1. Introduction The emergence of virtual prototyping and virtual reality technologies has improved traditional automotive design and testing methods. Digital cars designed using virtual prototyping technology are combined with complex and ever-changing virtual testing environments. Real drivers simulate driving, engaging the user through sight, sound, and touch, creating an immersive and interactive experience. The human, car, and environment become one, allowing direct experience of vibrations, tilting, and noise. Extreme tests such as collisions and rollovers can be conducted without danger or damage. This improves upon abstract numerical curve simulations, overcoming the difficulties of expressing complex driver sensations and reactions using mathematical models. It allows for the timely identification of potential problems in the early stages of design, enabling adjustments, corrections, and optimization. Advantages include cost savings, repeatability, and risk-free operation. Typical examples include Chrysler reducing the development cycle of its new car from 36 months to 24 months, and Ford announcing the development of a fully digital sedan at the end of 1999. Mercedes-Benz had already completed a digital car prototype before 1998 and achieved strong virtual reality technology, enabling intuitive and comprehensive simulation analysis, evaluation, and improvement of the overall performance matching and body system layout design of the car during the design phase. Therefore, based on extensive research, this paper designs a virtual automotive testing system, using handling stability testing as an example. Comparison between actual and virtual tests shows that virtual testing achieves better results than traditional simulation data processing, integrating the human-vehicle-environment system and providing a novel method for studying vehicle closed-loop systems. A detailed introduction follows. 2. System Structure Design The automotive virtual testing system is a complex system. Its design adheres to the principles of advancement, openness, and reliability, based on object-oriented design principles, and adopts a hierarchical, modular, and standardized design approach to achieve structural rationality while maintaining functionality. The automotive virtual testing system mainly consists of three parts: an input module, a virtual testing module, and an output module, as shown in Figure 1. The process involves importing a car model into a virtual environment. Based on user input and control commands, the system performs kinematic and dynamic analyses on the car model. Using this analysis data, the car testing process is virtually recreated in a virtual scene. Users experience the car's performance through various sensors, obtaining performance evaluations. Model parameters are then modified based on these evaluations. This process can be repeated continuously, modifying car parameters until the car achieves optimal performance. A brief description follows: 1. Input Module: This module primarily provides input information such as the vehicle and the parametric ground model. Real-world test data can also be input for comparison with the virtual test. 2. Virtual Test Module: This module mainly consists of a dynamics analysis module, an interface module, and a virtual test operation module. The dynamics analysis module calculates the vehicle's performance parameters in real time based on the parametric model and driver behavior, providing simulation data for the virtual test module. The virtual test operation module provides a realistic virtual test environment, realizing various state changes of the virtual vehicle during the test and feeding this information back to the user. The interface module organically connects the real-time vehicle simulation calculation, the virtual operating environment, sensor interface devices, and the user into a unified whole. 3. Performance Evaluation Module: In this module, users comprehensively consider various sensory experiences under different operating conditions during virtual testing to evaluate vehicle performance. Based on the evaluation results, the parameters of the vehicle model are modified for optimization. This method overcomes the shortcomings of previous methods, such as difficulty in establishing evaluation standards, difficulty in determining indicators, and poor visualization. This module also allows for comparison between virtual and real vehicle tests, and the vehicle model can be corrected based on the differences. 3. Development Example of a Virtual Testing System The authors developed a virtual test for handling stability as an example. First, a digital virtual prototype of the car was created using ADAMS/Car software, and handling stability simulation analysis was performed on it. The virtual testing environment was implemented using WTK (WorldToolKit) and Visual C++, and the virtual instrumentation was implemented using NI's ComponentWorks components. The simulation data from the dynamic analysis was transmitted to the virtual environment to realistically and vividly realize the test, and the virtual instruments accurately displayed the changes in the data. Human-computer interaction technology was used, allowing users not only to have a visual experience but also to immerse themselves in the virtual testing environment through the AGC—Viewer G 3D observation system, truly experiencing the car's performance and thus evaluating its performance. The following explanation uses the single lane change test in the closed-loop test of vehicle handling stability as an example. Using ADAMS/Car, the authors established a digital model of a car and designed a control file based on the single lane change test method (such as trajectory and speed). During the vehicle dynamics simulation, this file controls the vehicle to travel along a specified route. Then, the main characteristic parameters of handling stability (such as steering wheel angle, yaw rate, roll angle, and lateral acceleration) are extracted from the simulation analysis results. This data is then used in conjunction with the single lane change test scenario to dynamically observe the vehicle itself and the changes in parameters. To allow users to accurately know the specific parameter changes during the car's movement, a virtual instrument function was added. This virtual instrument, combined with the virtual scene, dynamically displays the changes in various parameters of the car during the simulation. To achieve a more realistic sense of immersion, a dual-viewport stereo scene display effect was developed. Through the AGC-ViewerG stereo observation system, users can immerse themselves in the virtual test scene and feel and experience the test process. Figure 2 shows the dynamics simulation analysis results (speed 100km/h), and Figure 3 shows the simulation scene and virtual instrument changes during the virtual single lane change test. (a) Lateral displacement change[/align] (b) Steering wheel angle, yaw rate, and lateral acceleration change[/align] Figure 3 Simulation scene and virtual instrument changes during virtual single lane change test[/align] 4. Summary With the application and development of virtual prototypes and virtual reality technology in the automotive industry, significant progress has been made in automotive design and testing research. Using multibody system dynamics analysis software to obtain vehicle performance parameters and virtually realize vehicle performance testing systems can fully realize "indoor virtual testing," representing an extension of future virtual prototype systems. By connecting some control devices and combining them with virtual scenes, users can feel the vibration, lateral slip, and tilt of the car, incorporating subjective human perception into the evaluation of vehicle performance, which has strong practical significance. With the development of computer, image technology, sensor, and other technologies, automotive virtual testing technology has a promising future. References [1] Wang Chengwei, Gao Wen, Wang Xingren. Theory, Implementation and Application of Virtual Reality Technology. Beijing: Tsinghua University Press, 1996. [2] Yang Baomin, Zhu Yining. Distributed Virtual Reality Technology and Its Application. Beijing: Science Press, 2000. [3] Wang Shufeng, Yu Qun. Implementation of Vehicle Virtual Test System. Journal of Agricultural Machinery, 2002, 33(3), 4-7. [4] Xiong Jian, Zeng Jiguo, Song Jian. Research on Virtual Simulation of Automobile Handling Stability. Automotive Engineering, 2002, 24(5): 430-433.