A virtual reality classroom simulation method

Abstract

The application provides a virtual reality classroom simulation method, and relates to the technical field of data processing, and the method comprises the following steps: obtaining a high-dimensional data matrix with space-time correlation characteristics; constructing an electrical connection relationship graph; obtaining quantum state evolution results of a power system transient process; calculating a plurality of physical field distribution parameters corresponding to the equipment; based on the plurality of physical field distribution parameters, combining molecular dynamics simulation and a fluid mechanics model to obtain arc dynamic characteristic data; collecting an electromyographic signal of an operator, and combining mechanical parameters of the equipment to generate force feedback control instructions through a neural network mapping model; and generating a virtual reality scene that is adaptive to visual response and operation state of the operator by using a dynamic rendering engine according to eye tracking data, arc dynamic characteristic data and the force feedback control instructions. The embodiment of the application can improve the authenticity and interactivity of virtual reality classroom simulation.

Description

Technical Field

[0001] This application relates to, but is not limited to, the field of data processing technology, and in particular to a virtual reality classroom simulation method. Background Technology

[0002] Currently, power system teaching mainly relies on textbook explanations, two-dimensional simulation software, and some practical training systems. Although some platforms have introduced virtual reality technology to achieve basic operation training, they are mostly based on static scenes and simple interactions, making it difficult to realistically reflect the dynamic operation and complex fault processes of power systems.

[0003] However, existing methods generally suffer from poor real-time performance, insufficient interactivity, and missing physical parameters, failing to meet the needs of immersive, data-driven teaching. They lack visualization of key parameters such as current and temperature, lack force feedback support for complex operations like circuit breakers and electric arcs, and have weak mechanisms for multi-person collaboration and behavior tracking. Therefore, the teaching effectiveness and assessment credibility need improvement. Summary of the Invention

[0004] In view of this, embodiments of this application provide at least one method for simulating a virtual reality classroom. The technical solution of embodiments of this application is implemented as follows:

[0005] On one hand, embodiments of this application provide a virtual reality classroom simulation method, including:

[0006] Tensor decomposition is performed on the SCADA historical data stream of the power system to obtain a high-dimensional data matrix with spatiotemporal correlation characteristics.

[0007] Based on the configuration parameters of power equipment, a graph neural network is used for topology self-learning to construct an electrical connection relationship graph.

[0008] The high-dimensional data matrix and the electrical connection relationship diagram are input into the quantum computing simulation module to obtain the quantum state evolution results of the transient process of the power system;

[0009] Based on the quantum state evolution results, the electromagnetic-thermal-mechanical coupled field calculation engine is invoked to calculate the multiphysics field distribution parameters corresponding to the device;

[0010] Based on the multiphysics field distribution parameters, combined with molecular dynamics simulation and fluid dynamics model, the dynamic characteristics data of the electric arc are obtained.

[0011] The operator's electromyography signals are collected and combined with the mechanical parameters of the equipment to generate force feedback control commands through a neural network mapping model;

[0012] Based on eye-tracking data, the dynamic characteristics of the electric arc, and the force feedback control commands, a virtual reality scene that adapts to the operator's visual response and operating state is generated using a dynamic rendering engine.

[0013] In some embodiments, the tensor decomposition process of the SCADA historical data stream of the power system to obtain a high-dimensional data matrix with spatiotemporal correlation characteristics includes:

[0014] Construct a three-dimensional tensor with timestamps, device identifiers, and electrical parameters as dimensions;

[0015] The three-dimensional tensor is decomposed using the Tucker tensor decomposition algorithm to obtain the core tensor and factor matrix;

[0016] The core tensor and the factor matrix are fused and reconstructed into the high-dimensional data matrix with spatiotemporal correlation characteristics;

[0017] The high-dimensional data matrix is ​​used to characterize the temporal dependency and spatial coupling relationship between nodes of the power system.

[0018] In some embodiments, the step of constructing an electrical connection graph using a graph neural network for topology self-learning based on the configuration parameters of the power equipment includes:

[0019] An initial electrical diagram is constructed using multiple key devices as nodes in a graph structure and the connection relationships between these key devices as edges in the graph structure; the categories of key devices include circuit breakers, transformers, and busbars.

[0020] The GraphSAGE algorithm in graph neural networks is used to extract node features and edge weight information of the graph structure to identify implicit electrical dependencies.

[0021] Output an electrical connection diagram with connection weights, device category labels, and adjacency structures.

[0022] In some embodiments, inputting the high-dimensional data matrix and the electrical connection relationship map into a quantum computing simulation module to obtain the quantum state evolution results of the power system transient process includes:

[0023] The spatiotemporal data obtained based on tensor decomposition and graph learning are transformed into a set of state variables, and the corresponding transient evolution differential equations are constructed.

[0024] The transient evolution differential equations are equivalently represented as a Hamiltonian model in the form of a quantum circuit.

[0025] The variable quantum eigenfunction solver is invoked to minimize the objective energy function, simulating the quantum state evolution process of the power system under transient disturbance conditions;

[0026] Quantitative feature values, including node voltage changes, frequency fluctuations, and phase angle shifts, are extracted through quantum measurement operations and used as the output of the quantum state evolution result.

[0027] In some embodiments, the step of calling the electromagnetic-thermal-mechanical coupled field calculation engine based on the quantum state evolution result to calculate the multiphysics field distribution parameters corresponding to the device includes:

[0028] Based on the quantum state evolution results, physical quantities at the corresponding time nodes are extracted; these physical quantities include current density, node voltage, and power flow information.

[0029] The physical quantities are input into the electromagnetic-thermal-mechanical coupled field calculation engine to construct a set of multi-physical quantity control equations.

[0030] Based on the set of governing equations for the multi-physical quantities, a joint partial differential equation system including Maxwell's equations, heat conduction equations, and stress-strain equations is solved, and the three-dimensional multi-physics field distribution parameters including the Lorentz force distribution, temperature gradient field, and mechanical stress tensor are output as multi-physics field distribution parameters.

[0031] In some embodiments, obtaining arc dynamic characteristic data based on the multiphysics field distribution parameters, combined with molecular dynamics simulation and fluid dynamics model, includes:

[0032] Molecular dynamics simulations were used to estimate the electron free path, collision cross section, and ionization rate.

[0033] The flow behavior of the electric arc plasma is solved by combining the k-ε turbulence model and the Navier-Stokes equations, and the three-dimensional dynamic vector data of the arc morphology evolution, core temperature distribution and ionization region range are output as the dynamic characteristic data of the arc.

[0034] In some embodiments, the process of acquiring the operator's electromyographic signals and combining them with the device's mechanical parameters to generate force feedback control commands through a neural network mapping model includes:

[0035] The surface electromyographic signals of the radial wrist flexor muscles of the operator were collected to obtain the acquired signals.

[0036] The acquired signal is processed by performing action recognition and intent classification on the acquired signal based on a one-dimensional convolutional neural network.

[0037] The processed acquired signal is combined with the mechanical characteristic curve of the circuit breaker to generate a corresponding damping force feedback signal waveform as the force feedback control command.

[0038] In some embodiments, generating a virtual reality scene that adapts to the operator's visual response and operating state using a dynamic rendering engine based on eye-tracking data, the arc dynamic characteristic data, and the force feedback control command includes:

[0039] Based on the eye-tracking system, the pupil diameter, fixation point position and saccade path are collected in real time to identify the operator's visual attention area and alertness state;

[0040] By combining the aforementioned arc dynamic characteristic data, the visual brightness, particle density, and flicker frequency of the arc shape in the virtual scene are dynamically adjusted to obtain the adjusted visual brightness, particle density, and flicker frequency of the arc shape.

[0041] Receive the force feedback control command and map the force feedback control command to the real-time deformation, interaction resistance and material response effect of the device in the virtual environment;

[0042] By integrating the operator's visual attention area and alertness state, the visual brightness of the adjusted electric arc shape, particle density and flicker frequency, the real-time deformation of the device, interaction resistance and material response effects, an immersive 3D scene image that conforms to the operator's current perceptual characteristics and operational intentions is generated in real time as the virtual reality scene.

[0043] In some embodiments, the method further includes:

[0044] Construct an evaluation vector sequence of the trainee's operational ability in the virtual reality scene; each vector in the evaluation vector sequence represents the trainee's score in handling various types of faults;

[0045] Based on the evaluation vector sequence, a reinforcement learning model is constructed using a deep deterministic policy gradient algorithm, and a personalized training scenario sequence adapted to the learner's ability state is generated according to the output of the reinforcement learning model.

[0046] By combining the statistical distribution patterns of historical fault data, random fault events are dynamically injected into the training scenario sequence.

[0047] In some embodiments, dynamically injecting random fault events into the training scenario sequence based on the statistical distribution patterns of historical fault data includes:

[0048] When the power system operates stably for a continuous period of time in the training environment for more than a set threshold, a random fault injection is triggered.

[0049] Based on the set weighted probability model, the target fault type is extracted from the preset rare fault event library;

[0050] The severity of the fault event of the target fault type is negatively correlated with the assessment vector of the trainee's current ability, so as to achieve targeted training for weak areas.

[0051] On the other hand, this application provides a computer system including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the program to implement the steps in the method described above.

[0052] The beneficial effects of this application include at least the following:

[0053] (1) This application achieves high-precision modeling and dynamic prediction of complex transient processes of power systems by performing tensor decomposition on SCADA historical data streams, combining the graph neural network self-learning structure of electrical topology map, and introducing quantum computing simulation module, which significantly improves the simulation system's ability to reproduce the real power grid operating state.

[0054] (2) An electromagnetic-thermal-mechanical coupling calculation engine was constructed, and further combined with molecular dynamics and fluid dynamics models to obtain three-dimensional dynamic evolution data of electric arc morphology. This can accurately characterize the multi-physics coupling effect of equipment response under typical fault scenarios, providing more physically based data support for subsequent perception feedback and teaching interaction.

[0055] (3) By introducing an electromyography signal acquisition and motion recognition neural network, an electromyography-mechanical signal mapping model is constructed to generate force feedback control commands in real time. The commands are then integrated with eye tracking data and arc characteristic data in the dynamic rendering engine to achieve a highly personalized, immersive, and responsive three-dimensional interactive environment, which significantly enhances the operator's sense of immersion and ability to present their operational intentions.

[0056] This application overcomes the key bottlenecks in existing power teaching methods, such as poor real-time performance, weak interactivity, insufficient physical accuracy, and inaccurate training feedback. It constructs an integrated virtual reality classroom system with high realism, multi-physics field support, closed-loop human-computer interaction, and personalized training mechanism, which has significant engineering practical value and promotion prospects. Attached Figure Description

[0057] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with this application and, together with the specification, serve to explain the technical solutions of this application.

[0058] Figure 1 This is a schematic diagram illustrating the implementation process of a virtual reality classroom simulation method provided in an embodiment of this application.

[0059] Figure 2 This is a schematic diagram of the hardware entity of a computer system provided in an embodiment of this application. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application are further described in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. In the following description, the term "some embodiments" refers to a subset of all possible embodiments. However, it is understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first / second / third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first / second / third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0061] This application provides a virtual reality classroom simulation method, which can be executed by a processor of a computer system. The computer system can refer to devices with data processing capabilities, such as servers, laptops, tablets, and desktop computers.

[0062] Figure 1 This is a schematic diagram illustrating the implementation process of a virtual reality classroom simulation method provided in an embodiment of this application, as shown below. Figure 1 As shown, the method includes:

[0063] Step S10: Perform tensor decomposition on the SCADA historical data stream of the power system to obtain a high-dimensional data matrix with spatiotemporal correlation characteristics.

[0064] In practice, this step involves tensor decomposition of the SCADA historical data stream of the power system. It is a basic data preprocessing step for virtual reality classroom simulation, aiming to extract high-dimensional structural information with spatiotemporal correlation characteristics from massive monitoring data to support subsequent power system state modeling and dynamic scene generation.

[0065] Specifically, the raw data collected by the SCADA system (such as timestamps, device IDs, voltages, currents, etc.) can be organized into a three-dimensional tensor. The dimensions are usually: time × device × electrical parameters. This structure preserves the temporal continuity and spatial topological distribution characteristics of the device operation data.

[0066] Next, the Tucker tensor decomposition algorithm can be used to decompose the above three-dimensional tensor into a low-rank core tensor and several factor matrices, from which the main variation features and coupling patterns can be extracted, effectively reducing redundant data dimensions and compressing storage overhead.

[0067] Ultimately, the core tensor and factor matrix can be fused to reconstruct a high-dimensional data matrix with stronger representation capabilities, which preserves key temporal evolution patterns and spatial dependencies between devices.

[0068] It is understandable that the final output high-dimensional data matrix can serve as the basic input for subsequent steps such as graph neural network modeling and quantum evolution calculation, thereby achieving a high-fidelity description and modeling of the power system state.

[0069] In some embodiments, step S10 may include:

[0070] Construct a three-dimensional tensor with timestamps, device identifiers, and electrical parameters as dimensions;

[0071] The Tucker tensor decomposition algorithm is used to decompose the three-dimensional tensor to obtain the core tensor and factor matrix;

[0072] The core tensor and factor matrix are fused and reconstructed into a high-dimensional data matrix with spatiotemporal correlation characteristics;

[0073] Among them, the high-dimensional data matrix is ​​used to characterize the temporal dependence and spatial coupling relationship between nodes in the power system.

[0074] Specifically, firstly, tensor modeling and decomposition of the SCADA historical data stream of the power system can be performed to obtain a high-dimensional data matrix with spatiotemporal correlation characteristics, which can be used for subsequent multi-model joint simulation and virtual reality generation processes.

[0075] Next, the raw SCADA data can be organized according to the following three dimensions to construct a three-dimensional tensor. :

[0076] A timestamp is a series of consecutive time points from which data is collected, denoted as... ;

[0077] Device IDs represent the various data acquisition devices in the system, such as circuit breakers and transformers, and are denoted as... ;

[0078] Electrical parameters can include voltage, current, active power, reactive power, etc. Electrical parameters.

[0079] Therefore, the tensor structure can be:

[0080]

[0081] in, Indicates the first The time point, the first The first device Each electrical parameter value.

[0082] Finally, we can work on tensors Using Tucker decomposition, it can be represented as:

[0083]

[0084] in, For the core tensor; These are the factor matrices obtained by decomposition along the three-dimensional directions of time, device, and electrical parameters, respectively. Indicates along the first Tensor products of dimensions.

[0085] The fusion and reconstruction of high-dimensional matrices can transform tensors The high-dimensional data matrix is ​​reconstructed by fusing the core tensor and factor matrix. ,For example:

[0086]

[0087] in, To determine the number of feature dimensions to be fused.

[0088] This high-dimensional data matrix It can effectively characterize the time-series dependencies between nodes in a power system: such as Reflecting the trend of sudden changes in operation; spatial coupling: such as Implicit connections between devices; cross-dimensional fusion capabilities: The main interaction structure and global features of the original tensor are preserved.

[0089] In summary, the high-dimensional structure data of this application, as the basic input for subsequent steps such as graph learning, quantum evolution simulation and multiphysics simulation, can significantly improve the accuracy and response speed of modeling.

[0090] Step S20: Based on the configuration parameters of the power equipment, a graph neural network is used for topology self-learning to construct an electrical connection relationship graph.

[0091] Based on the configuration parameters of power equipment, a graph neural network is used for topology self-learning to construct an electrical connection relationship graph, which aims to automatically identify the actual connection relationships and potential dependency structures between various devices in the power system.

[0092] Specifically, key power equipment such as circuit breakers, transformers, and busbars can first be modeled as nodes in a graph structure, and the connections between equipment can be modeled as edges. Then, based on configuration parameters (such as equipment type, rating, connection port, etc.), graph structure learning and embedding representation extraction are performed using graph neural networks (such as GraphSAGE or GAT models). The final output electrical connection graph not only retains the original equipment topology but also captures the potential functional coupling and state correlation between equipment, providing a more intelligent and accurate data foundation for subsequent physical modeling and system simulation.

[0093] In some embodiments, step S20 may include:

[0094] An initial electrical diagram is constructed using multiple key devices as nodes in the graph structure and the connections between these key devices as edges. The categories of key devices include circuit breakers, transformers, and busbars.

[0095] The GraphSAGE algorithm in graph neural networks is used to extract node features and edge weight information of graph structures to identify implicit electrical dependencies.

[0096] Output an electrical connection diagram with connection weights, device category labels, and adjacency structures.

[0097] Specifically, multiple key devices in the power system can first be abstracted as nodes in a graph structure. These key devices include, but are not limited to, circuit breakers, transformers, and busbars. Based on the primary wiring diagram or SCADA configuration data, the connection relationships between the devices are extracted and used as edges in the graph structure to form an initial electrical diagram.

[0098] The initial graph structure is then input into the graph neural network model, and the GraphSAGE (Graph Sample and Aggregate) algorithm is used to train the nodes in the graph. GraphSAGE learns the embedded feature representations of nodes by sampling and aggregating the neighbor information of each node. Specifically, this includes:

[0099] Node feature extraction: Considering equipment type, electrical capacity, status flags, etc.;

[0100] Edge feature extraction: Consider connection strength, whether it is switchable, connection method, etc.

[0101] Through multiple rounds of iterative updates, we uncover the implicit dependencies and mutual influences between nodes.

[0102] Finally, after the network training is complete, an enhanced graph structure can be output, in which each node is labeled with a device category, each edge contains connection weights (such as impedance, dependency strength, etc.), and the complete adjacency structure is preserved. This graph not only reflects physical connectivity relationships but also integrates state and functional information, which can be used for subsequent system modeling, simulation, and control optimization.

[0103] Through this topology self-learning process, the system can extract complex electrical network relationships from the original configuration, achieve intelligent perception and structural reconstruction, and enhance the accuracy and realism of subsequent multiphysics modeling and virtual simulation.

[0104] Step S30: Input the high-dimensional data matrix and electrical connection relationship graph into the quantum computing simulation module to obtain the quantum state evolution results of the transient process of the power system.

[0105] In practice, the high-dimensional data matrix and electrical connection relationship map obtained in the early stage can be used as joint inputs and fed into the quantum computing simulation module to simulate the dynamic evolution process of the power system under transient disturbance conditions. By constructing an equivalent quantum Hamiltonian model and solving it using a variable quantum algorithm, the quantum state evolution results characterizing information such as node voltage, frequency, and phase angle changes are finally obtained, providing quantitative support for subsequent multiphysics modeling.

[0106] In some embodiments, step S30 may include:

[0107] The spatiotemporal data obtained based on tensor decomposition and graph learning are transformed into a set of state variables, and the corresponding transient evolution differential equations are constructed.

[0108] The transient evolution differential equation system is equivalently represented as a Hamiltonian model in the form of a quantum circuit;

[0109] The variable quantum eigenfunction solver is invoked to minimize the objective energy function, simulating the quantum state evolution process of the power system under transient disturbance conditions;

[0110] Quantitative characteristic values, including node voltage changes, frequency fluctuations, and phase angle shifts, are extracted through quantum measurement operations and used as the output of quantum state evolution results.

[0111] Specifically, the high-dimensional spatiotemporal data matrix obtained through tensor decomposition can be fused with the electrical connection relationship map extracted by graph neural network to extract key parameters in power system operation, such as node voltage, current density, and electrical topology, and uniformly represent them as a set of state variables:

[0112]

[0113] The above state variables characterize the dynamic evolution trend of the system at different time points.

[0114] Next, based on the coupling relationship between state variables, a set of differential equations can be established to characterize the transient behavior of the system, such as dynamic power flow equations and transient voltage-current response equations, to describe the response evolution process after disturbance.

[0115] Then, the above system of differential equations can be mapped to the Hamiltonian representation of the quantum system through physical modeling, that is, the corresponding quantum circuit model can be constructed:

[0116]

[0117] in, The control parameters represent the parameterized variable quantum states.

[0118] Then, a Variational QuantumEigensolver (VQE) can be used to minimize the objective energy function:

[0119]

[0120] Parameters in the variational circuit are optimized through iterative optimization. It approximates the energy spectrum structure of the system's stable and excited states, and simulates the quantum state evolution process of the power system under transient disturbance conditions.

[0121] Finally, through quantum measurement operations, the quantum state is projected and measured to extract key physical quantities characterizing the system's operating state, such as node voltage change Δt, frequency fluctuation Δt, and phase angle shift Δt, which are output in the form of quantized characteristic values, constituting a quantum-level expression of the transient evolution behavior of the power system.

[0122] In summary, the quantum computing process of this application not only improves the simulation accuracy of system behavior under complex perturbation states, but also provides efficient and scalable underlying support for subsequent multiphysics field linkage analysis and operation interaction modeling.

[0123] Step S40: Based on the quantum state evolution results, call the electromagnetic-thermal-mechanical coupled field calculation engine to calculate the multiphysics field distribution parameters corresponding to the device.

[0124] In practical implementation, the quantum state evolution results of transient processes in a power system can be analyzed to obtain the changes in key physical quantities at each node under system disturbances, such as current density, voltage fluctuations, and power flow. Based on these results, an electromagnetic-thermal-mechanical coupled field calculation engine is further invoked to construct a multi-physics governing equation system using these physical quantities as boundary conditions or initial parameters. Subsequently, a system of partial differential equations, including Maxwell's equations, heat conduction equations, and stress-strain equations, is solved to obtain three-dimensional multi-physics field distribution parameters such as Lorentz force distribution, temperature gradient, and mechanical stress of the equipment under specific operating conditions. These parameters provide accurate inputs for subsequent steps such as arc characteristic extraction, virtual feedback generation, and equipment response modeling.

[0125] In some embodiments, step S40 may include:

[0126] Based on the quantum state evolution results, physical quantities at the corresponding time nodes are extracted; these physical quantities include current density, node voltage, and power flow information.

[0127] Input physical quantities into the electromagnetic-thermal-mechanical coupled field calculation engine to construct a set of multi-physical quantity control equations;

[0128] Based on the multi-physical governing equations, a joint partial differential equation system including Maxwell's equations, heat conduction equations, and stress-strain equations is solved, and the three-dimensional multi-physics field distribution parameters including Lorentz force distribution, temperature gradient field, and mechanical stress tensor are output as multi-physics field distribution parameters.

[0129] In practical implementation, in the virtual reality classroom simulation method, the quantum state evolution results of the transient process of the power system obtained through the quantum computing simulation module can accurately capture the transient behavior of the system under disturbance.

[0130] Specifically, target time points can be selected from quantum state evolution to extract key physical quantities related to the device state, including but not limited to: current density (used to describe the current distribution per unit area in a current-carrying conductor); node voltage (characterizing the voltage response characteristics of the system); and power flow information (reflecting the active and reactive power flow states of the device).

[0131] Next, a set of multi-physical-quantity governing equations can be constructed. The above physical quantities can be input as initial conditions or boundary conditions into the electromagnetic-thermal-mechanical coupled field calculation engine. The following coupled models are established in this module: Electromagnetic field governing equations: described by Maxwell's equations; Thermal field governing equations: the Joule heat generation and conduction process is modeled using the Fourier heat conduction equation; Mechanical stress field governing equations: constitutive relations and stress-strain governing equations are used to describe the deformation behavior of the equipment caused by thermal expansion, electromagnetic force, etc.

[0132] Next, the system of partial differential equations is solved jointly. The aforementioned governing equations are constructed in the form of a system of partial differential equations (PDEs), and the three are closely linked through coupling parameters (such as Joule heating generated by current and mechanical deformation induced by electromagnetic force). In the coupled simulation, the solver iteratively calculates the variation trends of each physical field with time and space.

[0133] Finally, the three-dimensional multiphysics field distribution parameters are output, and the following three-dimensional field parameter results closely related to the equipment's operating status are output: Lorentz force distribution: describes the volume force on the conductor under the action of electromagnetic field; temperature gradient field: characterizes the heat distribution inside or on the surface of the equipment caused by heat generation due to current flow; mechanical stress tensor: quantifies the deformation and stress state of the equipment under multiphysics action.

[0134] In summary, the three-dimensional multiphysics field distribution parameters of the present invention not only provide high-precision input data for subsequent modules such as arc behavior modeling and operator feedback generation, but also provide dynamic visualization support with a real physical basis for the teaching system.

[0135] Step S50: Based on the multi-physics field distribution parameters, and by combining molecular dynamics simulation and fluid dynamics model, obtain the dynamic characteristic data of the electric arc.

[0136] In practice, by further calculating and simulating the multiphysics field distribution parameters, and by combining molecular dynamics simulation and fluid dynamics modeling methods, high-precision electric arc dynamic characteristic data can be obtained to support the visualization and interactive feedback of electric arc behavior in a virtual reality environment.

[0137] This process integrates the dynamic behavior at the microscopic particle level with the evolution characteristics of the macroscopic plasma flow field, thereby achieving accurate modeling of arc morphology, temperature distribution, and changes in ionization regions, enhancing the realism and teaching value of the virtual training system.

[0138] In some embodiments, step S50 may include:

[0139] Molecular dynamics simulations were used to estimate the electron free path, collision cross section, and ionization rate.

[0140] By combining the k-ε turbulence model and the Navier-Stokes equations to solve the flow behavior of the electric arc plasma, three-dimensional dynamic vector data of the arc morphology evolution, core temperature distribution, and ionization region range are output as dynamic characteristic data of the electric arc.

[0141] Specifically, the acquisition of arc dynamics data depends on the collaborative modeling of molecular and fluid scales. First, molecular dynamics (MD) methods can be used to model the interactions between electrons, ions and neutral particles under a set microscopic physical environment.

[0142] Next, based on the particle's trajectory and collision patterns, key microscopic physical parameters can be estimated, including: the electron's mean free path; the collision cross-section with background particles; and the ionization rate at specific temperatures and electric field strengths.

[0143] With the support of molecular simulation parameters, a k-ε turbulence model is introduced to enclose the turbulence characteristics; a Navier-Stokes equations system suitable for high-temperature arc plasma is constructed, coupled with mass, momentum, and energy conservation equations; the above equations system can be numerically solved to obtain the following three-dimensional dynamic vector data as arc dynamic characteristic data: arc morphology evolution trajectory; arc core temperature distribution field; arc ionization region range and variation boundary.

[0144] The final output data can be used to guide the visual simulation and physical feedback modeling of electric arcs in virtual reality environments, enabling a highly immersive training experience.

[0145] Step S60: Collect the operator's electromyographic signals and combine them with the mechanical parameters of the equipment to generate force feedback control commands through a neural network mapping model.

[0146] In practice, the operator's electromyography (EMG) signals can be collected and combined with the mechanical parameters of the equipment to generate force feedback control commands through a neural network mapping model.

[0147] In some embodiments, step S60 may include:

[0148] The surface electromyographic signals of the radial wrist flexor muscles of the operator were collected to obtain the acquired signals.

[0149] The acquired signal is processed by performing action recognition and intent classification on the acquired signal based on a one-dimensional convolutional neural network.

[0150] The processed acquired signal is combined with the mechanical characteristic curve of the circuit breaker to generate a corresponding damping force feedback signal waveform as a force feedback control command.

[0151] Specifically, the surface electromyography (sEMG) signals of the operator's radial wrist flexor muscles can be collected in real time using a surface electromyography acquisition device. These signals can reflect the operator's forearm movement intentions and muscle contraction intensity, serving as the basis for subsequent movement recognition.

[0152] Next, the acquired raw electromyographic signals can be input into a one-dimensional convolutional neural network (1D-CNN) model for feature extraction and classification. This process includes:

[0153] Filtering, denoising, and normalization processing;

[0154] Multi-scale convolutional kernels extract local patterns from time series;

[0155] The corresponding operation action labels (such as closing, opening, emergency braking, etc.) and intent strength scores are output using the fully connected layer.

[0156] Finally, the identification results can be linked with the preset circuit breaker mechanical characteristic curves (including parameters such as spring stiffness, frictional resistance, and travel limit) to obtain a damping force feedback signal waveform that reflects the impedance change of the current operation.

[0157] The final output signal serves as a force feedback control command for the virtual device, used to control the device's dynamic response in the virtual reality environment, allowing the operator to obtain physical tactile and resistance feedback that matches their intention during operation.

[0158] Through the above methods, this application realizes virtual-real linkage control driven by electromyographic signals, which effectively enhances the immersion, responsiveness and muscle memory construction effect of the training system.

[0159] Step S70: Based on eye-tracking data, arc dynamic characteristic data, and force feedback control commands, use a dynamic rendering engine to generate a virtual reality scene that adapts to the operator's visual response and operating state.

[0160] In practice, by integrating the operator's eye-tracking data, arc dynamic characteristic data, and force feedback control commands, a dynamic rendering engine can be driven to generate a virtual reality scene in real time that highly matches the operator's visual focus area, operational intention, and feedback perception state. This process achieves adaptive linkage between vision, touch, and the physical scene, enhancing the immersion, interactivity, and realism of scene response during training.

[0161] In some embodiments, step S70 may include:

[0162] Based on the eye-tracking system, pupil diameter, fixation point position and saccade path are collected in real time to identify the operator's visual attention area and alertness state;

[0163] By combining the dynamic characteristics data of the electric arc, the visual brightness, particle density and flicker frequency of the electric arc shape in the virtual scene are dynamically adjusted to obtain the adjusted visual brightness, particle density and flicker frequency of the electric arc shape.

[0164] Receive force feedback control commands and map them to the real-time deformation, interaction resistance and material response effects of the device in the virtual environment;

[0165] By integrating the operator's visual attention area and alertness state, the visual brightness of the adjusted electric arc shape, particle density and flicker frequency, the real-time deformation of the device, interaction resistance and material response effects, an immersive 3D scene image that conforms to the operator's current perceptual characteristics and operational intentions is generated in real time as a virtual reality scene.

[0166] Specifically, an eye-tracking system can be used to collect multi-dimensional physiological signals such as the operator's pupil diameter, fixation point position, and saccade path in real time. By analyzing its spatial and temporal characteristics, the operator's visual attention area, focus of attention, and changes in alertness can be identified, thereby inferring the operator's current operational intention and focus of attention.

[0167] Next, the above recognition results can be linked with the arc dynamic characteristic data from the arc simulation module to adjust the visual features of the arc shape in the virtual scene in real time, including visual brightness (such as simulating high temperature and strong light), particle density (such as the arc plasma cloud dispersion range) and flicker frequency (such as the arc unstable flicker state) to enhance the realism and immersion of the visual response.

[0168] Then, force feedback control commands from the electromyographic signal-driven model can be received and parsed, and mapped to the deformation behavior (such as pressing, disconnecting, etc.), interactive resistance (such as reaction force during operation), and material response effects (such as physical feedback such as elasticity and hardness) of the virtual device under different operating conditions, thereby constructing a multimodal tactile feedback channel.

[0169] Finally, based on the dynamic rendering engine, visual attention information, electric arc behavior simulation data and force feedback control results can be fused and processed to drive a high-performance graphics rendering system to generate a three-dimensional immersive virtual reality image that matches the operator's current cognitive and action state in real time, thereby achieving a high-fidelity training experience that combines virtual and real interaction and human-machine collaboration.

[0170] In some embodiments, the method of this application may further include:

[0171] Construct an evaluation vector sequence of trainees' operational capabilities in a virtual reality scenario; each vector in the evaluation vector sequence represents a trainee's score in handling various types of faults.

[0172] Based on the evaluation vector sequence, a reinforcement learning model is constructed using a deep deterministic policy gradient algorithm, and a personalized training scenario sequence adapted to the learner's ability state is generated according to the output of the reinforcement learning model.

[0173] By combining the statistical distribution patterns of historical fault data, random fault events are dynamically injected into the training scenario sequence.

[0174] During training, the system continuously collects the trainee's interactive behavior and fault handling performance in the virtual reality scenario. For each typical fault type (such as short circuit, arc breakdown, bus fault, loss of synchronization, etc.), the system scores the trainee based on multiple indicators, including task completion time, accuracy, and operation sequence matching degree, constructing an evaluation vector for the trainee's operational ability.

[0175]

[0176] in, Indicates the first Evaluation results within each training cycle Indicates that the students are in the The score for handling fault-like tasks reflects the user's mastery and responsiveness. All evaluation vectors form a dynamically updated vector sequence. .

[0177] Using the aforementioned capability assessment vector sequence, the system establishes a training scenario generation policy based on the Deep Deterministic Policy Gradient (DDPG) algorithm:

[0178] This algorithm can optimize the policy function in a continuous action space and output personalized training scenario sequences for different ability states, so that the training content always matches the learner's ability level, thereby improving learning efficiency and the depth of skill mastery.

[0179] To enhance the robustness of training and its ability to respond to emergencies, the system also dynamically introduces random fault events by incorporating historical fault distribution data from actual power system operation. These events include:

[0180] This mechanism effectively avoids overly idealistic training, simulates sudden risk scenarios in real-world working conditions, and promotes trainees' comprehensive response capabilities under various emergency conditions.

[0181] In summary, through the aforementioned reinforcement learning strategy and dynamic fault injection mechanism, this application enables an intelligent training process that is "tailored to individual students and abilities," significantly improving the adaptability, challenge, and practical effectiveness of virtual reality teaching systems.

[0182] In some embodiments, the step of "dynamically injecting random fault events into the training scenario sequence by combining the statistical distribution patterns of historical fault data" may include:

[0183] When the power system operates stably for a continuous period of time in the training environment for more than a set threshold, random fault injection is triggered.

[0184] Based on the set weighted probability model, the target fault type is extracted from the preset rare fault event library;

[0185] By negatively correlated the severity of the target fault type with the trainee's current ability assessment vector, targeted training can be provided for weak areas.

[0186] Specifically, the system can determine whether the power system in the training environment is in a continuous and stable operation phase by monitoring its operating status in real time. The fault injection process is triggered when the following conditions are met:

[0187] This mechanism is used to simulate a "low alertness - high contingency" scenario in real-world operations, breaking the inertia of the training process and triggering more challenging operational demands.

[0188] Once the triggering conditions are met, the system will automatically activate the rare fault event library invocation module. This module is built from historical operating and statistical data of the power system, covering events such as:

[0189] Using a weighted probability model The candidate fault types are sampled, and the model weights are composed of the following factors:

[0190] This strategy ensures that the selected faults are both engineering-representative and maintain sufficient training diversity.

[0191] The system will use the current student's ability assessment vector Associating fault types, the following allocation logic is used:

[0192]

[0193] in, The score is given to the trainee's ability to handle this type of fault. To prevent small constants from being divided by zero.

[0194] Ultimately through normalization Furthermore, the sampling weights are reconstructed to prioritize the injection of high severity and low mastery types, thereby strengthening the focus on refining weak points during training and improving the relevance and practicality of training results.

[0195] In summary, this application utilizes intelligent triggering, probability regulation, and capability-oriented injection strategies to enable training content to achieve "mutational breakthroughs" in a stable state, and focuses on strengthening weak links, thereby significantly enhancing trainees' ability to handle low-probability, high-risk power failure events.

[0196] It should be noted that, in the embodiments of this application, if the above-described virtual reality classroom simulation method is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to the related technology, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), magnetic disks, or optical disks. Thus, the embodiments of this application are not limited to any specific hardware, software, or firmware, or any combination of hardware, software, and firmware.

[0197] This application provides a computer system including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the program, it implements some or all of the steps in the above-described method.

[0198] This application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements some or all of the steps in the above-described method. The computer-readable storage medium can be transient or non-transient.

[0199] This application provides a computer program including computer-readable code, wherein when the computer-readable code is executed in a computer device, a processor in the computer device performs some or all of the steps in the above-described method.

[0200] This application provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. When the computer program is read and executed by a computer, it implements some or all of the steps in the above-described method. This computer program product can be implemented specifically through hardware, software, or a combination thereof. In some embodiments, the computer program product is specifically embodied as a computer storage medium; in other embodiments, the computer program product is specifically embodied as a software product, such as a software development kit (SDK), etc.

[0201] It should be noted that the descriptions of the various embodiments above tend to emphasize the differences between them, while their similarities or commonalities can be referred to interchangeably. The descriptions of the above embodiments of the device, storage medium, computer program, and computer program product are similar to the descriptions of the above method embodiments and have similar beneficial effects. For technical details not disclosed in the embodiments of the device, storage medium, computer program, and computer program product of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0202] Figure 2 A hardware entity diagram of a computer system provided in an embodiment of this application is shown below. Figure 2 As shown, the hardware entity of the computer system 1000 includes a processor 1001 and a memory 1002, wherein the memory 1002 stores a computer program that can run on the processor 1001, and the processor 1001 executes the program to implement the steps in the method of any of the above embodiments.

[0203] The memory 1002 stores computer programs that can run on the processor. The memory 1002 is configured to store instructions and applications that can be executed by the processor 1001. It can also cache data to be processed or already processed (e.g., image data, audio data, voice communication data, and video communication data) in the processor 1001 and various modules in the computer system 1000. It can be implemented by flash memory or random access memory (RAM).

[0204] The processor 1001 executes the program to implement the steps of any of the above-mentioned virtual reality classroom simulation methods. The processor 1001 typically controls the overall operation of the computer system 1000.

[0205] This application provides a computer storage medium storing one or more programs that can be executed by one or more processors to implement the steps of the virtual reality classroom simulation method as described in any of the above embodiments.

[0206] It should be noted that the descriptions of the storage medium and device embodiments above are similar to those of the method embodiments above, and have similar beneficial effects. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding. The processor described above can be at least one of the following: Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), Central Processing Unit (CPU), Controller, Microcontroller, and Microprocessor. It is understood that the electronic device implementing the above processor function can also be other types, and this application does not specifically limit the specific types.

[0207] The aforementioned computer storage media / memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM), etc.; or it can be various terminals that include one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.

[0208] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence number of the above-described steps / processes does not imply the order of execution; the execution order of each step / process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments. It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. In the absence of further restrictions, an element defined by the phrase "including one..." does not preclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0209] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0210] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0211] In addition, each functional unit in the various embodiments of this application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0212] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0213] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence or the part that contributes to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, magnetic disks, or optical disks.

[0214] The above description is merely an embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A virtual reality classroom simulation method, characterized by, The method comprises the following steps: performing tensor decomposition processing on the SCADA historical data stream of the power system to obtain a high-dimensional data matrix with space-time correlation characteristics; based on the configuration parameters of the power equipment, performing topology self-learning using a graph neural network to construct an electrical connection relationship graph; inputting the high-dimensional data matrix and the electrical connection relationship graph into a quantum computing simulation module to obtain quantum state evolution results of the transient process of the power system; based on the quantum state evolution results, calling an electromagnetic-thermal-mechanical coupling field calculation engine to calculate the multi-physical field distribution parameters corresponding to the equipment; based on the multi-physical field distribution parameters, jointly simulating molecular dynamics and fluid mechanics to obtain arc dynamic characteristic data; collecting the electromyographic signals of the operator and combining the mechanical parameters of the equipment to generate force feedback control instructions through a neural network mapping model; based on the eye tracking data, the arc dynamic characteristic data, and the force feedback control instructions, generating a virtual reality scene that is adaptive to the visual response and operation state of the operator using a dynamic rendering engine; The evaluation vector sequence of the operation ability of the trainee in the virtual reality scene is constructed; each vector in the evaluation vector sequence represents the processing ability score of the trainee for each type of fault, and each vector in the evaluation vector sequence is represented as wherein, represents the evaluation result in the first training cycle, represents the processing score of the trainee in the first type of fault task, reflecting the mastery degree and response ability, and all evaluation vectors form a dynamically updated evaluation vector sequence . based on the evaluation vector sequence, constructing a reinforcement learning model using a deep deterministic policy gradient algorithm, and outputting a personalized training scene sequence that adapts to the ability state of the trainee according to the reinforcement learning model; The statistical distribution law of historical fault data is combined, and random fault events are dynamically injected in the training scene sequence, including: triggering random fault injection when the power system continuously and stably runs in the training environment for more than a set threshold; based on a set weighted probability model, extracting a target fault type from a preset rare fault event library; the severity of the fault event of the target fault type is negatively correlated with the evaluation vector of the current ability of the student, so as to realize the key training of the weak link, and the severity of the fault event of the target fault type and the evaluation vector of the current ability of the student are represented as: Wherein, is the ability score of the student in handling such faults, is a small constant to prevent division by zero.

2. The method of claim 1, wherein, the method of performing tensor decomposition processing on the SCADA historical data stream of the power system to obtain a high-dimensional data matrix with space-time correlation characteristics comprises the following steps: constructing a three-dimensional tensor with time stamp, device identifier, and electrical parameter as dimensions; performing decomposition on the three-dimensional tensor using a Tucker tensor decomposition algorithm to obtain a core tensor and a factor matrix; fusing and reconstructing the core tensor and the factor matrix into the high-dimensional data matrix with space-time correlation characteristics; wherein the high-dimensional data matrix is used to represent the time sequence dependence and spatial coupling relationship between nodes of the power system.

3. The method of claim 2, wherein, the method of constructing an electrical connection relationship graph based on the configuration parameters of the power equipment using a graph neural network comprises the following steps: constructing an initial electrical graph with multiple key devices as nodes of a graph structure and the connection relationship between the key devices as edges of the graph structure; the categories of the key devices include circuit breakers, transformers, and busbars; extracting node features and edge weight information of the graph structure using a GraphSAGE algorithm in the graph neural network to identify implicit electrical dependence relationships; outputting an electrical connection relationship graph with connection weights, device category labels, and adjacency structures.

4. The method of claim 3, wherein, the method of inputting the high-dimensional data matrix and the electrical connection relationship graph into a quantum computing simulation module to obtain quantum state evolution results of the transient process of the power system comprises the following steps: transforming the space-time data obtained based on tensor decomposition and graph learning into a set of state variables and constructing a corresponding transient evolution differential equation set; equivalent representation of the transient evolution differential equation set as a Hamiltonian model in the form of a quantum circuit; calling a variational quantum eigenvalue solver to minimize the target energy function to simulate the quantum state evolution process of the power system under transient disturbance conditions; Quantum measurement operation extracts quantized eigenvalues including node voltage change, frequency fluctuation and phase angle offset as the output of the quantum state evolution result.

5. The method of claim 4, wherein, According to the quantum state evolution result, an electromagnetic-thermal-mechanical coupling field calculation engine is called to calculate the multi-physical field distribution parameters corresponding to the device, including: According to the quantum state evolution result, the physical quantities at the corresponding time node are extracted; the physical quantities include current density, node voltage and power flow information; The physical quantities are input into the electromagnetic-thermal-mechanical coupling field calculation engine to construct a multi-physical quantity control equation set; According to the multi-physical quantity control equation set, a joint partial differential equation system containing Maxwell equation, heat conduction equation and stress-strain equation is solved, and three-dimensional multi-physical field distribution parameters including Lorentz force distribution, temperature gradient field and mechanical stress tensor are output as the multi-physical field distribution parameters.

6. The method of claim 5, wherein, Based on the multi-physical field distribution parameters, molecular dynamics simulation and fluid mechanics model are combined to obtain arc dynamic characteristic data, including: The electron mean free path, collision cross section and ionization rate are estimated by molecular dynamics simulation method; The k-ε turbulence model and Navier-Stokes equation are combined to solve the arc plasma flow behavior, and three-dimensional dynamic vector data of arc shape evolution, core temperature distribution and ionization region range are output as the arc dynamic characteristic data.

7. The method of claim 6, wherein, The operator's electromyographic signal is collected, and the force feedback control instruction is generated through a neural network mapping model combined with the mechanical parameters of the device, including: The surface electromyographic signal of the radial wrist flexor muscle group of the operator is collected to obtain the collected signal; The collected signal is subjected to action recognition and intention classification based on a one-dimensional convolutional neural network to obtain the processed collected signal; The processed collected signal is combined with the mechanical characteristic curve of the circuit breaker to generate the corresponding damping force feedback signal waveform as the force feedback control instruction.

8. The method of claim 7, wherein, According to the eye tracking data, the arc dynamic characteristic data and the force feedback control instruction, a dynamic rendering engine is used to generate a virtual reality scene that is adaptive to the operator's visual response and operation state, including: The pupil diameter, gaze point position and saccade path are collected in real time based on an eye tracking system to identify the operator's visual attention area and alert state; The visual brightness, particle density and flicker frequency of the arc shape in the virtual scene are dynamically adjusted in combination with the arc dynamic characteristic data; The force feedback control instruction is received and mapped to the real-time deformation, interactive resistance and material response effect of the device in the virtual environment; The operator's visual attention area and alert state, the adjusted visual brightness, particle density and flicker frequency of the arc shape, and the real-time deformation, interactive resistance and material response effect of the device are fused by the dynamic rendering engine to generate an immersive three-dimensional scene image that meets the operator's current perception characteristics and operation intention as the virtual reality scene.

Images