Energy flow interface 3d interaction method

Abstract

The application discloses an energy flow interface 3D interaction method, and relates to the technical field of vehicle electronics, and comprises the following steps: collecting real-time running messages of each vehicle subsystem through a vehicle communication network and analyzing the real-time running messages into engineering values with physical dimensions; determining a vehicle running state based on the engineering values, dynamically generating an energy flow topology table defining source nodes and target nodes; calculating three-dimensional energy flux vectors containing direction vectors and modulus scalars according to node anchor point coordinates; calculating a target observation matrix of a virtual camera based on the three-dimensional energy flux vectors, so that the sight direction and the spatial direction of the vector satisfy a preset geometric projection relationship; driving a graphics rendering pipeline to generate an interactive interface containing three-dimensional component models and dynamic particle flows according to the engineering values and the target observation matrix. The application realizes adaptive adjustment of the visual angle by constructing the energy flux vector, ensures the best visibility of the energy transmission path, and realizes real-time visualization driven by physical data.

Description

Technical Field

[0001] This application relates to the field of vehicle electronics technology, and in particular to a 3D interaction method for energy flow interface. Background Technology

[0002] With the improvement of computing power in smart cockpit hardware, 3D visualization technology is gradually being applied to in-vehicle human-machine interfaces, especially for displaying the energy flow status of hybrid or pure electric vehicles. Existing energy flow interfaces typically use real-time rendered 3D models to reproduce the vehicle's powertrain system, including the engine, motor, battery, and transmission structure, aiming to allow users to intuitively understand the vehicle's working logic during driving, charging, or energy recovery processes.

[0003] However, in practical applications, existing 3D energy flow interaction technologies tend to prioritize high-fidelity models and rich animations in their visual presentation, resulting in an excessive number of visual elements on the interface. During driving, drivers have limited information processing capabilities; overly complex or disorganized 3D animations can easily create visual redundancy, increasing the driver's cognitive load, thus distracting attention and affecting driving safety. Current technologies struggle to strike a balance between displaying detailed vehicle status and maintaining a clean, low-interference interface, lacking mechanisms for proactively filtering and hierarchically presenting key information.

[0004] In terms of interaction logic, existing 3D interfaces typically employ a fixed perspective or rely on manual user input to change the viewing angle. When the energy transmission path changes or critical components are obscured by the vehicle body structure, users often need to adjust the perspective using complex gestures such as clicking, swiping, rotating, or zooming. This manual intervention-dependent interaction method is difficult to execute accurately while the vehicle is in motion, increasing not only the user's operational difficulty and learning cost but also preventing users from obtaining crucial vehicle status information in a timely manner. Existing display systems lack an intelligent mechanism that can automatically adjust the optimal observation pose based on real-time energy transmission logic and adaptively avoid occlusion. Summary of the Invention

[0005] The purpose of this invention is to provide a 3D interaction method for energy flow interfaces, which at least solves one of the technical problems of key components being obscured, difficult operation, and visual information overload caused by fixed viewing angles or reliance on manual interaction in existing vehicle-mounted 3D energy flow interfaces.

[0006] This invention provides the following solution:

[0007] A 3D interaction method for an energy flow interface, the method running on a cockpit domain controller connected to a vehicle communication network and a human-machine interface display terminal, the interaction method comprising:

[0008] The vehicle communication network collects real-time operation messages from each subsystem of the vehicle, and parses and normalizes the real-time operation messages to obtain engineering values ​​with physical dimensions.

[0009] Based on the engineering values, the vehicle operating status is determined, and an energy flow topology table is dynamically generated. The energy flow topology table defines the source node ID and target node ID of the energy transmission at the current moment.

[0010] Based on the source node ID and target node ID, the corresponding source node anchor coordinates and target node anchor coordinates are indexed in the virtual 3D scene. Vector subtraction and normalization operations are performed on the target node anchor coordinates and the source node anchor coordinates to calculate and generate a 3D energy flux vector. The 3D energy flux vector includes a normalized direction vector and a magnitude scalar.

[0011] The optimal observation pose calculation logic is executed, and the target observation matrix of the virtual camera is calculated based on the three-dimensional energy flux vector, so that the line of sight of the virtual camera and the spatial orientation of the three-dimensional energy flux vector satisfy a preset geometric projection relationship.

[0012] The driving graphics rendering pipeline generates an energy flow interactive interface containing three-dimensional component models and dynamic particle flows based on the engineering values ​​and the target observation matrix, and displays it through the human-computer interaction display terminal.

[0013] By adopting the above technical solution, the present invention transforms the vehicle's physical signals into geometric vectors in virtual space, and uses three-dimensional energy flux vectors to dynamically calculate the optimal observation angle, so that the line of sight always adapts to the current energy transmission path, thus solving the display blind spot problem under a fixed viewpoint; at the same time, it realizes real-time rendering driven by physical data, ensuring the consistency between the interface display and the actual operating state of the vehicle.

[0014] Preferably, the step of determining the vehicle's operating status based on the engineering values ​​and dynamically generating an energy flow topology table specifically includes:

[0015] The engineering values ​​are compared with preset physical thresholds, and continuous physical quantities are mapped to discrete vehicle operating state enumeration values.

[0016] When it is determined to be a pure electric drive state or a rapid acceleration high energy consumption state, the power battery model is marked as the source node, the drive motor model and the wheel hub model are marked as the destination node, and the active flags of the destination node and the connection path are set to true.

[0017] When the braking energy recovery state is determined, the wheel hub model and drive motor model are marked as source nodes, and the power battery model is marked as sink node;

[0018] When the state is determined to be hybrid drive, a multi-level transmission path is established from the engine model to the generator model, and then branches to the power battery model or drive motor model.

[0019] By adopting the above technical solution, the state machine logic is used to convert continuously changing physical signals into discrete operating conditions, which can cover multiple operating conditions such as pure electric, hybrid and energy recovery, and construct corresponding node connection relationships accordingly, so as to ensure that the energy path topology in the three-dimensional scene accurately reflects the current mechanical working principle of the vehicle.

[0020] Preferably, in the step of calculating and generating the three-dimensional energy flux vector, when there are multiple active paths involving hybrid power operation, the step of performing vector arbitration based on power weight and operation priority specifically includes:

[0021] Read the real-time transmission power value corresponding to each path, and convert the real-time transmission power value into a dimensionless weighting coefficient as the power amplitude weight;

[0022] The criticality weight of each path is determined based on a pre-set lookup table.

[0023] Calculate the weighted sum of the power magnitude weight and the criticality weight of each path as the comprehensive score;

[0024] The vector corresponding to the transmission path with the highest comprehensive score is selected as the principal energy flux vector, and the principal energy flux vector is used as the sole reference for the optimal observation pose calculation logic.

[0025] By adopting the above technical solution, in the case of multiple paths coexisting, the weighted algorithm automatically selects the transmission path with the highest current power or the most critical operating condition as the main perspective tracking target, avoiding disorderly switching of perspective between different paths and ensuring that the observation focus is always locked on the main energy flow path.

[0026] Preferably, the step of executing the optimal observation pose calculation logic specifically includes:

[0027] Calculate the midpoint coordinates of the line connecting the anchor point coordinates corresponding to the source node ID and the anchor point coordinates corresponding to the target node ID, and use it as the gaze point of the virtual camera;

[0028] Following the principle of maximum projection, the three-dimensional energy flux vector is projected onto the horizontal reference plane of the virtual world to obtain the horizontal component vector. The normal vector orthogonal to the horizontal component vector is calculated, and the target azimuth angle of the virtual camera is determined based on the normal vector.

[0029] By applying dynamic scaling logic, the distance between the virtual camera and the gaze point is set to be proportional to the magnitude of the three-dimensional energy flux vector, and the target observation matrix is ​​synthesized.

[0030] By adopting the above technical solution and utilizing the principles of geometric projection and orthogonal normals, the virtual camera is kept in a side observation position of energy flow, ensuring that the projection length of the transmission path on the screen is maximized; at the same time, the observation distance is adaptively adjusted according to the physical length of the transmission path to ensure that energy paths of different lengths can be fully presented within the field of view.

[0031] Preferably, before generating the target observation matrix, the method further includes performing line-of-sight occlusion detection and camera obstacle avoidance correction logic:

[0032] Construct a virtual ray pointing from the candidate position of the virtual camera to the gaze point, and divide the objects in the 3D scene into a target layer and an occlusion layer;

[0033] The ray casting interface of the physics engine is invoked to detect whether the virtual ray intersects with the mesh collider of the occlusion layer;

[0034] If an intersection is detected, an obstacle avoidance iterative algorithm is initiated. In a spherical coordinate system centered at the gaze point, a strategy of moving tangentially along the observation track is adopted to apply a preset incremental offset to the azimuth angle of the virtual camera until the line of sight is restored or the maximum adjustment threshold is reached.

[0035] By adopting the above technical solution and introducing a line-of-sight detection and position correction mechanism, the vehicle body can be automatically avoided without changing the gaze point through tangential search on a spherical track, thus solving the problem of invisible internal components in complex model structures.

[0036] Preferably, the method further includes performing viewpoint smoothing and anti-shake processing:

[0037] Receive the target observation matrix, and when the difference between the new target pose and the current stable state exceeds a preset dead zone threshold and remains stable within a preset time window, confirm that the pose change request is valid.

[0038] For the position coordinates of the virtual camera, calculate the difference vector between the current position and the target position, and use a nonlinear interpolation operation with a smoothing coefficient positively correlated with the magnitude of the difference vector; for the rotation quaternion of the virtual camera, use a spherical linear interpolation algorithm for interpolation update.

[0039] By adopting the above technical solution, dead zone determination is used to filter out the viewing angle tremor caused by high-frequency noise from the sensor, and nonlinear interpolation and spherical interpolation algorithms are used to ensure the speed uniformity and axial stability of the viewing angle switching process, thus avoiding the deadlock phenomenon caused by Euler angle rotation.

[0040] Preferably, the driving graphics rendering pipeline step includes performing component activity level determination and dynamic material rendering:

[0041] Components whose real-time physical parameters are greater than the activation threshold are marked as active nodes and assigned a first-level weight.

[0042] Mark other components as background nodes and assign them secondary or tertiary weights;

[0043] Sort the mesh objects in the scene, assign background nodes to the basic geometry queue, and assign active nodes to the highlight overlay queue;

[0044] When rendering the highlight overlay queue, the depth test function is modified to always pass, and depth writing is disabled, so that the pixel color of the active node overlays the background pixels in the frame buffer.

[0045] By adopting the above technical solution and adjusting the depth testing strategy in the graphics rendering pipeline, pixel-level coverage of background components by active components was achieved, enabling a visual effect of perspective observation of the internal dynamic structure without cutting the mesh model.

[0046] Preferably, the dynamic rendering of the material further includes transparency calculation and self-illumination modulation based on the Fresnel effect:

[0047] For the background nodes in the basic geometry queue, calculate the dot product of the fragment normal vector and the view vector. When the dot product result is close to zero, set a higher transparency coefficient, and when the dot product result is close to 1, set a lower transparency coefficient.

[0048] For the active node, the normalized power value is read and mapped to the gain coefficient of luminous intensity, while the saturation of the texture sampled color is increased.

[0049] By adopting the above technical solution, the transparency is controlled by utilizing the correlation between the surface normal and the viewing angle, so that the background object presents an effect where the outline is visible and the center is transparent, reducing the obstruction of the internal core components; at the same time, the power value is mapped to the material luminous intensity, realizing the visualization of energy strength.

[0050] Preferably, the step of driving the graphics rendering pipeline includes performing dynamic generation and behavior control of the particle system:

[0051] Extract the anchor coordinates corresponding to the source node ID as the path start point, and extract the anchor coordinates corresponding to the target node ID as the path end point;

[0052] Based on the relative positions of the starting point and the ending point and the preset curvature parameters, the intermediate control point is calculated, and a cubic Bézier curve is constructed as the particle motion trajectory.

[0053] When rendering each frame, the particle world coordinates are calculated by substituting the particle's life cycle parameters into the Bézier curve equation, and the tangent vector at the particle's location is calculated. The main axis direction of the particle texture is then aligned with the direction of the tangent vector.

[0054] By adopting the above technical solution, a smooth motion trajectory is constructed using cubic Bézier curves, and combined with tangent alignment calculations, the flow path of energy particles conforms to the laws of physical motion, avoiding the visual abruptness caused by straight-line connections.

[0055] Preferably, the dynamic generation of the particle system further includes performing particle visual attribute mapping:

[0056] Extract the corresponding color vector based on the energy type in the engineering values, and mix it with the base texture color;

[0057] The normalized power or current values ​​are mapped to the emission rate of the particle emitter and the texture flow velocity of the particles.

[0058] In the vertex shader, a scaling matrix is ​​applied to the particles, increasing the scaling factor when the engineering value is in a high-power state and decreasing the scaling factor when it is in a low-power state.

[0059] By adopting the above technical solution, a mapping relationship between physical signals and particle visual attributes is established. Energy types are distinguished by color, and power and current intensity are reflected by speed, density, and size, enabling users to intuitively perceive the current energy consumption status and working mode of the vehicle.

[0060] The above solution achieves the following beneficial technical effects:

[0061] This application achieves adaptive adjustment of the viewing angle by constructing a three-dimensional energy flux vector and performing optimal observation pose calculation. The system can calculate the direction and span of the energy transmission path in virtual space in real time and automatically adjust the position and angle of the virtual camera according to the maximum projection principle. This mechanism enables the display interface to automatically adapt to the current energy flow path under complex operating conditions such as hybrid power, ensuring that the energy flow path is always presented within the screen's field of view with the maximum projection length, thus solving the problem of key components being easily obscured or the path not being fully displayed under traditional fixed viewing angles.

[0062] This application establishes a precise mapping relationship between vehicle physical engineering values ​​and visual particle attributes. By parsing and normalizing real-time operational data such as current and power, these values ​​are used to directly drive the emission rate of the particle emitter, texture flow velocity, and geometric scaling factor. This data-driven rendering method enables the particle dynamics in the 3D scene to reflect the actual energy consumption and power changes of the vehicle in real time and quantitatively. Users do not need to interpret complex instrument data; they can determine the current load simply by observing the density and speed of the particle flow, thus improving the efficiency of information interaction.

[0063] This application employs a dynamic material rendering strategy based on component activity. The system divides components into active and background nodes according to real-time physical parameters. Transparency calculations based on the Fresnel effect are applied to the background nodes, and pixel coverage is forced during the rendering of active nodes by modifying the depth test function and disabling depth writing. This technical solution achieves perspective display of the internal power components of the vehicle body without physically cutting the model mesh, effectively highlighting the core components involved in energy transmission and ensuring the visual hierarchy and logical clarity of the complex mechanical structure. Attached Figure Description

[0064] Figure 1 This is a software logic architecture and data processing flowchart provided in one embodiment of the present invention. Detailed Implementation

[0065] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] This embodiment discloses a 3D interaction method for energy flow interface.

[0067] like Figure 1 The software logic architecture and data processing flowchart shown below:

[0068] The core of this interaction method includes a cockpit domain controller. As the system's computational hub, the cockpit domain controller integrates a high-performance system-on-a-chip (SoC). This high-performance SoC contains at least a multi-core CPU and an independent graphics processing unit (GPU). The CPU is configured to perform logical updates of the 3D scene graph, geometric calculations of energy flux vectors, and adaptive calibration algorithms for view frustum parameters. The GPU is configured to perform vertex shading, fragment processing, and rasterization operations. The memory is connected to the cockpit domain controller via a high-speed bus and includes non-volatile flash memory and random access memory (RAM). The non-volatile flash memory contains pre-loaded high-precision 3D model data of vehicle powertrain components, shader program code, and particle system configuration files.

[0069] This interaction method connects to the vehicle's communication network via an onboard communication interface. The onboard communication interface includes a CAN-FD transceiver or an onboard Ethernet physical layer interface. The cockpit domain controller subscribes to and reads real-time operation messages from various vehicle subsystems via the onboard Ethernet physical layer interface at a preset sampling frequency (e.g., 10ms to 50ms).

[0070] The subsystem controllers connected to the vehicle communication network include the battery management system, the motor controller, and the vehicle controller.

[0071] The battery management system is configured to collect cell status and output the real-time total voltage, total current, remaining charge percentage, and charge / discharge relay status flags of the power battery pack via the bus.

[0072] The motor controller is configured to acquire the motor status through a rotary transformer or encoder and output the real-time speed, rotor position, output torque magnitude and torque direction of the drive motor.

[0073] The vehicle controller is configured to collect driver operation signals and output the opening signal of the accelerator pedal position sensor, the brake pedal switch signal, and vehicle gear status information.

[0074] The interaction method also includes a human-machine interface display terminal. This terminal is electrically connected to the video output interface of the cockpit domain controller via a serial video link such as FPD-Link or GMSL, and is configured to receive the rendered video stream and display a 3D energy flow interface in real time. A capacitive touch sensing layer is attached to the surface of the display terminal. This sensing layer is connected to the cockpit domain controller via an I2C or SPI interface, and is configured to collect user touch operation commands such as clicks and drags, and transmit coordinate data to the cockpit domain controller. These hardware modules constitute a closed-loop hardware circuit from physical signal acquisition and logical operation processing to 3D visualization output.

[0075] The software logic architecture of this interaction method resides in memory and is loaded into the runtime space by the processor of the cockpit domain controller. Its logical structure is vertically divided into a data preprocessing layer, a core operation logic layer, and a rendering presentation layer.

[0076] The data preprocessing layer is configured to establish a mapping channel between physical signals and logical data, and to perform signal cleaning. The data preprocessing layer includes a protocol parsing module, which loads a pre-stored database file (such as a DBC or ARXML file), performs masking and shifting operations on the received hexadecimal raw bus messages, and converts them into engineering values ​​with physical dimensions, including current, voltage, speed, and pedal opening. The data preprocessing layer also includes a signal filtering module, configured to apply a moving average algorithm or a first-order low-pass filter algorithm to smooth the continuously sampled engineering values ​​to suppress high-frequency noise caused by sensor jitter and ensure the temporal stability of the input data for subsequent rendering pipelines.

[0077] The core operational logic layer is configured to implement the logical control of three-dimensional energy flow, and it includes a state machine management unit, a spatial vector calculation unit, and a view frustum control unit.

[0078] The state machine management unit is configured to execute multi-condition decision logic. It receives preprocessed engineering values ​​and maps continuous physical quantities to discrete vehicle operating state enumeration values ​​based on preset physical thresholds. For example, it determines the vehicle to be in a driving state only when the battery output current is greater than the discharge threshold and the motor torque is positive; it determines the vehicle to be in an energy recovery state when the motor torque is negative and the battery current is in the charging direction. Based on the decision results, the state machine management unit dynamically generates an energy flow topology table, which explicitly defines the source node ID and target node ID of the energy transmission at the current moment.

[0079] The spatial vector calculation unit is configured to perform vector operations in the world coordinate system of the 3D virtual scene. Based on the source node ID and target node ID, the spatial vector calculation unit indexes the geometric center coordinates or preset anchor point coordinates of the corresponding component model from the scene graph data structure. Based on the start and end coordinates, the spatial vector calculation unit calculates and generates a 3D energy flux vector. The 3D energy flux vector includes a normalized direction vector and a magnitude scalar; the direction vector represents the spatial direction of energy flow, and the magnitude scalar represents the Euclidean distance of the transmission path in virtual space. For hybrid power operations involving multi-path transmission, the spatial vector calculation unit performs weighted arbitration based on a preset power weight lookup table, selecting the transmission path with the highest power value as the primary guiding vector.

[0080] The view frustum control unit is configured to calculate the optimal pose matrix of the virtual camera based on the energy flux vector. The view frustum control unit reads the master guidance vector data and, in conjunction with preset camera trajectory constraint parameters (including minimum / maximum distance and pitch angle limits), performs a geometric solution. The solution algorithm follows the maximum projection principle, that is, calculating a camera position that maximizes the cross product magnitude of the view direction vector and the energy flux vector, thereby ensuring the longest possible projection length of the energy transmission path on the 2D screen plane. Furthermore, the view frustum control unit integrates occlusion detection logic based on ray casting, configured to detect whether the line connecting the camera and the target gaze point intersects with a non-transparent mesh, and trigger a position correction or material transparency command when occlusion is detected.

[0081] The rendering presentation layer is configured to receive output instructions from the core computational logic layer and drive the graphics rendering pipeline to generate the final frame. The rendering presentation layer includes a material control module and a particle system module.

[0082] The material control module dynamically modifies uniform variables in the shader, including diffuse color vector, emissivity intensity coefficient, and transparency alpha value, based on the vehicle's operating status and the activity level of the components, in order to achieve perspective or highlight effects.

[0083] The particle system module is configured to generate a dynamic particle flow on the Bézier curve between the source and target nodes. The particle system module maps the vehicle's physical dynamics parameters to the emission rate of the particle emitter and the texture flow velocity of the particles.

[0084] The rendering layer submits the computed vertex buffer objects and texture data to the graphics processing unit (GPU) by calling drawing instructions from the graphics application programming interface (API). The GPU performs rasterization and pixel shading, ultimately synthesizing a 3D energy flow interactive interface image in the frame buffer and outputting it through the video interface.

[0085] S1. The interaction method first executes the interaction function activation and initialization process.

[0086] S101. Monitoring and Determination of Interactive Activation Signals. The cockpit domain controller's processing unit continuously monitors touch event signals from the human-machine interface display terminal and vehicle status messages from the vehicle communication network. Interactive activation signals originate from user-initiated trigger commands and vehicle status change trigger commands. User-initiated trigger commands are generated when a user clicks on a specific icon (such as new energy management or energy flow) in the main menu interface of the display terminal. The processing unit parses the touch coordinate data, and when the touch point falls within the preset UI control response area, an activation command is generated. Vehicle status change trigger commands are generated by specific vehicle status change events, including detecting a charging connection confirmation signal being set, a driving mode switching request signal being activated, or the system detecting a battery thermal runaway warning flag flipping.

[0087] S102. Loading and instantiating 3D scene resources. In response to the generated activation command, the processing unit retrieves the 3D model data package of vehicle components from non-volatile memory. The 3D model data package contains vertex buffer data, index buffer data, and texture mapping data for each component (including but not limited to battery pack, motor, engine, wheel hub, and body frame). The processing unit asynchronously loads the above data into the video memory space of the graphics processing unit and instantiates each component object in the virtual world coordinate system according to the preset scene graph structure file. During this process, the processing unit parses the transformation matrix of each component object, maps the local coordinate system of each component to the world coordinate system, establishes its relative positional relationship in the virtual 3D space, and ensures that the topology of the virtual vehicle model is consistent with that of the physical vehicle.

[0088] S103. Initialize the rendering environment and set the default viewpoint. After the model is loaded, the system activates the graphics rendering pipeline. The processing unit initializes the virtual camera parameters and places the virtual camera at a preset initial viewpoint. This initial viewpoint is set to a top-down view of the vehicle model to ensure that the user can have a general overview of the vehicle even without energy flow data input. Simultaneously, the processing unit initializes the lighting environment in the scene, loading preset ambient occlusion maps and reflection probe data to provide basic physically based lighting effects for the component models. At this point, all component models are set to their default material state, typically a low-opacity semi-transparent material or a uniform metallic material, awaiting subsequent real-time data updates.

[0089] S104. Perform synchronization verification of the execution data channel. Before rendering the first frame of the 3D scene, the processing unit sends a handshake request to the vehicle communication network to confirm that the battery management system and motor controller are online and in normal communication status. If the verification passes, the system opens the high-frequency data subscription channel to prepare to receive real-time power system parameters; if the verification fails or the communication times out, the system calls the exception handling routine, loads a static schematic diagram or text prompt information on the interface, and terminates the 3D rendering thread to save computing resources.

[0090] S2. After initialization is complete, the real-time data processing and state mapping phase begins.

[0091] S201. The system performs raw physical signal acquisition and normalization. The arithmetic logic unit (ALU) reads real-time messages from the vehicle's powertrain system via the vehicle communication interface at preset time intervals (e.g., 10 to 50 milliseconds). Key parameters read include total battery pack current, total battery pack voltage, actual output torque of the drive motor, drive motor speed, engine speed, and accelerator pedal travel. The ALU parses the raw messages, removes signal bias, applies a scaling factor, and converts them to standard physical unit values. Subsequently, the system performs normalization, mapping the current, torque, and power values ​​to a floating-point range of 0.0 to 1.0. This normalized value will directly serve as the control input for the particle velocity factor and luminous intensity factor in subsequent rendering pipelines.

[0092] S202. Execute the state machine determination of the vehicle's operating condition. The arithmetic logic unit has multiple state determination logics pre-set, which determine the current energy flow state of the vehicle by comparing real-time physical values ​​with preset thresholds. Specifically, when the battery pack output current is detected to be greater than zero and the drive motor output torque is positive, the system determines it to be in pure electric drive mode. Based on this, the arithmetic logic unit further calculates the time derivative of the accelerator pedal travel value. When this rate of change exceeds a preset rapid acceleration threshold, or when the drive motor output torque exceeds a preset percentage of the rated peak torque, the system refines the state determination to be a rapid acceleration high-energy consumption state. When the drive motor output torque is detected to be negative and the battery pack current direction indicates charging input, the system determines it to be in regenerative braking mode. When the engine speed is detected to be greater than the starting threshold and the generator output power is greater than zero and supplies power to the battery pack or drive motor, the system determines it to be in hybrid drive mode or driving charging mode.

[0093] S203. Constructing and marking the energy transmission topology. Based on the operating conditions determined in step S202, the computational logic unit establishes dynamic connections in the component map in the virtual three-dimensional space, clarifying the source and destination nodes for energy transmission. In pure electric drive and rapid acceleration states, the system marks the power battery model as the source node, the drive motor model and the wheel hub model as destination nodes, and sets the active flag of these nodes and their connection paths to true. In regenerative braking state, the system marks the wheel hub model and the drive motor model as source nodes, and the power battery model as the destination node. In hybrid power state, the system establishes a multi-level transmission path from the engine model to the generator model, and then branches to the power battery model or the drive motor model. For components not participating in the current energy transmission process, the system sets their active flag to false.

[0094] S204. Encapsulation and output of rendering control parameters. The computational logic unit packages the logic state data generated in the above steps into a rendering instruction set. This rendering instruction set includes a node activation list, energy flow vector identifiers, and dynamic intensity coefficients. The dynamic intensity coefficients are scalar values ​​calculated based on the actual physical parameters of the vehicle. The rendering instruction set is transmitted to the rendering presentation layer for real-time updates of the material properties of objects in the scene graph and the behavior logic of particle emitters.

[0095] S3 maintains operation through a three-dimensional energy flow interactive closed-loop control logic.

[0096] S301. Perform frame synchronization and time step control. The graphics processing unit of the cockpit domain controller generates a vertical synchronization signal based on the hardware refresh rate of the display terminal. The main loop thread of the processing unit captures the vertical synchronization signal and establishes the time reference for a single rendering frame. The processing unit calculates the time increment from the previous frame to the current frame. This time increment serves as the numerical basis for subsequent physics simulation and interpolation calculations to ensure that the animation rate does not change with fluctuations in the system frame rate.

[0097] S302. Real-time acquisition of rendering state parameters. At the beginning of each frame, the main loop thread reads the latest instruction set output by the core computational logic layer from the shared memory region or data bus cache. The latest instruction set contains normalized power parameters, a list of active node states, and energy transfer path identifiers. The processing unit compares the current frame data with the previous frame data. If a change in state data is detected, the relevant nodes in the scene graph are marked as pending updates.

[0098] S303, Perform logical calculations and transformation updates for the scene objects. Step S303 is the core of closed-loop control; the processing unit recalculates the object attributes within the virtual scene based on the read state parameters. Specifically, it includes:

[0099] The target position and rotation quaternion of the virtual camera in the current frame are calculated based on the energy flux vector algorithm, and the actual rendering pose matrix of the current frame is calculated using the spherical linear interpolation algorithm.

[0100] Based on the list of active nodes, calculate the transparency coefficient and self-illumination intensity value of the material of each component model;

[0101] The simulated state of the particle system is updated based on the dynamic parameters, and the position coordinates of the particle emitter, the emission rate, and the lifetime parameters of the particles are calculated.

[0102] S304. Execute the submission and drawing instructions for graphics resources. The processing unit writes the model-view projection matrix, material attribute floating-point numbers, and particle vertex data calculated in step S303 into the constant buffer and vertex buffer of the graphics processing unit. Subsequently, the processing unit sends a drawing call instruction to the graphics rendering pipeline. The graphics processing unit performs vertex shading, rasterization, and fragment shading operations according to the instruction, projecting the 3D scene onto the 2D frame buffer. During this process, the system performs occlusion culling based on the depth information of the components to ensure the correct spatial hierarchy of the image.

[0103] S305. Perform frame buffer swapping and display output. After the graphics processing unit completes all rendering work for the current frame, the system performs a double buffer swapping operation, mapping the rendered back buffer data to the front buffer, and driving the LCD panel of the display terminal to update the screen through the video output interface.

[0104] S4. In order to achieve adaptive adjustment of the viewpoint, execute the energy flux vector construction and multi-path priority arbitration logic.

[0105] S401. Perform coordinate indexing and positioning of 3D spatial nodes. The spatial vector calculation unit accesses the scene graph database in memory. The scene graph database stores the spatial attribute information of each component object in the virtual vehicle model. For each component, the database has pre-set corresponding 3D anchor point coordinates. When the system receives the list of active nodes output from the state machine management unit, the spatial vector calculation unit retrieves and extracts the corresponding source node anchor point coordinates and target node anchor point coordinates from the database based on the component identifier.

[0106] S402. Perform geometric calculation of the initial energy flux vector. For each determined source node and target node pair, the computing unit performs vector subtraction. Specifically, the anchor coordinate vector of the target node is subtracted from the anchor coordinate vector of the source node to generate the original displacement vector from the source to the target. The computing unit then performs normalization on this original displacement vector to obtain a unit direction vector. Simultaneously, the computing unit retains the magnitude of the original vector, which represents the physical span of the energy transmission path in virtual space. This unit direction vector and the magnitude together constitute the basic definition of the energy flux vector.

[0107] S403. Perform multi-path concurrent conflict detection and pre-screening. The computation unit traverses all active paths in the current frame and constructs a candidate vector list. If only a single path exists in the list, it is directly marked as the main energy flux vector;

[0108] If multiple paths exist in the list, the multi-path arbitration subroutine is triggered.

[0109] S404. Perform vector arbitration based on power weight and operating condition priority. The multi-path arbitration subroutine introduces priority determination logic to select the primary viewpoint tracking target. The determination logic is based on two dimensions: power amplitude weight and operating condition criticality weight. In the power amplitude dimension, the system reads the real-time transmission power value corresponding to each path and converts it into a dimensionless weight coefficient. In the operating condition criticality dimension, the system has a pre-set lookup table that defines the visual attention level under different operating conditions. For example, the priority of the driving condition is set higher than that of the auxiliary power consumption condition, and the priority of the braking energy recovery condition is set higher than that of the idling charging condition. The calculation unit calculates the comprehensive score of each candidate vector, which is the weighted sum of the power weight and the criticality weight. The system compares the comprehensive scores of each candidate vector and selects the vector with the highest score as the primary energy flux vector for the current frame.

[0110] S405, Output and buffering of the primary vector. The primary energy flux vector determined through arbitration is written to the shared data bus for use by the frustum control unit. This output data includes the vector's start and end coordinates, as well as the normalized direction vector. For secondary paths not selected as primary vectors, the system only outputs their rendering instructions but does not use them as a reference for camera view calculations.

[0111] S5. Based on the determined principal energy flux vector, execute the optimal observation pose calculation logic.

[0112] S501. Calculate the visual focus center point. The frustum control unit reads the main energy flux vector data output by the spatial vector calculation unit and extracts its source node coordinates and target node coordinates. The control unit performs linear interpolation to calculate the midpoint coordinates of the line connecting the source node and the target node. These midpoint coordinates are set as the gaze point of the virtual camera to ensure that the energy transmission path is always located in the center area of ​​the display interface.

[0113] S502. Calculate the optimal horizontal observation azimuth. Step S502 follows the maximum projection principle, meaning that the projection length of the observed vector on the view plane reaches its maximum value when the line of sight is perpendicular to the observed vector. The control unit first projects the three-dimensional energy flux vector onto the horizontal reference plane of the virtual world to obtain the horizontal component vector. Subsequently, the control unit calculates the normal vector orthogonal to this horizontal component vector. This normal vector indicates the side observation direction relative to the energy flow direction. The control unit determines the target azimuth of the camera on the horizontal plane based on this normal vector.

[0114] S503. Perform constraint optimization for camera distance and pitch angle. Based on the determined horizontal azimuth angle, the control unit introduces distance and height constraints to construct the final three-dimensional coordinates. Regarding the distance constraint, the control unit calculates the magnitude of the energy flux vector. The system is configured with dynamic scaling logic to set the distance between the camera and the gaze point to be proportional to this magnitude. Regarding the height constraint, the control unit introduces a pitch angle limitation range (e.g., 15 degrees to 60 degrees). Regardless of the normal vector calculated in step S502, the control unit forcibly restricts the camera's height coordinates to a preset comfortable viewing range.

[0115] S504. Perform target observation matrix synthesis. Based on the gaze point coordinates determined in step S501, the camera position coordinates determined in steps S502 and S603, and the preset world up vector, the control unit performs observation matrix transformation operation. This operation constructs a 4×4 transformation matrix containing translation and rotation components. The transformation matrix defines the absolute position of the virtual camera in the world coordinate system, ensuring that the camera lens center is aligned with the midpoint of the energy path.

[0116] S6. To avoid visual obstruction, execute line-of-sight obstruction detection and camera obstacle avoidance correction logic.

[0117] S601. Construct and hierarchically filter the line-of-sight detection ray. The frustum control unit reads the candidate position coordinates of the virtual camera and the focus center coordinates of the energy flux vector. The control unit constructs a virtual ray starting from the candidate position coordinates and pointing to the focus center. To optimize detection efficiency and eliminate interference, the system hierarchically divides objects in the 3D scene. The system marks core power components such as batteries, motors, and wheel hubs as target layers, and non-core components such as the body frame, seat interiors, and chassis covers as occlusion layers. The line-of-sight detection ray is configured to only perform intersection tests with mesh colliders belonging to the occlusion layers.

[0118] S602. Perform raycasting and occlusion determination. The control unit calls the raycasting interface of the physics engine to emit a virtual ray and obtains collision feedback information between the ray and the scene geometry. This feedback information includes the Boolean state of the collision, the spatial coordinates of the collision point, and the object identifier to which the colliding object belongs. The control unit calculates the Euclidean distance from the candidate camera position to the collision point and compares it with the straight-line distance from the camera to the focus center. If a collision is detected and the former is less than the latter, the system determines that there is physical occlusion on the current line of sight.

[0119] S603. Perform camera position correction based on a spherical track. In response to the occlusion determination signal, the control unit initiates an obstacle avoidance iterative algorithm. The obstacle avoidance iterative algorithm employs a strategy of moving tangentially along the observation track. Specifically, the control unit applies a preset incremental offset to the camera's azimuth angle in a spherical coordinate system centered at the focus center point. The system recalculates the camera's Cartesian coordinates using the updated azimuth angle, while maintaining the radial distance between the camera and the center unchanged or making only minor adjustments.

[0120] S604. Execute the secondary verification and rollback strategy for the corrected position. For the corrected position generated in step 603, the system performs raycasting detection again. If the line of sight is restored, the corrected position is confirmed as the final target observation position and output.

[0121] If occlusion persists, the angle offset continues to accumulate until the preset maximum adjustment threshold is reached. If an unoccluded observation point cannot be found within the maximum adjustment threshold range, the system triggers a rollback rendering strategy, meaning the camera is no longer moved, and the occlusion problem is resolved by adjusting the material transparency.

[0122] S7. Performs smooth transitions and anti-shake processing of the viewing angle.

[0123] S701. Hysteresis determination of pose change request. The motion control module receives the target observation pose matrix. Internally, the module maintains a current stable state variable and a state hold timer. When the Euclidean distance or quaternion angle difference between the new target observation position and the current stable state exceeds a preset dead zone threshold, the system starts the state hold timer. The system only confirms the pose change request as valid if the new target pose remains stable within a consecutive preset time window or if the change amplitude is within the tolerance range.

[0124] S702. Perform nonlinear interpolation calculations for the position coordinates. Once the target observation pose is confirmed and updated, the system initiates the interpolation motion logic. For the camera's position coordinates, the motion control module employs either an exponential smoothing damping algorithm or a linear interpolation algorithm. In each rendered frame, the module calculates the difference vector between the current camera position and the target position and multiplies it by a smoothing coefficient related to the frame time interval. This smoothing coefficient is configured as a dynamic variable, positively correlated with the magnitude of the difference vector, achieving a nonlinear velocity response.

[0125] S703. Perform spherical linear interpolation calculation for rotation angle. For the camera's orientation parameter, i.e., the quaternion, the motion control module uses a spherical linear interpolation algorithm to calculate the rotation quaternion for the current frame. This algorithm interpolates the current quaternion and the target quaternion along a great circle path on a four-dimensional hypersphere, ensuring that the camera's angular velocity remains uniform during rotation and that the rotation axis remains stable, avoiding the gimbal lock problem caused by Euler angle interpolation.

[0126] S704. Perform clamping and locking of minute displacements. In the final stage of the interpolation motion, when the difference between the current pose and the target pose is less than a preset micro-threshold, the motion control module forcibly stops the interpolation calculation, directly assigns the current pose as the target pose, and marks the motion state as stationary to eliminate the minute jitter caused by floating-point calculation.

[0127] S8. In terms of visual presentation, the logic for determining the activity level of components is implemented.

[0128] S801, Performs threshold reading and preprocessing of real-time physical parameters. The rendering strategy module subscribes to real-time operating data of each subsystem from the data bus and reads preset activation threshold parameters from the configuration register. The module performs absolute value calculations on the real-time data to eliminate the influence of the current or torque direction sign on amplitude determination.

[0129] S802. Perform binarization determination of the active status of the components. The rendering strategy module traverses each functional node in the vehicle topology and compares its corresponding real-time physical parameters with the activation threshold. When the real-time output or input current amplitude of the power battery pack is greater than the current activation threshold, the system marks the power battery pack node as an active node and sets the active status flag to 1;

[0130] Conversely, if the node is active, it is marked as a background node, and the active state flag is set to 0. Similarly, the system marks the active states of components such as drive motors and internal combustion engines accordingly.

[0131] S803. Perform visual hierarchy weight allocation and mapping. After completing the basic binarization determination, the rendering strategy module further assigns rendering hierarchy weights to active nodes and background nodes. For components marked as active nodes, the module assigns a first-level weight, corresponding to high opacity and high saturation materials;

[0132] For components marked as background nodes, the module assigns a secondary or tertiary weight, corresponding to low opacity or wireframe mode.

[0133] S804, De-jitter filtering for state transitions. When the state of a component transitions from a background node to an active node, the system immediately responds and executes the state change;

[0134] When switching from an active node to a background node, the system starts a delay counter. The system will only officially revoke the active status of the node if the physical parameter remains below the activation threshold for a preset delay period.

[0135] S9. Based on the above classification, execute dynamic material rendering and visual enhancement logic.

[0136] S901. The rendering engine reorganizes the rendering queue and assigns material handles. After receiving a list of nodes containing active status flags, the rendering engine sorts and groups all mesh objects in the scene. The system establishes a basic geometry queue and a highlight overlay queue. Background nodes are assigned to the basic geometry queue and fitted with semi-transparent material handles;

[0137] Active nodes are assigned to the highlight overlay queue and have their entity specular material handles attached.

[0138] S902. Perform dynamic transparency calculation based on the Fresnel effect. For background nodes in the basic geometry queue, the fragment shader performs transparency calculations related to the view angle. The shader first performs a dot product operation: ;in, For the normalized virtual camera gaze vector, This is the normalized current fragment surface normal vector. Subsequently, the shader calculates the edge lighting coefficients using a simplified form of the Fresnel equation, as follows: ;in, For the transparency of the final output, It is a power function. The preset edge light index (e.g., 3.0 to 5.0). When the dot product result... When the line of sight is close to zero (i.e., the line of sight is perpendicular to the normal and at the edge of the object), the calculated value is... A higher value keeps the object's outline visible; when the dot product result... When the value is close to 1 (i.e., the line of sight is parallel to the normal and located at the center of the object), the calculated value is... The value is extremely low or close to zero, thus achieving an X-ray perspective visual effect with bright edges and a transparent center.

[0139] S903. Differentiated configuration for executing depth test rules. When rendering the basic geometry queue, the system enables depth writing to ensure correct occlusion relationships between background objects. When rendering the highlight overlay queue, the system modifies the depth test function to always pass or be greater than or equal to, and disables depth writing. This configuration ensures that the pixel color of active nodes forcibly overwrites the background pixels in the framebuffer, achieving logical perspective.

[0140] S904. Perform real-time modulation of emissivity intensity and color saturation. For active nodes, the system modulates the material's emissivity properties based on its actual physical power. The shader reads the normalized power value and maps it to a gain coefficient for emissivity intensity. Simultaneously, the system increases the saturation of the texture sampled colors for active nodes. Conversely, for background nodes, the system performs desaturation processing.

[0141] S905. Perform smooth inter-frame transitions for material properties. When a component's state switches between active and background, the rendering pipeline does not immediately hard-switch the material handle; instead, it initiates linear interpolation logic. The specific interpolation formula is as follows: ;in, This is the material property value for the current frame. The attribute value of the previous frame. The attribute value for the target state. The time increment between two frames. This is the preset interpolation rate coefficient. Under the control of this transition formula, the system smoothly changes the material's transparency, luminous intensity, and depth test mode.

[0142] S10, Execute the dynamic generation and behavior control logic of the energy flow particle system.

[0143] S1001. Initialize and pre-allocate memory for the particle object pool. The particle system module pre-builds the particle object pool during program startup. This object pool is a fixed-size circular queue, and each particle object contains a data structure containing its position coordinates, rotation angle, scaling factor, current lifecycle, and texture index.

[0144] S1002. Constructing the transmission path curve and calculating control points. For the main energy flux vector, the particle system module constructs the corresponding Bézier curve as the particle trajectory. The module extracts the geometric anchor points of the source node components as the path starting point. Extract the geometric anchor points of the target node components as the path endpoints. The module determines the tangent vector based on the relative positions of the start and end points and the direction of the surface normal. And using curvature parameters Calculate two intermediate control points and The formula is: Based on these four points, the module solves the path using cubic Bessel equations: ;in The interpolation factor is between 0 and 1.

[0145] S1003. Perform emission rate and attribute mapping calculation. The particle system module reads the real-time physical parameters corresponding to the current energy transmission path. The module inputs this physical value into the linear mapping function to calculate the number of particles to be activated in the current frame and the particle step velocity parameters. The system extracts the corresponding number of idle particles from the inactive list of the object pool and resets them to the active state.

[0146] S1004. Perform real-time updates and tangential alignment of particle poses. During the rendering of each frame, the particle system module performs parallel updates of the base light strip and discrete particles. For the base light strip, dynamic offset calculations of texture coordinates are performed. This achieves a visual effect of continuous flow. For discrete particles, their lifecycle parameters are updated: The module will be updated. Substitute the value into step S1002 The equation is solved to calculate the world coordinates of the particle in the current frame. The module further differentiates the curve equation to calculate the tangent vector at the particle's location. The system constructs a rotation matrix to align the principal axis of the particle texture to the direction of this tangent vector, ensuring that the particle always moves tangentially along the path.

[0147] S1005. Perform boundary detection and object recycling. When a parameter greater than or equal to 1.0 is detected, the system executes the recycling logic, sets the activity flag of the particle to invalid, and returns its data structure to the object pool.

[0148] S11, Execute particle visual attribute mapping and state response logic.

[0149] S1101. Perform energy type-based color space mapping. The material processing unit has a pre-set energy type color lookup table. When the system determines that the current energy flow is electrically driven, it extracts the cool-toned color vector;

[0150] When the output is determined to be mechanical energy, extract the warm-toned color vector;

[0151] When energy recovery is detected, a specific functional color vector is extracted. This color vector is then blended with the base texture color.

[0152] S1102. Execute shape scaling control based on power amplitude. The material processing unit implements dynamic scaling logic. The processing unit applies a scaling matrix to the particle vertex coordinates in the vertex shader. When the vehicle is in a high-power state, the scaling factor increases;

[0153] When in a low-power state, the scaling factor decreases. Furthermore, this scaling logic is coupled with the particle's lifecycle parameters, causing the particle to exhibit a tapered geometry at both ends.

[0154] S1103. Perform a sequence frame animation update of texture coordinates. The material processing unit updates the texture coordinate offset of each particle according to the system time increment. During rendering, the shader samples different sub-regions in the texture atlas according to the texture coordinate offset to achieve dynamic lighting effects inside the particles.

[0155] S1104. Smooth transition processing when the energy flow direction is reversed. When the energy flow direction is reversed, the material processing unit activates the dual-channel mixing mechanism. The system first locks all surviving particles from the previous state, forcing their transparency channels to decay rapidly to zero over time. Simultaneously, the system immediately activates the particle emitter for the new state. During the transition window, forward-facing particles gradually disappear, while reverse-facing particles gradually generate, achieving a smooth transition.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A 3D interactive method for energy flow interfaces, characterized in that, The interaction method operates on a cockpit domain controller, which is connected to the vehicle communication network and a human-machine interface display terminal. The interaction method includes the following steps: The vehicle communication network collects real-time operation messages from each subsystem of the vehicle, and parses and normalizes the real-time operation messages to obtain engineering values ​​with physical dimensions. Based on the engineering values, the vehicle operating status is determined, and an energy flow topology table is dynamically generated. The energy flow topology table defines the source node ID and target node ID of the energy transmission at the current moment. Based on the source node ID and target node ID, the corresponding source node anchor coordinates and target node anchor coordinates are indexed in the virtual 3D scene. Vector subtraction and normalization operations are performed on the target node anchor coordinates and the source node anchor coordinates to calculate and generate a 3D energy flux vector. The 3D energy flux vector includes a normalized direction vector and a magnitude scalar. The optimal observation pose calculation logic is executed, and the target observation matrix of the virtual camera is calculated based on the three-dimensional energy flux vector, so that the line of sight of the virtual camera and the spatial orientation of the three-dimensional energy flux vector satisfy a preset geometric projection relationship. The driving graphics rendering pipeline generates an energy flow interactive interface containing three-dimensional component models and dynamic particle flows based on the engineering values ​​and the target observation matrix, and displays it through the human-computer interaction display terminal.

2. The energy flow interface 3D interaction method according to claim 1, characterized in that, The steps of determining vehicle operating status based on the aforementioned engineering values ​​and dynamically generating an energy flow topology table specifically include: The engineering values ​​are compared with preset physical thresholds, and continuous physical quantities are mapped to discrete vehicle operating state enumeration values. When it is determined to be a pure electric drive state or a rapid acceleration high energy consumption state, the power battery model is marked as the source node, the drive motor model and the wheel hub model are marked as the destination node, and the active flags of the destination node and the connection path are set to true. When the braking energy recovery state is determined, the wheel hub model and drive motor model are marked as source nodes, and the power battery model is marked as sink node; When the state is determined to be hybrid drive, a multi-level transmission path is established from the engine model to the generator model, and then branches to the power battery model or drive motor model.

3. The energy flow interface 3D interaction method according to claim 1, characterized in that, In the step of calculating and generating the three-dimensional energy flux vector, when there are multiple active paths involving hybrid power operation, the step of performing vector arbitration based on power weight and operation priority specifically includes: Read the real-time transmission power value corresponding to each path, and convert the real-time transmission power value into a dimensionless weighting coefficient as the power amplitude weight; The criticality weight of each path is determined based on a pre-set lookup table. Calculate the weighted sum of the power magnitude weight and the criticality weight of each path as the comprehensive score; The vector corresponding to the transmission path with the highest comprehensive score is selected as the principal energy flux vector, and the principal energy flux vector is used as the sole reference for the optimal observation pose calculation logic.

4. The energy flow interface 3D interaction method according to claim 1, characterized in that, The steps for executing the optimal observation pose calculation logic specifically include: Calculate the midpoint coordinates of the line connecting the anchor point coordinates corresponding to the source node ID and the anchor point coordinates corresponding to the target node ID, and use it as the gaze point of the virtual camera; Following the principle of maximum projection, the three-dimensional energy flux vector is projected onto the horizontal reference plane of the virtual world to obtain the horizontal component vector. The normal vector orthogonal to the horizontal component vector is calculated, and the target azimuth angle of the virtual camera is determined based on the normal vector. By applying dynamic scaling logic, the distance between the virtual camera and the gaze point is set to be proportional to the magnitude of the three-dimensional energy flux vector, and the target observation matrix is ​​synthesized.

5. The energy flow interface 3D interaction method according to claim 4, characterized in that, Before generating the target observation matrix, the logic for line-of-sight occlusion detection and camera obstacle avoidance correction is also included: Construct a virtual ray pointing from the candidate position of the virtual camera to the gaze point, and divide the objects in the 3D scene into a target layer and an occlusion layer; The ray casting interface of the physics engine is invoked to detect whether the virtual ray intersects with the mesh collider of the occlusion layer; If an intersection is detected, an obstacle avoidance iterative algorithm is initiated. In a spherical coordinate system with the gaze point as the center, a strategy of moving tangentially along the observation track is adopted to apply a preset incremental offset to the azimuth angle of the virtual camera until the line of sight is restored or the maximum adjustment threshold is reached.

6. The energy flow interface 3D interaction method according to claim 1, characterized in that, The method also includes performing viewpoint smoothing and anti-shake processing: Receive the target observation matrix, and when the difference between the new target pose and the current stable state exceeds a preset dead zone threshold and remains stable within a preset time window, confirm that the pose change request is valid. For the position coordinates of the virtual camera, calculate the difference vector between the current position and the target position, and use a nonlinear interpolation operation with a smoothing coefficient positively correlated with the magnitude of the difference vector; The rotation quaternion of the virtual camera is updated using a spherical linear interpolation algorithm.

7. The energy flow interface 3D interaction method according to claim 1, characterized in that, The steps of driving the graphics rendering pipeline include performing component activity level determination and dynamic material rendering: Components whose real-time physical parameters are greater than the activation threshold are marked as active nodes and assigned a first-level weight. Mark other components as background nodes and assign them secondary or tertiary weights; Sort the mesh objects in the scene, assign background nodes to the basic geometry queue, and assign active nodes to the highlight overlay queue; When rendering the highlight overlay queue, the depth test function is modified to always pass, and depth writing is disabled, so that the pixel color of the active node overlays the background pixels in the frame buffer.

8. The energy flow interface 3D interaction method according to claim 7, characterized in that, The dynamic rendering of materials also includes transparency calculation and self-illumination modulation based on the Fresnel effect: For the background nodes in the basic geometry queue, calculate the dot product of the fragment normal vector and the view vector. When the dot product result is close to zero, set a higher transparency coefficient, and when the dot product result is close to 1, set a lower transparency coefficient. For the active node, the normalized power value is read and mapped to the gain coefficient of luminous intensity, while the saturation of the texture sampled color is increased.

9. The energy flow interface 3D interaction method according to claim 1, characterized in that, The driving graphics rendering pipeline steps include performing dynamic generation and behavior control of particle systems: Extract the anchor coordinates corresponding to the source node ID as the path start point, and extract the anchor coordinates corresponding to the target node ID as the path end point; Based on the relative positions of the starting point and the ending point and the preset curvature parameters, the intermediate control point is calculated, and a cubic Bézier curve is constructed as the particle motion trajectory. When rendering each frame, the particle world coordinates are calculated by substituting the particle's life cycle parameters into the Bézier curve equation, and the tangent vector at the particle's location is calculated. The main axis direction of the particle texture is then aligned with the direction of the tangent vector.

10. A 3D interactive method for an energy flow interface according to claim 9, characterized in that, The dynamic generation of the particle system also includes performing particle visual attribute mapping: Extract the corresponding color vector based on the energy type in the engineering values, and mix it with the base texture color; The normalized power or current values ​​are mapped to the emission rate of the particle emitter and the texture flow velocity of the particles. In the vertex shader, a scaling matrix is ​​applied to the particles, increasing the scaling factor when the engineering value is in a high-power state and decreasing the scaling factor when it is in a low-power state.

Images