Semi-physical simulation system and method for tracked vehicle driving training

Abstract

The invention provides a semi-physical simulation system and method for tracked vehicle driving training. The system comprises a computer, a control and input device, a motion feedback device and an immersive display and positioning device. The computer is used for resolving the kinetic model and rendering a virtual scene; the control and input device is used for collecting a control instruction of a user; the motion feedback device is used for receiving and presenting the posture change of the virtual vehicle in real time and providing somatosensory feedback; the immersive display and positioning device comprises virtual reality glasses, a matched positioning base station and a tracker. The technical barriers that traditional professional simulation software is closed in environment and difficult to be deeply integrated with a high-immersion VR system are broken through, and a more open and extensible solution is provided for high-performance simulation of the tracked vehicle.

Description

Technical Field

[0001] This invention relates to the field of virtual simulation technology, and in particular to a hardware-in-the-loop simulation system and method for training tracked vehicle drivers. Background Technology

[0002] Entering the 21st century, with the rapid development of computer graphics and real-time rendering technologies, traditional simulation technologies, due to their insufficient immersion, have become unable to meet the growing demands, driving 3D virtual simulation technology to become a key tool for industry and research. Against this backdrop, the Unity engine, with its powerful cross-platform capabilities, open ecosystem, and continuously optimized performance, has become one of the most widely used 3D development platforms globally.

[0003] Tracked running gear is widely used in special vehicles in fields such as construction, training, and mining due to its high passability and mobility. In-depth research into its dynamics and handling response characteristics is of great significance for improving the reliability and passability of vehicles under complex working conditions.

[0004] Currently, simulation studies of tracked locomotives largely rely on specialized multibody dynamics software such as DADS, ADAMS / ATV, and RecurDyn. While these tools excel in the precise theoretical analysis of the mechanisms themselves, they have limitations in handling complex environmental interactions, multi-rigid-flexible body coupling simulations, and meeting the demands for high real-time performance and highly immersive visualization and interaction. Consequently, they are difficult to directly apply to real-time simulation scenarios requiring a "human-in-the-loop" presence, such as driver training and operational condition pre-simulation.

[0005] While the Unity engine has significantly improved computational performance through technologies such as ECS and Burst Compiler, and achieved highly realistic rendering effects with the help of technologies like HDRP and Shader Graph, its native physics engine is not designed for complex multibody systems like tracked vehicles. Therefore, Unity lacks a dedicated model and solution capable of accurately describing the complex dynamic coupling relationships between tracks, drive wheels, road wheels, guide wheels, and the ground. This technological gap prevents the full utilization of Unity's high market share and powerful rendering capabilities to build a tracked vehicle simulation training system that combines high dynamic accuracy with a high level of immersion. Summary of the Invention

[0006] Purpose of the invention: The technical problem to be solved by the present invention is to provide a hardware-in-the-loop simulation system and method for tracked vehicle driving training, which addresses the shortcomings of existing technologies. This system and method resolves the technical contradiction that existing professional simulation software cannot meet the requirements of highly immersive real-time interactive training, while the Unity platform lacks a high-precision tracked vehicle-specific dynamic model. The invention achieves integrated simulation of tracked vehicles in complex virtual working conditions, from dynamic calculation to visual-physical feedback.

[0007] The system includes a computer, a control and input device, a motion feedback device, and an immersive display and positioning device;

[0008] The computer is used to run simulation software and to solve dynamic models and render virtual scenes;

[0009] The control and input device is a semi-physical driver's console, including a steering wheel, gear shift, brake, and accelerator. The control and input device is connected to a computer via a Universal Serial Bus (USB) to collect user control commands.

[0010] The motion feedback device includes a six-degree-of-freedom platform, which is connected to a computer via a network cable to receive and present the posture changes of the virtual vehicle in real time, providing haptic feedback.

[0011] The immersive display and positioning device includes virtual reality (VR) glasses and a matching positioning base station and tracker. The VR glasses are connected to a computer via a DisplayPort (DP) cable and a Universal Serial Bus (USB) cable to display virtual scenes. The positioning base station tracks the user's head position and posture in real time and updates the virtual viewpoint via optical positioning and Bluetooth communication. The tracker tracks the position and posture of the user's torso and limbs in real time via inertial positioning and Bluetooth communication, and the data is synchronously mapped into the virtual world.

[0012] The system's workflow includes:

[0013] Command issuance and scenario initialization: The user selects the training scenario through the command and control software, and the simulation software loads the corresponding virtual environment and tracked vehicle model;

[0014] Real-time simulation of human in the loop: The trainer wears virtual reality (VR) glasses and operates a semi-physical control console, and the control commands are collected in real time and input into the computer;

[0015] The core model for solution: The simulation software calculates the multibody dynamics coupling model system of the track mechanism in real time based on control commands and the virtual environment;

[0016] A multi-channel synchronous feedback mechanism is implemented, including a somatosensory feedback channel, a visual feedback channel, and a directional tactile feedback channel.

[0017] The somatosensory feedback channel includes: the system calculates the high-frequency vehicle motion state, and drives the six-degree-of-freedom platform in real time to reproduce the corresponding physical motion through a high-precision motion mapping algorithm and a low-latency communication protocol, and adds plastic flow correction on the cap yield surface of the Drucker-Prager criterion.

[0018] The visual feedback channel includes: based on real-time updated vehicle dynamics and attitude data, driving the rendering engine to generate corresponding scene images, introducing an adaptive radial blur post-processing algorithm, dynamically adjusting the rendering effect in the Unity (Unity Engine, Unity) High Definition Render Pipeline (HDRP) virtual simulation platform according to the calculated vehicle speed and acceleration, and presenting realistic motion blur that conforms to human high-speed visual perception in virtual reality VR glasses;

[0019] The adaptive radial blur post-processing algorithm includes: firstly, calculating the radial blur vector of the circular field of view region according to the following formula:

[0020]

[0021] in For radial fuzzy vectors, The vector of the center point of the field of view. R is the screen pixel vector, and R is the radial blur radius;

[0022] The formula used is: R = k1v + k2|a|, where k1 and k2 are weighting coefficients, and v and a are the vehicle's speed and acceleration, respectively.

[0023] Change The field-of-view center vector value can simulate the fuzzy center of the velocity direction, as shown in the formula:

[0024]

[0025] in, Let K be the vector of the screen center point, K be the offset matrix, and θ be the vehicle's attitude matrix, expressed as:

[0026]

[0027] Where, k y k is the yaw-to-horizontal offset coefficient. p k is the pitch-to-vertical offset coefficient. r k′ is the roll-off horizontal offset coefficient. r , where is the roll-to-vertical offset coefficient, pitch is the vehicle's pitch angle, yaw is the vehicle's yaw angle, and roll is the vehicle's roll angle.

[0028] Using the initial sampling formula in Let i.texcoord be the initial sampled screen pixels, and i.texcoord be the normalized texture vector. The k-th sample is... in For the k-th sampled screen pixel, For the radial blur vector, the formula for calculating the accumulated color value A is:

[0029]

[0030] Where Iteration is the total number of sampling iterations. The value of Iteration needs to be balanced between effect quality and performance overhead. Low quality Iteration = 4 to 8, medium quality Iteration = 8 to 16, and high quality Iteration = 16 to 32.

[0031] T is a bilinear sampling function, expressed as:

[0032] T(u, v) = (1-t) x (1-t) y )Tex(x0+y0)+t x (1-t y )Tex(x1+y0)+(1-t x )t y Tex(x0+y1)+t x t y Tex(x1+y1),

[0033] Where Tex is the discrete pixel matrix of the texture, x0 and y0 are the x and y coordinates of the bottom left pixel of the screen, respectively, and x1 and y1 are the x and y coordinates of the top right pixel of the screen, respectively. x , t y It is the interpolation factor for the sampling point in a 2×2 pixel grid.

[0034] The directional tactile feedback channel includes: adopting a directional tactile feedback mechanism, the system simulates continuous acceleration inertia by retracting the electric seat belt according to longitudinal acceleration, and simulates emergency braking inertia by servo push rods on the seat back, providing directional force stimulation to the user's chest and back on a static base in sync with vision and body sensation, forming a multi-dimensional sensory closed loop;

[0035] The inertial force on the human body is calculated based on the vehicle's acceleration, and the tension F of the seatbelt is obtained from this inertial force. b :

[0036]

[0037] Where m is the occupant mass, θ is the angle between the seat belt direction and the direction of motion, and μ is the frictional force between the human body and the seat belt;

[0038] A spring-damper model is used to describe the force F generated by the backrest push rod pressing against the occupant. c :

[0039]

[0040] Where k s c is the equivalent stiffness between the occupant and the backrest. s For contact damping, x sb Let x represent the displacement of the backrest, and let x represent the displacement of the occupant's chest and back relative to the vehicle body. The speed at which the backrest moves. For the occupant's chest and back speed;

[0041] The contribution ratio of the two channels is dynamically allocated based on the magnitude and direction of vehicle acceleration:

[0042] w b (a)+w c (a) = 1,

[0043] Where w b (a) represents the force weight of the seatbelt aisle, w c (a) represents the force weight of the backrest push rod channel, calculated using the following formula:

[0044]

[0045] The present invention also provides a hardware-in-the-loop simulation method for tracked vehicle driving training based on the aforementioned system, comprising the following steps:

[0046] Step 1, System Initialization and Command Acquisition:

[0047] Build a virtual prototype of a tracked vehicle in Unity and continuously receive control commands from a semi-physical driver's console;

[0048] Step 2: Solve the multibody dynamics coupling model system of the track mechanism and update the tension that drives the track movement;

[0049] To simulate power input, torque is directly applied to the drive wheel. To establish the drive wheel-track coupling relationship, the traction force F obtained by the track is calculated according to the following formula. t :

[0050] F t =M k / r k ;

[0051] Among them, F t M is the pulling force that drives the tracks to move the drive wheels.k For the driving wheel torque, r k The driving radius of the drive wheel;

[0052] Step 3, Ground Interaction Modeling: Calculate the tangential reaction force of the ground on the tracks using the following formula. and normal stress F R :

[0053]

[0054] in, The tangential reaction force exerted by the ground on the tracks, where f is the ground deformation resistance coefficient, F R F is the normal stress exerted by the ground on the track. d The equivalent driving force on the road wheel, The unit vector of the sliding velocity of the track contact point relative to the ground. The projection vector of the forward vector onto the ground tangent plane. The projection vector of the rightward vector onto the ground tangent plane;

[0055]

[0056] Among them, the function F R Let b be the normal stress exerted by the ground on the track, and k be the track width. d R is the ground modulus. rw Let be the radius of the road wheel, and s be the amount of sinkage of the road wheel. For angular coordinates, The angle of contact wheel arc is n, and the pressure-sinking index is n.

[0057] Incorporating the plastic flow correction on the yield surface of the Drucker-Prager cap criterion, the corrected normal stress... for:

[0058]

[0059] Where f pl It is the plastic flow factor. The Drucker-Prager yield function value;

[0060] Corrected tangential reaction force for:

[0061]

[0062] Step 4: Update the position of the road wheels based on the ground normal stress. Then, by establishing a suspension model, calculate the vertical force F exerted by the suspension on the vehicle body according to the following formula.z :

[0063]

[0064] Among them, F z The force applied to the suspension, k s Let z be the equivalent torsional stiffness of the suspension, z be the vertical position of the road wheel, c be the equivalent damping constant of the suspension, t be the simulation step size, and δ be the increment; update the vertical position of the vehicle body.

[0065] Step 5: Simulate internal system losses to drive the vehicle.

[0066] Step 5 includes: To simulate internal system losses, calculate the sliding friction force F between the road wheel and the track according to the following formula. f The sliding friction force F between the guide wheel and the track id :

[0067] F f =signμ|F R |,

[0068] Among them, F f F is the frictional force between the road wheel and the track, μ is the coefficient of friction between the road wheel and the track, |F R | represents the normal force acting on the contact surface between the road wheel and the track. The parameter sign is taken as ±1 according to the direction of the friction force.

[0069] F id =(T0-T7)(R id +δ t / 2) / R id ,

[0070] Among them, F id R is the frictional force between the guide wheel and the track, T0 and T are the tension forces around the two sides of the guide wheel, and R is the tension force between the guide wheel and the track. id Let δ be the radius of the guide wheel. t Track plate thickness;

[0071] Finally, by combining all the resultant forces acting on the tracks, the final traction force acting on the vehicle body is calculated, driving the vehicle to move.

[0072] The present invention also provides an electronic device, including a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method.

[0073] The present invention also provides a storage medium storing a computer program or instructions that, when the computer program or instructions are run on a computer, execute the steps of the method described.

[0074] Beneficial Effects: A breakthrough and integrated innovation in high-end simulation technology has been achieved through platform-based approaches. Deep integration of core dynamic aspects such as track chain drive, wheel-track coupling, and high-order ground interaction within the Unity platform has created a high-fidelity simulation solution covering closed-loop vehicle-ground interaction. This breaks down the technical barriers of traditional professional simulation software environments being closed and difficult to integrate deeply with highly immersive VR systems, providing a more open and scalable solution for high-performance simulation of tracked vehicles.

[0075] Achieving a harmonious balance between high model fidelity and strong real-time performance: Through the reasonable simplification of complex physical models (such as the Drucker-Prager cap model) and the innovative application of Unity's underlying physical components (such as HingeJoint), high-fidelity simulation of the complex dynamics of tracked systems, especially the continuous deformation and failure behavior of soil, was realized on consumer-grade hardware, successfully resolving the inherent contradiction between engineering-grade accuracy and real-time rendering frame rate.

[0076] A closed-loop training paradigm with deep multi-sensory coupling was constructed: high-precision dynamic models, semi-physical interactive hardware, and VR rendering pipelines are deeply integrated. Through a real-time closed loop of "command input - dynamic calculation - multimodal feedback," an immersive training environment with highly synchronized visual, auditory, and tactile senses is created. This not only greatly enhances the realism and immersion of training but also provides a safe, controllable, and efficient standardized platform for tracked vehicle driving training under complex and high-risk conditions. Attached Figure Description

[0077] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0078] Figure 1 This is the system architecture diagram of the present invention.

[0079] Figure 2 This is a flowchart of the method of the present invention.

[0080] Figure 3 This is a simulation effect diagram of the present invention under virtual terrain.

[0081] Figure 4 It is a platform that simulates different weather effects. Detailed Implementation

[0082] This invention provides a hardware-in-the-loop simulation system for tracked vehicle driving training, such as... Figure 1 As shown, this system is not simply a collection of hardware components, but rather deeply integrates virtual simulation and physical feedback through a specific dataflow design and integration architecture. The system mainly includes:

[0083] The core of the simulation calculation is a high-performance computer that runs simulation software developed based on the method of this invention, and is responsible for all dynamic model solving and virtual scene rendering.

[0084] Control and input devices: a semi-physical driver's console, including a steering wheel, gear shift, brake, accelerator, etc., which is connected to a computer via USB to collect user control commands.

[0085] Motion feedback device: A six-degrees-of-freedom platform connected to a computer via a gigabit network cable, used to receive and present the virtual vehicle's posture changes in real time, providing haptic feedback.

[0086] Immersive display and positioning devices: VR glasses and supporting positioning base stations and trackers. The VR glasses connect to the computer via DP and USB cables to display virtual scenes; the positioning base stations use optical positioning and Bluetooth communication to track the user's head position and posture in real time and update the virtual viewpoint.

[0087] The innovative integration workflow of this system is as follows:

[0088] Command issuance and scenario initialization: The user selects the training scenario through the command and control software, and the simulation software then loads the corresponding virtual environment and tracked vehicle model.

[0089] Real-time human-in-the-loop simulation: The trainee wears VR glasses and operates a hardware-in-the-loop control system. Their control commands (such as steering and throttle) are collected in real time and input into the simulation calculation core.

[0090] Core model solution: The multibody dynamics coupling model system of the track mechanism in the simulation software (see Example 2) performs real-time solution based on control commands and the virtual environment. By introducing the "modified Drucker-Prager cap constitutive model", this system can simulate complex mechanical behaviors such as soil shear yielding and compaction hardening with high accuracy, thereby outputting key dynamic states (such as acceleration, pitch angle, and roll angle) that can truly reflect the interaction between the vehicle and the ground.

[0091] Multi-channel synchronous feedback mechanism:

[0092] The motion feedback channel: The system uses high-frequency vehicle motion states calculated from the data, along with a high-precision motion mapping algorithm and a low-latency communication protocol, to drive the six-degree-of-freedom platform in real time to reproduce the corresponding physical motion. In particular, the "modified Drucker-Prager cap model" integrated into the dynamics kernel provides the platform with real underlying data from high-order ground mechanics feedback by accurately simulating the elastoplastic deformation of the soil, significantly improving the realism of the motion feedback for complex terrains such as bumps and sinkholes.

[0093] Visual Feedback Channel: Based on real-time updated vehicle dynamics and attitude data, the rendering engine generates corresponding scene images. To compensate for the limitations of the platform's physical motion range, an "adaptive radial blur" post-processing algorithm is specially introduced. According to the calculated vehicle speed and acceleration, the rendering effect is dynamically adjusted in the Unity HDRP pipeline, presenting realistic motion blur in VR glasses that conforms to human high-speed visual perception, enhancing the sense of speed and immersion at the visual level.

[0094] First, the radial blur vector of the circular field of view region is calculated using the following formula.

[0095]

[0096] in For radial fuzzy vectors, The vector of the center point of the field of view. Let R be the screen pixel vector, and let R be the radial blur radius. The blur radius implies the blur intensity. If a constant is used, it will lead to distortion under dynamic blur at different speeds. Therefore, we can use the formula R = k1v + k2|a|, where k1 and k2 are weighting coefficients, and v and a are the vehicle's speed and acceleration. Similarly, changing... The field-of-view center vector value can be used to simulate the fuzzy center of the velocity direction, as shown in the following formula:

[0097]

[0098] Where K is the offset matrix and θ is the vehicle's attitude matrix. This is the vector of the center point of the screen.

[0099]

[0100] Where, k y k is the yaw-to-horizontal offset coefficient. p k is the pitch-to-vertical offset coefficient. r k is the roll-off horizontal offset coefficient. ′ r This is the vertical offset coefficient for the roll pair.

[0101] Using the initial sampling formula Sampling screen pixels, the k-th sample is The accumulated color value is then expressed as follows:

[0102]

[0103] Where T is the bilinear sampling function, and its expression is as follows:

[0104] T(u, v) = (1-t) x(1-t) y )Tex(x0+y0)+t x (1-t y )Tex(x1+y0)+(1-t x )t y Tex(x0+y1)+t x t y Tex(x1+y1)

[0105] Where Tex is the discrete pixel matrix of the texture, representing the return of a four-dimensional color vector value located at integer pixel coordinates.

[0106] Directional haptic feedback channel: Innovatively adopting a "directional haptic feedback" mechanism as a supplement to platform motion. Based on longitudinal acceleration, the system simulates continuous acceleration inertia through the retraction of the electric seat belt and simulates emergency braking inertia through the servo push rod of the seat back, providing the user's chest and back with directional force stimulation synchronized with vision and body sensation on a static base, forming a multi-dimensional sensory closed loop.

[0107] The inertial force on the human body can be calculated from the vehicle's acceleration, and the tension F of the seatbelt can be deduced from this inertial force. b As shown in the following formula:

[0108]

[0109] Where m is the occupant mass, a is the vehicle acceleration, θ is the angle between the seat belt direction and the direction of motion, and μ is the frictional force between the human body and the seat belt.

[0110] At the same time, a spring-damper model is used to describe the force generated by the backrest push rod pressing against the occupant.

[0111]

[0112] Where k s c is the equivalent stiffness between the occupant and the backrest. s For contact damping, x sb Let x represent the displacement of the backrest (end of the push rod), and let x represent the displacement of the occupant's chest and back relative to the vehicle body. The speed at which the backrest moves. The velocity is the speed of the occupant's chest and back.

[0113] In order for the seat belt tension and backrest thrust to realistically simulate the longitudinal inertia of a vehicle, the contribution ratio of the two channels must be dynamically allocated according to the magnitude and direction of the vehicle's acceleration, as shown in the following formula.

[0114] w b (a)+w c (a) = 1

[0115] a is the vehicle acceleration, wb (a) represents the force weight of the seatbelt aisle, w c (a) represents the force weights for the backrest push rod channel. The backrest push rod simulates the force acting on the back, and the seatbelt simulates the force acting on the chest; therefore, the two weights can be automatically given directly from the force ratio:

[0116]

[0117] Through the highly coordinated and low-latency closed-loop data flow of the aforementioned multi-channel synchronous feedback mechanism, this system successfully solves the technical challenge of synchronizing the virtual environment with physical feedback, achieving truly immersive training.

[0118] This embodiment also provides a hardware-in-the-loop simulation method for tracked vehicle driving training based on the aforementioned system. This embodiment focuses on the specific implementation method of building and running the "multibody dynamics coupling model system" in Unity. For example... Figure 2 As shown in the flowchart, the key to this method lies in transforming the theoretical model into concrete components and logic that Unity can execute efficiently.

[0119] S1: System Initialization and Command Acquisition

[0120] Build a virtual prototype of a tracked vehicle in Unity and continuously receive control commands (accelerator, brake, steering wheel angle, etc.) from a semi-physical driver's console.

[0121] S2-S8: Solving and updating the core dynamics model.

[0122] This process is a concentrated embodiment of the model system, with each step interconnected and working synergistically:

[0123] Power transmission modeling (S2-S3): Torque is directly applied to the drive wheel to simulate power input. To establish the drive wheel-track coupling relationship, according to formula (1) F t =M k / r k Calculate the traction force obtained by the tracks. This step converts a simple torque input into the source of power for the system's motion.

[0124] Ground interaction modeling (S4): This is the core of the simulation accuracy. To construct the road wheel-track-ground coupling model, formula (2) is used. and (3) Calculate the tangential reaction force of the ground on the tracks respectively. and normal stress F R .

[0125] Traditional calculations struggle to quantify the interaction between the vehicle and complex terrain, so a plastic flow correction based on the Drucker-Prager cap yield surface was incorporated into the theory above. The corrected normal stress... As shown in the following formula:

[0126]

[0127] Where f pl It is the plastic flow factor. This represents the Drucker–Prager yield function value. The corrected tangential reaction force. As shown in the following formula:

[0128]

[0129] Vehicle attitude update modeling (S5-S6): Update the position of the road wheels based on the ground normal stress. Subsequently, by establishing a suspension model, according to formula (4) The vertical forces exerted by the suspension on the vehicle body are calculated, and the vertical position of the vehicle body is updated accordingly. This realizes a complete dynamic chain from ground excitation to vehicle body response.

[0130] Modeling of internal friction and final driving force (S7-S8): To simulate the internal losses of the system, according to formula (5) F f =signμ|F R | and (6)F id =(T0-T)(R id +δ t / 2) / R id Calculate the sliding friction force F between the road roller, guide roller and track. f and F id Finally, by combining all the resultant forces acting on the tracks, the final traction force acting on the vehicle body is calculated, driving the vehicle to move.

[0131] This embodiment also provides a hardware-in-the-loop simulation method based on the aforementioned system for training tracked vehicle drivers under different terrain obstacles and simulated weather conditions. This embodiment focuses on illustrating the effects of tracked vehicles under different terrain obstacles and varying weather conditions, including sunlight and rain. Figure 3 , Figure 4 As shown, this illustrates the specific implementation of multi-channel feedback to the driver based on changes in vehicle attitude and speed.

[0132] Figure 3 The first image shows a simulation of a tracked vehicle "crossing over a protruding rock with one side of its road wheel".

[0133] Step 1: Simulation initialization;

[0134] Road surface model: dry hard sandy land, soil subsidence of 0.001m, pressure-subsidence index of 1.0, Drucker-Prager yield function of 200kPa, and plastic flow factor of 50kPa.

[0135] Stone model: A rigid ellipsoid, 0.4m long, 0.3m wide, and 0.25m high, partially buried in the ground. Its surface contacts the track plates using a high-rigidity elastic contact model.

[0136] Vehicle parameters: vehicle mass 50t, track width 0.5m, road wheel radius 0.5m, vehicle travels in a straight line at a constant speed of 20km / h.

[0137] Step 2: Real-time dynamics solution;

[0138] The vehicle was traveling in a straight line when its left front road wheel came into contact with a rock. The rock, acting as a rigid obstacle, caused instantaneous deformation of the road wheel's contact envelope. The rock was 0.4m long, but the road wheel only made contact with a small section of the top of the track, with a contact area of ​​approximately 0.04m². 2 At the boundary of the rock, the soil will undergo strong plastic deformation, and its equivalent shear resistance will increase significantly. Therefore, according to local correction, the effective bearing capacity of the remaining soil is often significantly reduced, approaching 0 locally around the rock and approaching 0.75 far away from the rock.

[0139] Due to the severe asymmetry of the impact force, the vehicle generates a rolling moment about the longitudinal axis and a yaw moment about the vertical axis. The roll rate instantaneously reaches 15° / s to the right, the yaw rate generates a small leftward disturbance of 2° / s, and the roll rate reaches its peak in about 0.2–0.3 seconds, and then rapidly decays under the action of suspension damping.

[0140] Step 3: Multi-channel collaborative feedback;

[0141] Somatosensory feedback channel: At the moment of impact, the six-degree-of-freedom platform quickly rolls to the right (simulating tilt) based on the instantaneous impact and roll angular velocity, while simultaneously performing a sharp vertical lift to simulate the combined sensation of the wheel "hitting" a stone and lifting the vehicle body.

[0142] Visual feedback channel: The adaptive radial blur maintains its basic intensity as the vehicle speed remains constant. However, due to instantaneous yaw disturbance, the blur direction undergoes a slight dynamic change. The image rendered by the VR glasses tilts synchronously to the lower right as the platform moves and then immediately returns to center, perfectly matching the tactile sensation.

[0143] Directional tactile feedback channel: Because the main directions of local collision are vertical and lateral, the peak longitudinal acceleration is insufficient, so the seat belt and backrest push rod do not activate.

[0144] Figure 4To simulate different weather effects on the platform, we now combine... Figure 3 The second image illustrates a simulation scenario of a tracked vehicle traversing a long, slippery trench. The background shows the vehicle needing to cross a muddy trench formed by continuous rainfall in a straight line. The soil inside the trench is saturated and loose, and the sidewalls are slippery. The objective is to cross the trench in the optimal manner while maintaining directional stability, avoiding getting stuck or skidding.

[0145] Step 1: Simulation initialization;

[0146] Road surface model: A long, muddy ditch formed by continuous rainfall, 3.5m wide and 0.8m deep, with a soil subsidence of 0.03m, a pressure-subsidence index of 0.6, a Drucker-Prager yield function of 40kPa, and a plastic flow factor of 30kPa.

[0147] Weather conditions: Initially sunny, followed by moderate to heavy rain after 1 hour, with a rainfall intensity of 30 mm / h, lasting for more than 30 minutes.

[0148] Vehicle parameters: vehicle mass 50t, vehicle length 9m, track width 0.5m, road wheel radius 0.5m, vehicle travels in a straight line at a constant speed of 10km / h.

[0149] Step 2: Real-time dynamics solution;

[0150] During the entry phase (front end sinking), the peak instantaneous normal force at the front of the track is close to 500-600 kN (considering the additional load caused by inertial impact and mud compression), slightly higher than the static load of the entire vehicle, but consistent with the peak characteristics of impact dynamics. Due to the sudden drop in soil strength, despite the torque provided by the engine, the effective traction force generated by the front track is severely insufficient. The actual tangential force is much lower than the dry ground level, causing the track to slip, with a speed drop of approximately 2 km / h. When the front end sinks into the trench while the rear is still on higher ground, the vehicle will experience significant forward pitch, with a short-term pitch angle reaching 1°.

[0151] During the exit phase (front end raised), the driver maintains a low and stable speed. The first pair of road wheels at the front of the vehicle first contact and climb onto the hard surface opposite the trench, while the middle and rear of the vehicle remain in the muddy trench. The actual short-term peak normal force at the front of the tracks may reach 500-600kN. While maintaining speed, the short-term pitch angle can reach 1°.

[0152] Step 3: Multi-channel collaborative feedback;

[0153] 1. Entering the driving phase;

[0154] Somatosensory feedback channel: Simulates front wheel impact and sinking, platform tilts forward 1°, simulates the sinking posture of the front of the car.

[0155] Visual feedback channel: The VR camera's tilt angle is synchronously adjusted to 1°, and the viewpoint rotates downwards, so the driver sees the hood lifting and the ground approaching. The blur effect automatically adjusts according to the vehicle's deceleration, reducing the forward flow speed to give the driver a clear "sense of deceleration," with the blur intensity increasing by approximately 15% based on the deceleration value.

[0156] Directional tactile feedback: At the moment of descent, the seatbelt contracts rapidly with a force of 150-200N for 0.5 seconds, simulating inertial forward restraint. The backrest push bar does not move (the longitudinal acceleration changes little during this phase).

[0157] 2. Departure phase;

[0158] Somatosensory feedback channel: The platform performs a "lift-tilt" compound motion to simulate the front wheels climbing hard ground, and the platform tilts back 1° to simulate the front of the car tilting up.

[0159] Visual feedback channel: The VR camera's tilt angle is synchronously adjusted to 1°, the viewpoint rotates upwards, and the driver sees the hood lower and the sky appear larger. Adaptive radial blur adjusts according to the stable speed, the blur intensity maintains the base value, and the blur center shifts slightly upwards (simulating the change in visual focus when looking up).

[0160] Somatosensory feedback channel: When the front of the vehicle is climbing, the longitudinal acceleration shows a slight backward change, so the seat belt relaxes momentarily for 0.3 seconds, giving the driver a natural "backward" inertial feeling; during the stabilization phase, the seat belt returns to moderate tension, simulating the stable acceleration environment after the vehicle's posture is restored.

[0161] The hardware-in-the-loop simulation method of this embodiment has the following significant advantages compared to existing driving simulation technologies:

[0162] 1. High-precision vehicle dynamics simulation;

[0163] This scheme uses the Drucker–Prager yield surface model combined with plastic flow factor correction to perform high-precision modeling of the local interaction between the road wheels of tracked vehicles and the ground.

[0164] It can accurately simulate the peak normal force, pitch angle, roll rate and vertical acceleration changes during the collision of a single-sided load-bearing wheel with a protruding rock and the sinking / climbing process in a muddy ditch, improving the accuracy of simulation dynamics and adapting to various complex terrains and different weather conditions.

[0165] 2. Multi-channel feedback and coordination;

[0166] This solution constructs a three-channel collaborative feedback system integrating somatosensory, visual, and directional tactile feedback:

[0167] A six-degree-of-freedom platform simulates pitch, roll, and vertical impact.

[0168] VR vision synchronously adjusts the pitch angle and adaptive radial blur to achieve speed and acceleration perception;

[0169] The seat belt and backrest provide tactile feedback response to longitudinal acceleration, thereby restraining inertial forward and backward movement.

[0170] The highly consistent feedback across the three channels provides the driver with an immersive experience that closely resembles driving a real vehicle.

[0171] 3. Adaptability to complex weather and terrain;

[0172] It can simulate vehicle dynamics changes in real time under conditions such as rainfall, slippery muddy ditches, etc., including front end sinking, track slippage, speed fluctuations and pitch angle changes.

[0173] Drivers can train in a variety of complex environments to improve their ability to cope with different road conditions and weather changes.

[0174] 4. Excellent driver training results;

[0175] High-precision simulation and multi-channel synchronous feedback enable drivers to experience the dynamic behavior of vehicles in complex terrain and inclement weather in advance.

[0176] It significantly improves the efficiency of skill transfer, safety awareness, and emergency response capabilities, and the training effect is superior to traditional single-channel or simplified dynamics simulation systems.

[0177] This embodiment also provides a computer program product, such as an executable program stored on a USB flash drive, hard disk, or server. When the program is executed by a processor, it can implement all the steps of the method.

[0178] This invention provides a hardware-in-the-loop simulation system and method for tracked vehicle driving training. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A hardware-in-the-loop simulation system for tracked vehicle driving training, characterized in that, Includes computers, control and input devices, motion feedback devices, and immersive display and positioning devices; The computer is used to run simulation software and to solve dynamic models and render virtual scenes; The control and input device is a semi-physical driver's console, including a steering wheel, gear shift, brake, and accelerator. The control and input device is connected to a computer via a universal serial bus to collect user control commands. The motion feedback device includes a six-degree-of-freedom platform, which is connected to a computer via a network cable to receive and present the posture changes of the virtual vehicle in real time, providing haptic feedback. The immersive display and positioning device includes virtual reality (VR) glasses and a matching positioning base station and tracker. The VR glasses are connected to a computer via a display port cable and a universal serial bus (USB) cable to display virtual scenes. The positioning base station tracks the user's head position and posture in real time and updates the virtual viewpoint via optical positioning and Bluetooth communication. The tracker tracks the position and posture of the user's torso and limbs in real time via inertial positioning and Bluetooth communication, and the data is synchronously mapped into the virtual world.

2. The system according to claim 1, characterized in that, The system's workflow includes: Command issuance and scenario initialization: The user selects the training scenario through the command and control software, and the simulation software loads the corresponding virtual environment and tracked vehicle model; Real-time simulation of human in the loop: The trainer wears virtual reality (VR) glasses and operates a semi-physical control console, and the control commands are collected in real time and input into the computer; The core model for solution: The simulation software calculates the multibody dynamics coupling model system of the track mechanism in real time based on control commands and the virtual environment; A multi-channel synchronous feedback mechanism is implemented, including a somatosensory feedback channel, a visual feedback channel, and a directional tactile feedback channel.

3. The system according to claim 2, characterized in that, The somatosensory feedback channel includes: the system calculates the high-frequency vehicle motion state, and through motion mapping algorithm and low-latency communication protocol, drives the six-degree-of-freedom platform in real time to reproduce the corresponding physical motion, and adds plastic flow correction on the yield surface of the Drucker-Prager cap.

4. The system according to claim 3, characterized in that, The visual feedback channel includes: based on real-time updated vehicle dynamics and attitude data, driving the rendering engine to generate corresponding scene images, introducing an adaptive radial blur post-processing algorithm, dynamically adjusting the rendering effect in the Unity High Precision Rendering Pipeline (HDRP) of the virtual simulation platform according to the calculated vehicle speed and acceleration, and presenting realistic motion blur in virtual reality (VR) glasses that conforms to human high-speed visual perception.

5. The system according to claim 4, characterized in that, The adaptive radial blur post-processing algorithm includes: firstly, calculating the radial blur vector of the circular field of view region according to the following formula: in For radial fuzzy vectors, The vector of the center point of the field of view. R is the screen pixel vector, and R is the radial blur radius; The formula used is: R = k1v + k2|a|, where k1 and k2 are weighting coefficients, and v and a are the vehicle's speed and acceleration, respectively. Change The field-of-view center vector value can simulate the fuzzy center of the velocity direction, as shown in the formula: in, Let K be the vector of the screen center point, K be the offset matrix, and θ be the vehicle's attitude matrix, expressed as: Where, k y k is the yaw-to-horizontal offset coefficient. p k is the pitch-to-vertical offset coefficient. r k′ is the roll-off horizontal offset coefficient. r , where is the roll-to-vertical offset coefficient, pitch is the vehicle's pitch angle, yaw is the vehicle's yaw angle, and roll is the vehicle's roll angle. Using the initial sampling formula in Let i.texcoord be the initial sampled screen pixels, and i.texcoord be the normalized texture vector. The k-th sample is... in For the k-th sampled screen pixel, For the radial blur vector, the formula for calculating the accumulated color value A is: Where Iteration is the total number of sampling iterations; T is a bilinear sampling function, expressed as: T(u,v)=(1-t x )(1-t y )Tex(x0+y0)+t x (1-t y )Tex(x1+y0)+(1-t x )t y Tex(x0+y1)+t x t y Tex(x1+y1), Where Tex is the discrete pixel matrix of the texture, x0 and y0 are the x and y coordinates of the bottom left pixel of the screen, respectively, and x1 and y1 are the x and y coordinates of the top right pixel of the screen, respectively. x , t y It is the interpolation factor for the sampling point in a 2×2 pixel grid.

6. The system according to claim 5, characterized in that, The directional tactile feedback channel includes: adopting a directional tactile feedback mechanism, the system simulates continuous acceleration inertia by retracting the electric seat belt according to longitudinal acceleration, and simulates emergency braking inertia by servo push rods on the seat back, providing directional force stimulation to the user's chest and back on a static base in sync with vision and body sensation, forming a multi-dimensional sensory closed loop; The inertial force on the human body is calculated based on the vehicle's acceleration, and the tension F of the seatbelt is obtained from this inertial force. b : Where m is the occupant mass, θ is the angle between the seat belt direction and the direction of motion, and μ is the frictional force between the human body and the seat belt; A spring-damper model is used to describe the force F generated by the backrest push rod pressing against the occupant. c : Where k s c is the equivalent stiffness between the occupant and the backrest. s For contact damping, x sb Let x represent the displacement of the backrest, and let x represent the displacement of the occupant's chest and back relative to the vehicle body. The speed at which the backrest moves. For the occupant's chest and back speed; The contribution ratio of the two channels is dynamically allocated based on the magnitude and direction of vehicle acceleration: w b (a)+w c (a)=1, Where w b (a) represents the force weight of the seatbelt aisle, w c (a) represents the force weight of the backrest push rod channel, calculated using the following formula:

7. A hardware-in-the-loop simulation method for tracked vehicle driving training based on the system described in any one of claims 1 to 6, characterized in that, Includes the following steps: Step 1, System Initialization and Command Acquisition: Build a virtual prototype of a tracked vehicle in Unity and continuously receive control commands from a semi-physical driver's console; Step 2: Solve the multibody dynamics coupling model system of the track mechanism and update the tension that drives the track movement; To simulate power input, torque is directly applied to the drive wheel. To establish the drive wheel-track coupling relationship, the traction force F obtained by the track is calculated according to the following formula. t : F t =M k / r k ; Among them, F t M is the pulling force that drives the tracks to move the drive wheels. k For the driving wheel torque, r k The driving radius of the drive wheel; Step 3, Ground Interaction Modeling: Calculate the tangential reaction force of the ground on the tracks using the following formula. and normal stress F R : in, The tangential reaction force exerted by the ground on the tracks, where f is the ground deformation resistance coefficient, F R F is the normal stress exerted by the ground on the track. d The equivalent driving force on the road wheel, The unit vector of the sliding velocity of the track contact point relative to the ground. The projection vector of the forward vector onto the ground tangent plane. The projection vector of the rightward vector onto the ground tangent plane; Among them, the function F R Let b be the normal stress exerted by the ground on the track, and k be the track width. d R is the ground modulus. rw Let be the radius of the road wheel, and s be the amount of sinkage of the road wheel. For angular coordinates, The angle of contact wheel arc is n, and the pressure-sinking index is n. Incorporating the plastic flow correction on the yield surface of the Drucker-Prager cap criterion, the corrected normal stress... for: Where f pl It is the plastic flow factor. The Drucker-Prager yield function value; Corrected tangential reaction force for: Step 4: Update the position of the road wheels based on the ground normal stress. Then, by establishing a suspension model, calculate the vertical force F exerted by the suspension on the vehicle body according to the following formula. z : Among them, F z The force applied to the suspension, k s Let z be the equivalent torsional stiffness of the suspension, z be the vertical position of the road wheel, c be the equivalent damping constant of the suspension, t be the simulation step size, and δ be the increment; update the vertical position of the vehicle body. Step 5: Simulate internal system losses to drive the vehicle.

8. The method according to claim 7, characterized in that, Step 5 includes: To simulate internal system losses, calculate the sliding friction force F between the road wheel and the track according to the following formula. f The sliding friction force F between the guide wheel and the track id : F f =signμ|F R |, Among them, F f F is the frictional force between the road wheel and the track, μ is the coefficient of friction between the road wheel and the track, |F R | represents the normal force acting on the contact surface between the road wheel and the track. The parameter sign is taken as ±1 according to the direction of the friction force. Among them, F id For the friction between the guide wheel and the track, T0 and For the tension force around both sides of the guide wheel, R id Let δ be the radius of the guide wheel. t Track plate thickness; Finally, by combining all the resultant forces acting on the tracks, the final traction force acting on the vehicle body is calculated, driving the vehicle to move.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method as described in claim 7.

10. A storage medium, characterized in that, It stores a computer program or instructions that, when run on a computer, perform the steps of the method as described in claim 7.

Images