual reality evacuation experiment method based on multi-person online collaboration is provided, the method comprising:

[0007] Based on the scene templates in the preset scene model library, a virtual reality disaster scene is built to simulate the disaster process in the real disaster scene through the test equipment used by the test subjects;

[0008] After triggering the virtual reality disaster scene to run, obtaining evacuation experimental data generated by the test subject based on the virtual reality disaster scene displayed by the test device;

[0009] Through each non-player character NPC based on the intelligent agent model in the virtual reality disaster scene, the movement and control strategy of the NPC is determined based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so that each NPC is displayed according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interacts with the test subject to perform an evacuation experiment; wherein different NPCs correspond to different intelligent agent models.

[0010] In a possible implementation, determining, by each NPC based on the agent model in the virtual reality disaster scene, the movement and control strategy of the NPC based on the evacuation experiment data and the scene data of the virtual reality disaster scene, includes:

[0011] For each NPC based on the agent model, obtaining setting parameters of the NPC; wherein the setting parameters include an initial movement position and an end movement position of the NPC;

[0012] By means of the NPC, based on the evacuation experiment data and the obstacle information and the topological map indicated by the scene data, planning an initialization movement path of the NPC from the movement initial position to the movement end position in the virtual reality disaster scene;

[0013] The NPC moves according to the initialized motion path, and the behavior of the NPC is determined based on preset heuristic rules during the movement, so as to simulate the behavior of a real person when moving in the actual disaster scene.

[0014] In a possible implementation, the decision of the behavior of the NPC based on a preset heuristic rule during the movement process includes:

[0015] Acquire the physical information, social information and environmental information indicated by the evacuation experiment data and the scenario data;

[0016] Behaviors indicated by the heuristic rules corresponding to the physical information, the social information, and the environmental information are determined.

[0017] In a possible implementation, the setting parameters further include the NPC's generation position, disappearance position and / or return position; accordingly, the method further includes:

[0018] Controlling the generation of the NPC at a generation location of the virtual reality disaster scene to trigger the step of determining the movement and control strategy of the NPC based on the evacuation experiment data and the scene data of the virtual reality disaster scene through each agent-based NPC in the virtual reality disaster scene;

[0019] and / or,

[0020] Controlling the NPC to disappear at the disappearance position of the virtual reality disaster scene;

[0021] and / or,

[0022] At the return position of the virtual reality disaster scene, the NPC is controlled to return to the initial position of movement to trigger the execution of the step of planning, by the NPC, an initialization movement path of the NPC from the initial position of movement to the end position of movement in the virtual reality disaster scene based on the obstacle information and topological map indicated by the scene data, and subsequent steps.

[0023] In a possible implementation, the virtual reality disaster scene is constructed based on a scene template in a preset scene model library, including:

[0024] Calling a three-dimensional engine to load a pre-built scene map of the virtual reality disaster scene;

[0025] Obtaining a selection instruction for a scene template in the scene model library;

[0026] The scene template indicated by the selection instruction is loaded into the scene map by the three-dimensional engine, so as to generate the virtual reality disaster scene by combining the scene template indicated by the selection instruction and the scene map;

[0027] Wherein, the scene map and the scene template are both in preset formats supported by the three-dimensional engine for loading.

[0028] In a possible implementation, after triggering the virtual reality disaster scene to run, the method further includes:

[0029] Calling simulation software to perform disaster dynamics simulation on the virtual reality disaster scene;

[0030] The simulation results of the simulation software are loaded to be displayed in the virtual reality disaster scene of the device under test.

[0031] In a possible implementation, the virtual reality disaster scene corresponds to at least two test subjects, and the method further includes:

[0032] For each test subject, the evacuation experiment data of the test subject is sent in a preset data format through a preset data interface to other test devices used by other test subjects different from the test subject, so that the other test devices can display the test subject virtual image corresponding to the test subject in the virtual reality disaster scene based on the evacuation experiment data.

[0033] According to another aspect of the present disclosure, a virtual reality evacuation experiment system based on multi-person online collaboration is provided, the system comprising: a test device, and a server communicatively connected to the test device;

[0034] The server is used to build a virtual reality disaster scene based on a scene template in a preset scene model library;

[0035] The test device is used to display the virtual reality disaster scene to simulate the disaster process in the real disaster scene; after triggering the operation of the virtual reality disaster scene, the evacuation experimental data generated by the test object based on the virtual reality disaster scene displayed by the test device is collected;

[0036] The server is further used to obtain the evacuation experiment data after triggering the operation of the virtual reality disaster scene; determine the movement and control strategy of the NPC based on the intelligent agent model in the virtual reality disaster scene based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so as to display each NPC according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interact with the test object; wherein different NPCs correspond to different intelligent agent models.

[0037] According to another aspect of the present disclosure, a virtual reality evacuation experimental device based on multi-person online collaboration is provided, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0038] According to another aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions implement the above method when executed by a processor.

[0039] According to another aspect of the present disclosure, a computer program product is provided, including a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0040] By building a virtual reality disaster scene based on a scene template in a preset scene model library, the disaster process in the real disaster scene is simulated through the test equipment used by the test subject; after the virtual reality disaster scene is triggered to run, the evacuation experiment data generated by the test subject based on the virtual reality disaster scene displayed by the test equipment is obtained; through each NPC based on the intelligent body model in the virtual reality disaster scene, based on the evacuation experiment data and the scene data of the virtual reality disaster scene, the movement and control strategy of the NPC is determined, so that each NPC is displayed according to the movement and control strategy in the virtual reality disaster scene displayed by the test equipment, and the evacuation experiment is carried out by interacting with the test subject; the problems of low efficiency, poor effect, high cost and uncontrollable risk caused by conducting evacuation experiments in real scenes can be avoided; since the disaster scene constructed by the virtual reality disaster scene is more realistic and there is no need to configure experimental related facilities in the real scene, the efficiency and effect of the evacuation experiment can be improved and the experimental cost can be reduced; at the same time, since the test subject does not need to conduct an evacuation experiment in a real disaster scene, the risk of accidents to the test subject can be reduced.

[0041] In addition, when building a virtual reality disaster scene, there is no need to draw the virtual reality scene through a scene management device. Instead, the existing scene template can be used to quickly build the scene, which can improve the efficiency of building the virtual reality disaster scene.

[0042] In addition, by setting an intelligent agent model for each NPC, different NPCs can generate personalized movement and control strategies, which can improve the flexibility of NPC movement and control.

[0043] In addition, multiple test devices can follow the same rules in data interaction with virtual reality disaster scenes and user operation feedback. On the one hand, it can realize the synchronous collection of evacuation experiment data of different test subjects, thereby realizing multi-person collaborative experiments; on the other hand, it can enable test subjects to use different types of equipment to collaboratively conduct evacuation experiments with data compatibility and scalability, thereby improving the versatility of the evacuation experiment method.

[0044] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0046] Figure 1 A structural schematic diagram of a virtual reality evacuation experiment system based on multi-person online collaboration according to an embodiment of the present disclosure is shown;

[0047] Figure 2 A flowchart of a virtual reality evacuation experiment method based on multi-person online collaboration according to an embodiment of the present disclosure is shown;

[0048] Figure 3 A block diagram of a virtual reality evacuation experimental device based on multi-person online collaboration according to an embodiment of the present disclosure is shown;

[0049] Figure 4 A block diagram of a virtual reality evacuation experimental device based on multi-person online collaboration according to another embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0050] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0051] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0052] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0053] In view of the current problems of difficulty, low efficiency and poor effect in conducting group evacuation experiments, this application provides a virtual reality evacuation experiment system based on multi-person online collaboration, which allows multiple subjects to participate online at the same time and interact with non-player characters (NPCs), and can simulate real evacuation scenes through an immersive virtual reality (VR) environment. The evacuation experiment system is introduced below.

[0054] Figure 1 FIG. 1 is a schematic diagram showing a structure of a virtual reality evacuation experiment system based on multi-person online collaboration according to an embodiment of the present disclosure. Figure 1 As shown, the system includes: a test device 110, a scene management device 120, and a server 130 which is communicatively connected to the test device 110 and the scene management device 120 respectively.

[0055] The scene management device 120 is used by the scene manager to manage the virtual reality disaster scene, such as triggering the generation of the virtual reality disaster scene, setting NPC for the virtual reality disaster scene, etc. This embodiment does not limit the management content of the scene management device 120. The scene manager (also known as the organizer of the evacuation experiment, this embodiment does not limit the name of the scene manager) is used to set the task objectives of the virtual reality evacuation experiment, such as the type of disaster simulated, the number of people in the experiment, etc. Optionally, the scene management device 120 can be an electronic device with computing power and human-computer interaction capabilities, such as a VR device, a computer, a tablet computer, or a mobile phone. This embodiment does not limit the implementation method of the scene management device 120.

[0056] The test device 110 is used by the test subject to provide the test subject with a virtual reality disaster scene so that the test subject can interact with the virtual reality disaster scene to realize the evacuation experiment. Optionally, the test device 110 and the scene management device 120 can be the same device, that is, the scene manager and the test subject use the same device (that is, the scene management device 120 can be not configured at this time); or, the scene management device 120 and the test device 110 can also be different devices. In this case, the device types of the scene management device 120 and the test device 110 can be the same or different. The test device 110 can be a VR device, a computer, a tablet computer, or a mobile phone, etc., an electronic device with computing power and human-computer interaction capabilities. This embodiment does not limit the implementation method of the test device 110.

[0057] Optionally, there is one or at least two test subjects. When there are at least two test subjects, there are at least two test devices 110 , which correspond one-to-one to each test subject.

[0058] Optionally, since the subject interacts with the virtual reality disaster scene during the evacuation experiment, the test device 110 includes at least one data acquisition device, which can collect the evacuation experiment data of the subject, and the evacuation experiment data is used to indicate the subject's reaction to the current virtual reality disaster scene. Optionally, the data acquisition device includes but is not limited to at least one of the following:

[0059] 1. Posture sensor, used to collect the position and direction of the subject in the virtual reality disaster scene. Accordingly, the evacuation experiment data includes the posture data of the subject.

[0060] 2. Electromyographic equipment, used to collect electromyographic data of the subjects. Accordingly, the evacuation experiment data includes the electromyographic data of the subjects.

[0061] 3. EEG equipment, used to collect EEG data of the subjects. Accordingly, the evacuation experiment data includes the EEG data of the subjects.

[0062] 4. Motion capture equipment is used to collect the body movements of the subjects. Accordingly, the evacuation experiment data includes the body movement data of the subjects.

[0063] 5. Eye tracker, used to collect eye movement data of the subjects. Accordingly, the evacuation experiment data includes the eye movement data of the subjects.

[0064] 6. Near infrared equipment, used to collect near infrared data of the subjects, such as brain oxygenated hemoglobin and oxygen-free hemoglobin data. Accordingly, the evacuation experiment data includes the near infrared data of the subjects.

[0065] In actual implementation, the data acquisition device may also include other types, and accordingly, the evacuation experiment data may also include other types of data, which will not be listed one by one in this embodiment.

[0066] In this embodiment, the tested device and the scene management device further include a display device to display the experimental process of the virtual reality evacuation experiment.

[0067] The server 130 is used to provide virtual reality evacuation experiment services for the scene management device 120 and the test device 110. Specifically, the server 130, the scene management device 120 and the test device 110 run a virtual reality evacuation experiment platform (or a virtual reality evacuation experiment program). The virtual reality evacuation experiment platform is used to provide complete virtual reality evacuation experiment services. The platform includes a background service module running in the server, and a front-end interaction module running on the front end (including the test device 110 and the scene management device 120). The front-end interaction module communicates with the background service module to exchange data to provide virtual reality evacuation experiment services for managers and test subjects. Optionally, the server 130 is a server cluster, such as: Figure 1 As shown, the tested device 110 and the scene management device 120 correspond to a server respectively, and the server architecture adopts a master-slave architecture, the server that is connected to the pre-defined master tested device among the multiple tested devices 110 is the master server, and the server that is connected to the other devices (including other tested devices and scene management devices) other than the master tested device is the slave server. In other embodiments, the server architecture of the server 130 can also be other types, and this embodiment does not limit the implementation method of the server 130.

[0068] In this embodiment, the server 130 is used to: build a virtual reality disaster scene based on a scene template in a preset scene model library, so as to simulate the disaster process in the real disaster scene through the test device used by the subject; after triggering the operation of the virtual reality disaster scene, obtain the evacuation experiment data generated by the subject based on the virtual reality disaster scene displayed by the test device; through each NPC based on the intelligent agent model in the virtual reality disaster scene, determine the movement and control strategy of the NPC based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so as to display each NPC according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interact with the subject to conduct an evacuation experiment; wherein different NPCs correspond to different intelligent agent models.

[0069] In this embodiment, the scene data is used to indicate the scene information within the preset scene range corresponding to the NPC, which is determined based on the NPC position in the virtual reality disaster scene. Among them, the preset scene range determined based on the NPC position includes: a rectangular area range of a first size determined with the NPC position as the center; or a circular area range of a second size determined with the NPC position as the center. The first size and the second size are pre-stored in the server, and this embodiment does not limit the shape and size of the preset area. The scene information includes virtual object information (such as virtual rooms, virtual roads, etc.) in the virtual reality disaster scene except for the virtual image of the subject corresponding to the subject, as well as weather information and other information that affects the NPC action.

[0070] In this embodiment, by building a virtual reality disaster scene for the test equipment to display for evacuation experiments, the problems of low efficiency, poor effect, high cost and uncontrollable risk caused by conducting evacuation experiments in real scenes can be avoided; since the disaster scene constructed by the virtual reality disaster scene is more realistic and there is no need to configure experimental related facilities in the real scene, the efficiency and effect of the evacuation experiment can be improved and the experimental cost can be reduced; at the same time, since the subjects do not need to conduct evacuation experiments in real disaster scenes, the risk of accidents to the subjects can be reduced.

[0071] In addition, when building a virtual reality disaster scene, there is no need to draw the virtual reality scene through the scene management device 120, but the existing scene template can be used to achieve rapid scene building, which can improve the efficiency of building the virtual reality disaster scene.

[0072] In addition, by setting an intelligent agent model for each NPC, different NPCs can generate personalized movement and control strategies, which can improve the flexibility of NPC movement and control.

[0073] The following is a detailed description of the virtual reality evacuation experimental method based on multi-person online collaboration provided in this embodiment. Figure 1 The server 130 in the system shown is taken as an example for explanation.

[0074] Figure 2 FIG. 1 is a flowchart of a virtual reality evacuation experiment method based on multi-person online collaboration according to an embodiment of the present disclosure. Figure 2 As shown, the method includes:

[0075] Step 201, based on the scene template in the preset scene model library, a virtual reality disaster scene is constructed to simulate the disaster process in the real disaster scene through the test equipment used by the test subject.

[0076] Among them, the scene model library includes at least one scene template, each scene template is used to simulate a disaster process of a real disaster scene, such as: simulating the process of fire, specifically simulating the change process of the scene (including smoke and fire) from the start of the fire to the extinguishing of the fire in a certain place; for example: simulating the process of earthquake, specifically simulating the various changes of the scene from the start of the earthquake in a certain place to the end of the earthquake (for example, including damage to buildings, cracks in the road surface, etc.). This embodiment does not limit the type of disaster and the implementation method of the scene template corresponding to each disaster.

[0077] Optionally, the scene model library includes scene templates that come with the virtual reality evacuation experiment platform; and / or scene templates edited and uploaded to the server by the scene management device and other devices. This embodiment does not limit the method for obtaining scene templates in the scene model library.

[0078] The scene model library is pre-built in the virtual reality evacuation experiment platform. By running the virtual reality evacuation experiment platform, at least one scene template in the scene model library can be obtained. In this way, when building a virtual reality disaster scene, there is no need to draw the virtual reality scene through the scene management device. Instead, the existing scene template can be used to achieve rapid scene building, which can improve the efficiency of building the virtual reality disaster scene.

[0079] In one example, a virtual reality disaster scene is constructed based on a scene template in a preset scene model library, including: calling a three-dimensional engine to load a scene map of a pre-built virtual reality disaster scene; obtaining a selection instruction for a scene template in the scene model library; and loading the scene template indicated by the selection instruction into the scene map through the three-dimensional engine to generate a virtual reality disaster scene by combining the scene template indicated by the selection instruction and the scene map.

[0080] The 3D engine is used to create and render virtual reality content in real time. Optionally, the 3D engine may be Unreal Engine, specifically Unreal Engine 5. In other embodiments, the 3D engine may also be implemented in other ways, and this embodiment does not limit the implementation of the 3D engine.

[0081] In this embodiment, the virtual reality evacuation experiment platform is established based on a three-dimensional engine. When a virtual reality disaster scene needs to be built, the server calls the three-dimensional engine to load the scene template and the scene map, and adapts the scene template and the scene map to generate a virtual reality disaster scene. Among them, the scene map and the scene template are both preset formats supported by the three-dimensional engine, for example: the preset format is FBX format, which is a universal three-dimensional file format. Files in FBX format can interact with content between different three-dimensional modeling and animation software. Based on this, by setting both the scene map and the scene template to FBX format, it is possible to load and adapt the scene maps and scene templates created based on different three-dimensional engines in the virtual reality evacuation experiment platform, thereby improving the versatility and compatibility of the virtual reality evacuation experiment platform.

[0082] The scene map is used to indicate the geometric layout and / or style of the virtual reality disaster scene expected by the manager, so that the scene template is loaded into the geometric layout and / or style expected by the manager. Optionally, the scene map is constructed by the scene management device. Specifically, the scene management device constructs the scene map, including: calling the three-dimensional engine to display the map editing interface; receiving the map drawing operation acting on the map editing interface to obtain the scene map. Afterwards, the scene management device sends the scene map to the server.

[0083] In other embodiments, the server may also provide a map model library, which includes templates of multiple scene maps; accordingly, the scene map may also be a template of the scene map selected by the scene management device from the map model library. This embodiment does not limit the method of constructing the scene map.

[0084] The selection instruction is used to indicate a scene template in the scene model library. Optionally, the selection instruction is generated by the scene management device based on the received selection operation. Specifically, the scene management device displays each scene template in the scene model library through the scene building page; when a selection operation acting on any scene template is received, a selection instruction is generated and sent to the server. Optionally, the selection operation can be an operation of clicking on a scene template, or an operation of selecting a scene template by voice instruction, etc. This embodiment does not limit the implementation method of the selection operation.

[0085] After the server completes building the virtual reality disaster scene, the virtual reality disaster scene is sent to the test device for display, so as to simulate the disaster process in the real disaster scene through the test device used by the test subject.

[0086] Step 202, after triggering the virtual reality disaster scene to run, obtaining evacuation experimental data generated by the test subject based on the virtual reality disaster scene displayed by the test device.

[0087] Optionally, the device under test triggers the operation of the virtual reality disaster scene; or, the scene management device triggers the operation of the virtual reality disaster scene; the triggering method includes but is not limited to: physical triggering (i.e., triggering through physical buttons on the device), scene triggering (i.e., triggering operations acting on the virtual reality disaster scene), or voice triggering (i.e., triggering based on received voice instructions), etc. This embodiment does not limit the triggering method for starting the operation of the virtual reality disaster scene. When the device under test or the scene management device triggers the operation of the virtual reality disaster scene, a trigger instruction is sent to the server. Accordingly, the server determines that the virtual reality disaster scene has been triggered to run based on the trigger instruction, so as to execute the step of obtaining the evacuation experimental data generated by the subject based on the virtual reality disaster scene displayed by the device under test.

[0088] Triggering the operation of the virtual reality disaster scene means that the virtual reality disaster scene begins to perform plot deduction. At this time, the virtual reality evacuation experiment platform renders the scene plot of the virtual reality disaster scene in real time according to the scene template to display the real-time deduced disaster plot in the tested device.

[0089] In one example, during the disaster scenario simulation process, that is, after triggering the virtual reality disaster scene to run, the server is also used to: call simulation software to perform disaster dynamics simulation on the virtual reality disaster scene; load the simulation results of the simulation software to display in the virtual reality disaster scene.

[0090] The simulation software is used to simulate the development trend and results of disasters based on the changes of various factors in the virtual reality disaster scene. For example, based on the measures taken by the subjects (i.e. evacuation experimental data), the impact of the measures on the disaster in the virtual reality disaster scene is simulated to obtain the simulation results.

[0091] The simulation software includes, but is not limited to, a mathematical model based on a Bayesian network or a mathematical model based on an artificial intelligence network, etc. This embodiment does not limit the implementation of the simulation software. Schematically, the simulation software is a third-party software integrated in the virtual reality evacuation experiment platform. During the operation of the virtual reality evacuation experiment platform, the simulation software is called to perform a disaster dynamics simulation in combination with the current disaster-related data of the virtual reality disaster scene to obtain a simulation result. Afterwards, the simulation result is loaded and displayed through the test device.

[0092] Among them, disaster-related data refers to data that affects the development of disasters. The disaster-related data includes but is not limited to at least one of the following: weather data of virtual reality disaster scenes, evacuation experiment data performed by subjects, action data performed by NPCs, disaster duration, disaster location, etc. This embodiment does not limit the content of disaster-related data.

[0093] The evacuation experiment data is collected by the test device and uploaded to the server, so that the server can update the scene content of the virtual reality disaster scene currently displayed by the test device in real time based on the evacuation experiment data. For example: when the evacuation experiment data indicates that the test subject rotates to the left, the scene content of the virtual reality disaster scene currently displayed by the test device is updated to: the scene content covered by the line of sight after rotating to the left at a certain angle indicated by the evacuation experiment data; for another example: when the evacuation experiment data indicates that the test subject is moving forward, the scene content of the virtual reality disaster scene currently displayed by the test device is updated to: the scene content covered by the line of sight after moving forward a certain distance indicated by the evacuation experiment data.

[0094] Optionally, the test device may also display a test subject's corresponding virtual image in the virtual reality disaster scene. The test subject's virtual image may be generated by the server based on the test subject's corresponding height, image and other object information, or the test subject's virtual image may be a preset image generated by the server. This embodiment does not limit the generation method of the test subject's corresponding virtual image. In other embodiments, the test device may not display the test subject's corresponding virtual image in the virtual reality disaster scene, but only display the scene content of the virtual reality disaster scene visible from the test subject's perspective.

[0095] Step 203, through each NPC based on the intelligent agent model in the virtual reality disaster scene, based on the evacuation experiment data and the scene data of the virtual reality disaster scene, determine the movement and control strategy of the NPC, so as to display each NPC according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interact with the subject to conduct an evacuation experiment; wherein different NPCs correspond to different intelligent agent models.

[0096] The server pre-configures NPCs for virtual reality disaster scenarios. Optionally, NPCs configured for different virtual disaster scenarios may be preset by the server, or may be configured by a scenario management device or a test device.

[0097] If the NPC is configured by the scene management device or the tested device, the configuration method of the NPC in the virtual reality disaster scene includes but is not limited to: after the virtual reality disaster scene is built or the virtual reality disaster scene is run, the scene management device or the tested device receives the NPC setting operation and generates the NPC in the built virtual reality disaster scene. The NPC setting operation is used to set the number of NPCs, the type of NPCs, the initial movement position and / or the end movement position of the NPC, etc.

[0098] The method for setting the initial position of the NPC movement includes: receiving a selection operation acting on any geographical location in the virtual reality disaster scene, and determining the geographical location indicated by the selection operation as the initial position of the movement. For example, after the scene management device sets up the virtual reality disaster scene, the scene content of the virtual reality disaster scene displayed by the manager's current perspective is displayed through the action performed by the manager (the collection method is the same as the collection method of the evacuation experiment data, which will not be repeated in this embodiment); when receiving a selection operation acting on a geographical location in the scene content, the geographical location acted by the selection operation is determined as the initial position of the movement.

[0099] The method for setting the end position of the NPC's movement includes: after receiving a selection operation of the initial position of the movement, if a selection operation acting on another geographical location in the virtual reality disaster scene is received again, the selected operation received again is determined as the end position of the movement.

[0100] The method for setting the type of the NPC includes: displaying an NPC setting interface, the device interface including an NPC type selection control; and setting the type of the NPC based on a selection operation acting on the type selection control.

[0101] Optionally, the types of NPCs include, but are not limited to: strangers (such as passers-by), familiar people (such as friends, family members, neighbors, etc.), rescuers, etc. This embodiment does not limit the types of NPCs.

[0102] Optionally, the NPC setting interface can pop up in the form of a window after the initial position and the end position of the movement are selected; or, the display interface of the virtual reality disaster scene displays an NPC setting control, and the NPC setting interface is displayed when a trigger operation acting on the NPC setting control is received. This embodiment does not limit the display method of the device interface.

[0103] The method of setting the number of NPCs includes: displaying an NPC setting interface, the setting interface including a number setting control; setting the number of NPCs based on a number setting operation acting on the number setting control; or determining the number of NPCs based on a set number of a movement initial position or a movement end position.

[0104] After the scene management device or the test device completes the configuration of the NPC, the setting parameters of the NPC are sent to the server so that the server can configure the corresponding intelligent agent model for each NPC, and generate the NPC through the intelligent agent model, so that the NPC can determine the movement and control strategy of the NPC according to the setting parameters. In this embodiment, each NPC corresponds to an intelligent agent model, and different NPCs correspond to different intelligent agent models. The NPC refers to a virtual image generated by running the intelligent agent model in a virtual reality disaster scene. The operation of the intelligent agent model can realize the control of the NPC, that is, the behavior and action, disappearance and generation of the NPC in the virtual reality disaster scene can be determined by the calculation of the intelligent agent model. The intelligent agent model refers to a mathematical model used to generate the movement and control strategy of the NPC. The mathematical model includes but is not limited to a neural network model, etc. This embodiment does not limit the mathematical model for realizing the intelligent agent model.

[0105] In this embodiment, the scene data is used to indicate the scene information within the preset scene range corresponding to the NPC, which is determined based on the NPC position in the virtual reality disaster scene. Among them, the preset scene range determined based on the NPC position includes: a rectangular area range of a first size determined with the NPC position as the center; or a circular area range of a second size determined with the NPC position as the center. The first size and the second size are pre-stored in the server, and this embodiment does not limit the shape and size of the preset area. The scene information includes virtual object information (such as virtual rooms, virtual roads, etc.) in the virtual reality disaster scene except for the virtual image of the subject corresponding to the subject, as well as weather information and other information that affects the NPC action. Through each NPC based on the intelligent agent model in the virtual reality disaster scene, based on the evacuation experiment data and the scene data of the virtual reality disaster scene, the movement and control strategy of the NPC is determined, including: for each NPC based on the intelligent agent model, the setting parameters of the NPC are obtained; wherein the setting parameters include the initial movement position and the end movement position of the NPC; through the NPC, based on the obstacle information and topological map indicated by the evacuation experiment data and the scene data, the initialization movement path of the NPC from the initial movement position to the end movement position in the virtual reality disaster scene is planned; through the NPC, according to the initialization movement path, the NPC is moved, and the behavior of the NPC is decided based on the preset heuristic rules during the movement, so as to simulate the behavior of real people when moving in the real disaster scene.

[0106] In one example, determining a topological map based on scene data includes: calling a physical navigation system at the bottom of a three-dimensional engine to generate a navigation grid based on the scene data, the navigation grid being used to indicate a walkable area of ​​a virtual reality disaster scene; determining key navigation points in the walkable area, and connecting the key navigation points to form a topological map. The key navigation points may be set by a scene management device, or may be automatically identified by a three-dimensional engine, and the key navigation points include but are not limited to: intersections, room entrances, tops and bottoms of stairs, etc.

[0107] In one example, the NPC planning initialization motion path based on the agent model includes: based on the reciprocal velocity obstacle model (RVO2) algorithm, using the obstacle information indicated by the evacuation experiment data and the scene data to perform local path search and local motion behavior planning to obtain the initialization motion path. The initialization motion path allows the NPC to move without colliding with obstacles. Optionally, the obstacle information includes but is not limited to: obstacle information of static obstacles in the virtual reality disaster scene, such as the position, size, and shape of the static obstacles, and the static obstacles include but are not limited to: obstacles with fixed positions such as walls, tables, and cabinets; and obstacle information of dynamic obstacles, such as the shape, size, movement speed, and movement direction of the static obstacles, and the dynamic obstacles include but are not limited to: other NPCs different from the current NPC, the subject's virtual image corresponding to the subject, vehicles in the virtual reality disaster scene, animals, etc., and this embodiment does not limit the implementation of static obstacles, dynamic obstacles, and obstacle information.

[0108] In other embodiments, the motion path planning algorithm may also be other algorithms, such as: D*Lite algorithm, etc. This embodiment does not limit the implementation method of the motion path planning algorithm.

[0109] In one example, the behavior of an NPC is determined based on preset heuristic rules during movement, including: obtaining physical information, social information, and environmental information indicated by evacuation experiment data and scene data; and determining the behavior indicated by the heuristic rules corresponding to the physical information, social information, and environmental information.

[0110] Physical information includes obstacle information for NPCs, social information includes information related to social rules, such as traffic light signals, and environmental information is used to indicate the environment around NPCs, such as the distance between NPCs and disaster areas, etc. This embodiment does not limit the information types of physical information, social information, and environmental information.

[0111] Heuristic rules are formulated based on the behaviors of real people under the conditions of corresponding physical information, social information and environmental information. For example, if the physical information indicates that there is an obstacle ahead, the heuristic rule is that the NPC will go around it; if the physical information indicates that the distance between the obstacle and the NPC is less than the preset minimum avoidance distance, the heuristic rule is to generate a collision action when the NPC collides with the obstacle; if the environmental information indicates that a disaster has been encountered, the heuristic rule is that the NPC will speed up the movement in the direction away from the disaster; if the social information indicates that the subject's virtual image corresponding to the subject, other people or other NPCs have been encountered, the heuristic rule is that the NPC will move according to social rules. Among them, the minimum avoidance distance refers to the minimum distance that the NPC can avoid the NPC.

[0112] Optionally, when the obstacle that collides with the NPC is the subject's virtual image, the server also generates a collision feedback instruction, and sends the collision feedback instruction to the subject device corresponding to the subject's virtual image, so that the subject device can output collision feedback based on the collision feedback instruction, and the collision feedback includes but is not limited to: physical feedback, such as: handle vibration; and / or, visual feedback, such as: the virtual reality disaster scene screen displayed by the subject device corresponding to the collided subject virtual image becomes black, etc. Correspondingly, the server can also generate the collision action of the collided subject virtual image, and synchronize the collision action to other subject devices in the virtual reality disaster scene, so as to display the collision action of the collided subject virtual image in other subject devices.

[0113] The agent model learns heuristic rules in advance and decides the NPC's behavior based on the heuristic rules.

[0114] Optionally, after deciding the behavior of the NPC according to the heuristic rules, the movement path of the NPC may be planned again, and the behavior of the NPC may be decided again based on the heuristic rules during the movement, so as to achieve real-time update of the movement path and behavior.

[0115] Optionally, in this embodiment, the generation, disappearance and / or repeated movement of the corresponding NPC in the virtual reality disaster scene can also be controlled by running the intelligent agent model. At this time, the setting parameters also include at least one of the following:

[0116] 1. The NPC generation location. That is, the geographical location where the NPC appears in the virtual reality disaster scene. At this time, the server controls the NPC generation at the generation location of the virtual reality disaster scene by running the agent model to trigger the execution of the steps of determining the movement and control strategy of the NPC based on the evacuation experiment data and the scene data of the virtual reality disaster scene through each NPC agent model based on the agent model in the virtual reality disaster scene. Optionally, the generation location is the same as the initial movement location.

[0117] 2. The disappearance position of the NPC. That is, the geographical location of the NPC is stopped from being displayed in the virtual reality disaster scene. At this time, the NPC is controlled to disappear at the disappearance position of the virtual reality disaster scene by running the intelligent agent model.

[0118] 3. NPC's return position. That is, if the NPC reaches a certain position in the virtual reality disaster scene, it returns to the initial position of the movement to achieve the repeated movement of the NPC. At this time, the server controls the NPC to return to the initial position of the movement at the return position of the virtual reality disaster scene by running the intelligent body model, so as to trigger the execution of the steps of planning the initialization movement path of the NPC from the initial position of the movement to the end position of the movement in the virtual reality disaster scene and the subsequent steps based on the obstacle information and topological map indicated by the scene data.

[0119] In this embodiment, a corresponding intelligent body model is set for each NPC for control, so that each NPC can automatically find and change the path according to the surrounding information. There is no need for manual repeated dragging and setting of paths for the NPC, which can improve the control efficiency and flexibility of the NPC.

[0120] Optionally, the virtual reality disaster scene includes at least two test subjects, in which case the evacuation experiment data is sent by at least two test devices. Accordingly, for each test subject, the evacuation experiment data of the test subject is sent to other test devices used by other test subjects different from the test subject through a preset data interface in a preset data format, so that the other test devices can display the test subject's corresponding test subject virtual image in the virtual reality disaster scene based on the evacuation experiment data.

[0121] For example: Figure 1 In the process, the main server obtains the eye movement data, movement position and movement direction collected by the main test device, and these data are transmitted to the main server in real time for visualization, that is, the server determines the position and posture of the subject virtual image corresponding to the main test device in the virtual reality disaster scene based on the evacuation experiment data, obtains the subject virtual image corresponding to the main test device after visualization, and synchronizes the visualized subject virtual image corresponding to the main test device with other slave test devices connected to the slave server communication. Therefore, the subject virtual image corresponding to the main test device can be displayed on the slave test device.

[0122] Similarly, the movement position and direction of the slave subject collected from the test device are transmitted in real time to the slave server connected to the test device for communication, visualized, and synchronized to the main server, so that the main test device connected to the main server can display the virtual image of the subject corresponding to the slave test device.

[0123] In addition, since the master test device and the slave test device may use different device types to run the virtual reality evacuation experimental platform, for example, the device types of multiple test devices include desktop VR (i.e., a virtual reality device that displays a virtual reality scene through a display screen), head-mounted VR (i.e., a virtual reality device that uses a head-mounted display to display a virtual reality scene) and CAVE system (i.e., a virtual reality device that includes projection screens on multiple walls and floor projection screens), etc., the formats of data collected by different devices may be different. In this embodiment, in order to ensure data compatibility, data collection uses a unified data format, and data transmission uses a unified data interface. At this time, different test devices can follow the same rules in data interaction with virtual reality disaster scenes, user operation feedback, etc., so that the test subjects can use different types of equipment to collaboratively conduct evacuation experiments, thereby improving the versatility of the evacuation test method.

[0124] In summary, the virtual reality evacuation experiment method based on multi-person online collaboration provided in this embodiment builds a virtual reality disaster scene based on a scene template in a preset scene model library, so as to simulate the disaster process in the real disaster scene through the test device used by the test subject; after the virtual reality disaster scene is triggered to run, the evacuation experiment data generated by the test subject based on the virtual reality disaster scene displayed by the test device is obtained; through each NPC based on the intelligent body model in the virtual reality disaster scene, based on the evacuation experiment data and the scene data of the virtual reality disaster scene, the movement and control strategy of the NPC is determined, so that each NPC is displayed according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and the evacuation experiment is performed by interacting with the test subject; the problems of low efficiency, poor effect, high cost and uncontrollable risk caused by conducting evacuation experiments in real scenes can be avoided; since the disaster scene constructed by the virtual reality disaster scene is more realistic and there is no need to configure experiment-related facilities in the real scene, the efficiency and effect of the evacuation experiment can be improved and the experimental cost can be reduced; at the same time, since the test subject does not need to conduct an evacuation experiment in a real disaster scene, the risk of accidents to the test subject can be reduced.

[0125] In addition, when building a virtual reality disaster scene, there is no need to draw the virtual reality scene through a scene management device. Instead, the existing scene template can be used to quickly build the scene, which can improve the efficiency of building the virtual reality disaster scene.

[0126] In addition, by setting an intelligent agent model for each NPC, different NPCs can generate personalized movement and control strategies, which can improve the flexibility of NPC movement and control.

[0127] In addition, multiple test devices can follow the same rules in data interaction with virtual reality disaster scenes and user operation feedback. On the one hand, it can realize the synchronous collection of evacuation experiment data of different test subjects, thereby realizing multi-person collaborative experiments; on the other hand, it can enable test subjects to use different types of equipment to collaboratively conduct evacuation experiments with data compatibility and scalability, thereby improving the versatility of the evacuation experiment method.

[0128] Optionally, based on the above embodiment, after step 203, the server may also store the evacuation experiment data of each subject obtained during the virtual reality evacuation experiment, and generate the movement trajectory of the subject based on the movement data of the subject indicated by the evacuation experiment data, so as to evaluate the evacuation experiment. For example: matching the movement trajectory with the expected trajectory indicated by the task goal; if the movement trajectory does not match the expected trajectory, it is determined that the evacuation experiment of the subject does not meet the requirements; if the movement trajectory matches the expected trajectory, it is determined that the evacuation experiment of the subject meets the requirements.

[0129] In this embodiment, by storing the evacuation experiment data of the subject and generating the movement trajectory of the subject, an evacuation experiment evaluation is performed. There is no need to manually monitor the evacuation experiment process, and the evacuation experiment results can also be obtained after the evacuation experiment is completed, which can improve the efficiency and flexibility of obtaining the experimental results.

[0130] Figure 3 A block diagram of a virtual reality evacuation experiment device based on multi-person online collaboration according to an embodiment of the present disclosure is shown. The device at least includes: a virtual reality scene building module 310, an experiment data acquisition module 320 and an NPC motion control module 330.

[0131] A virtual reality scene building module 310 is used to build a virtual reality disaster scene based on a scene template in a preset scene model library, so as to simulate a disaster process in a real disaster scene through a test device used by a test subject;

[0132] The experimental data acquisition module 320 is used to obtain the evacuation experimental data generated by the test subject based on the virtual reality disaster scene displayed by the test device after the virtual reality disaster scene is triggered to run;

[0133] The NPC motion control module 330 is used to determine the motion and control strategy of each NPC based on the intelligent agent model in the virtual reality disaster scene based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so as to display each NPC according to the motion and control strategy in the virtual reality disaster scene displayed by the test device, and interact with the test subject to perform an evacuation experiment; wherein different NPCs correspond to different intelligent agent models.

[0134] For relevant details, please refer to the above method embodiment.

[0135] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0136] The embodiment of the present disclosure also provides a computer-readable storage medium on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor. The computer-readable storage medium can be a volatile or non-volatile computer-readable storage medium.

[0137] An embodiment of the present disclosure further proposes an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0138] The embodiments of the present disclosure also provide a computer program product, including a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0139] Figure 4 1 is a block diagram of a virtual reality evacuation experiment device 1900 based on multi-person online collaboration according to an exemplary embodiment. For example, the device 1900 can be provided as a server. Figure 4 , the apparatus 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions, such as an application, that can be executed by the processing component 1922. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0140] The device 1900 may also include a power supply component 1926 configured to perform power management of the device 1900, a wired or wireless network interface 1950 configured to connect the device 1900 to a network, and an input / output interface 1958 (I / O interface). The device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server 2000. TM etc. operating systems capable of running 3D engines.

[0141] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the device 1900 to perform the above method.

[0142] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A virtual reality evacuation experimental method based on multi-person online collaboration, characterized in that: The method comprises: Based on the scene templates in the preset scene model library, a virtual reality disaster scene is built to simulate the disaster process in the real disaster scene through the test equipment used by the test subjects; After triggering the virtual reality disaster scene to run, obtaining evacuation experimental data generated by the test subject based on the virtual reality disaster scene displayed by the test device; Through each non-player character NPC based on the intelligent agent model in the virtual reality disaster scene, the movement and control strategy of the NPC is determined based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so that each NPC is displayed according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interacts with the test subject to perform an evacuation experiment; wherein different NPCs correspond to different intelligent agent models.

2. The method according to claim 1, characterized in that: The step of determining the movement and control strategy of each NPC based on the agent model in the virtual reality disaster scene based on the evacuation experiment data and the scene data of the virtual reality disaster scene includes: For each NPC based on the agent model, obtaining setting parameters of the NPC; wherein the setting parameters include an initial movement position and an end movement position of the NPC; By means of the NPC, based on the evacuation experiment data and the obstacle information and the topological map indicated by the scene data, planning an initialization movement path of the NPC from the movement initial position to the movement end position in the virtual reality disaster scene; The NPC moves according to the initialized motion path, and the behavior of the NPC is determined based on preset heuristic rules during the movement, so as to simulate the behavior of a real person when moving in the actual disaster scene.

3. The method according to claim 2, characterized in that The NPC's behavior is determined based on preset heuristic rules during the movement, including: Acquire the physical information, social information and environmental information indicated by the evacuation experiment data and the scenario data; Behaviors indicated by the heuristic rules corresponding to the physical information, the social information, and the environmental information are determined.

4. The method according to claim 2, characterized in that: The setting parameters also include the NPC's generation position, disappearance position and / or return position; accordingly, the method also includes: Controlling the generation of the NPC at a generation location of the virtual reality disaster scene to trigger the step of determining the movement and control strategy of the NPC based on the evacuation experiment data and the scene data of the virtual reality disaster scene through each agent-based NPC in the virtual reality disaster scene; and / or, Controlling the NPC to disappear at the disappearance position of the virtual reality disaster scene; and / or, At the return position of the virtual reality disaster scene, the NPC is controlled to return to the initial position of movement to trigger the execution of the step of planning, by the NPC, an initialization movement path of the NPC from the initial position of movement to the end position of movement in the virtual reality disaster scene based on the obstacle information and topological map indicated by the scene data, and subsequent steps.

5. The method according to any one of claims 1 to 4, characterized in that: The virtual reality disaster scene is constructed based on the scene template in the preset scene model library, including: Calling a three-dimensional engine to load a pre-built scene map of the virtual reality disaster scene; Obtaining a selection instruction for a scene template in the scene model library; The scene template indicated by the selection instruction is loaded into the scene map by the three-dimensional engine, so as to generate the virtual reality disaster scene by combining the scene template indicated by the selection instruction and the scene map; Wherein, the scene map and the scene template are both in preset formats supported by the three-dimensional engine for loading.

6. The method according to any one of claims 1 to 4, characterized in that: After triggering the virtual reality disaster scene to run, the method further includes: Calling simulation software to perform disaster dynamics simulation on the virtual reality disaster scene; The simulation results of the simulation software are loaded to be displayed in the virtual reality disaster scene of the device under test.

7. The method according to any one of claims 1 to 4, characterized in that: The virtual reality disaster scene includes at least two test subjects, and the method further includes: For each test subject, the evacuation experiment data of the test subject is sent in a preset data format through a preset data interface to other test devices used by other test subjects different from the test subject, so that the other test devices can display the test subject virtual image corresponding to the test subject in the virtual reality disaster scene based on the evacuation experiment data.

8. A virtual reality evacuation experiment system based on multi-person online collaboration, characterized in that: The system comprises: a device under test, and a server connected to the device under test in communication; The server is used to build a virtual reality disaster scene based on a scene template in a preset scene model library; The test device is used to display the virtual reality disaster scene to simulate the disaster process in the real disaster scene; after triggering the operation of the virtual reality disaster scene, the evacuation experimental data generated by the test object based on the virtual reality disaster scene displayed by the test device is collected; The server is further used to obtain the evacuation experiment data after triggering the operation of the virtual reality disaster scene; determine the movement and control strategy of the NPC based on the intelligent agent model in the virtual reality disaster scene based on the evacuation experiment data and the scene data of the virtual reality disaster scene, so as to display each NPC according to the movement and control strategy in the virtual reality disaster scene displayed by the test device, and interact with the test object; wherein different NPCs correspond to different intelligent agent models.

9. A virtual reality evacuation experimental device based on multi-person online collaboration, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to implement the method described in any one of claims 1 to 7 when executing the instructions stored in the memory.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images