Application Research of a V2X Semi-Physical Simulation Platform in Vehicle–Road Collaboration Experiments

Abstract

As a core application of V2X technology, vehicle–road collaboration enables dynamic coordination among road users (pedestrians, vehicles), infrastructure, and networks through real-time, omnidirectional information exchange. This system represents a pivotal solution for addressing critical transportation challenges, including traffic congestion, safety risks, and environmental sustainability. Its experimental teaching, as the core linkage of theoretical innovation and technical verification, is of vital importance to the cultivation of intelligent transportation talents. Compared with traditional experimental teaching, the V2X semi-physical simulation platform effectively reduces capital investment, completely eliminates the safety risks of actual road tests, and emulates the real traffic environment. To verify the teaching effectiveness of this platform, based on the OBE concept and the BOPPPS teaching method, this study constructed an experimental curriculum framework driven by learning goals and conducted an empirical analysis taking global path planning as an example. Teaching evaluation adopts a combination of subjective and objective methods: Subjective evaluation is conducted through questionnaire surveys, and the proportion of those satisfied with the teaching effect reached more than 80%. The objective evaluation consists of eight performance indicators before class, during class and after class. Through reliability analysis, the performance of students in the observation group was shown to increase by 17.39% compared with that in the control group. The results show that the experimental teaching mode based on the V2X semi-physical simulation platform significantly improves the teaching effectiveness of the vehicle–road collaboration course compared with traditional methods.

1. Introduction

V2X is a general term for vehicle-to-vehicle (V2V), vehicle-to-transportation infrastructure (V2I), vehicle-to-pedestrian [1] (V2P), vehicle-to-network (V2N) and vehicle-to-other traffic elements for information interaction and communication [2,3], and is a key component of future global intelligent transportation systems [4,5]. Vehicle–road collaboration is the application of wireless communication, sensors, cloud computing and other technologies to achieve vehicle–vehicle and vehicle–road status information real-time interaction and sharing [6], as well as the collection and analysis of broad space-time traffic information, to achieve intelligent cooperation between vehicles and infrastructure and safe autonomous driving; it is a representative application field of V2X technology [7]. Vehicle–road collaboration can realize the intelligent perception of road traffic information and optimal control of traffic systems, ensure vehicle safety in extremely complex traffic environments, optimize the use of traffic resources, improve traffic efficiency and ensure traffic safety [8]. Since the individual vehicle intelligence does not have the ability to go beyond visual range and multi-dimensional global perception, it cannot optimize traffic safety and traffic efficiency based on broader urban information [9], so it has become the consensus of China’s industry of intelligent connected vehicles to prefer the development of vehicle–road collaboration technology instead of individual vehicle intelligent technology [10]. In the cultivation of talent for vehicle–road collaboration technologies, fostering students’ proactive learning capabilities is particularly crucial. Given the highly interdisciplinary nature and rapid technological evolution characteristic of this field, learners are required not only to acquire theoretical knowledge but, more importantly, to actively engage in practical exploration of real-world challenges such as system integration and algorithm optimization through hands-on practice [11]. At present, colleges in China and developed countries in Europe and the United States have begun to set up vehicle–road collaboration technology and other related majors and courses, focusing on the training talents of intelligent vehicle decision-making control, traffic signal optimization, vehicle networking communication protocols, data analysis and processing in the vehicle–road collaboration field [12]. According to a survey conducted by the authors, colleges are faced with a series of difficulties in the teaching of vehicle–road collaboration-related courses, especially experimental courses, which are embodied in the following three aspects.

(1) To build a real V2X traffic experiment scenario, it is usually necessary to purchase advanced roadside intelligent sensing equipment, an intelligent connected vehicle, cloud servers and various software systems, which is technically complex and costly [13]; this may lead to high costs for experimental courses, and ordinary colleges are often faced with the problem of insufficient funds. Subject to Moore’s Law, the sensors, computing units and communication equipment in the vehicle–road collaboration system need to be constantly upgraded to provide ever-increasing computing power to deal with massive complex data [14]. The frequent updating and iteration of such equipment will bring continuous economic pressure to colleges.

(2) Vehicle–road collaboration experiment courses usually require teachers and students to go deep into the actual road environment and use actual vehicles to carry out relevant experimental activities, which may bring certain safety risks, and colleges need to invest additional resources to ensure the safety of teachers, students and equipment. At the same time, on-road experiments between teachers and students will bring certain social pressure to government road traffic managers, especially when it involves the personalized setting of road experiment scenarios, changing the timing of traffic signals and other special needs, which are not allowed by traffic management departments.

(3) Although computer simulation software can build scenarios that are difficult to achieve in real traffic experimental systems, e.g., complex traffic scenarios such as large-scale road networks and complex traffic intersections, such technology can only realize research and experiments on vehicle collaborative algorithms at the software level. It ignores the impact of physical factors such as wireless communication delay [15], signal interference [16] and vehicle performance on V2X interaction in real traffic environment, which is quite different from the real traffic environment.

In short, the field of vehicle–road collaborative experiment teaching and research is in urgent need of an experimental platform with low capital investment, no traffic safety risk and close to the real traffic environment. Based on this, this paper organically combines V2X theory with semi-physical simulation technology to build a vehicle–road collaborative simulation environment in the laboratory environment. This can provide effective hardware and software environment infrastructure and broad technology development space for real vehicle–road collaborative systems and the experimental courses of colleges in this field. The V2X semi-physical simulation platform described in this paper is based on the teaching requirements of vehicle–road collaboration, and integrates all elements and teaching points of the intelligent transportation system, including intelligent supervision cloud platform, network communication system, high-precision positioning system, intelligent terminals (intelligent miniature vehicle, integrated display terminal, etc.), smart transportation sand table, etc. It can display the operational principles of intelligent connected vehicles, verification of algorithms, and simulation and verification [17] of the comprehensive efficiency of advanced road traffic systems, providing a complete teaching environment and perfect method guidance for teaching and research work in vehicle–road collaboration.

2. Physical and Logical Framework of the Platform

2.1. Platform Physical Structure

The V2X semi-physical simulation platform simulates the principle and form of the vehicle–road collaboration system from multiple levels of intelligent supervision: cloud platform, communication system, positioning system, intelligent terminal and simulation scenario. The topology of the hardware system of the simulation platform is shown in Figure 1.

Figure 1. Physical topology of the V2X semi-physical simulation platform.

2.1.1. Intelligent Supervision Cloud Platform

The intelligent supervision cloud platform receives information from elements such as sand table infrastructure, intelligent miniature car, and indoor high-precision positioning through the indoor communication network to realize the monitoring and control scheduling of the entire smart transportation sand table status. The intelligent transportation simulation system and sand table dynamic real-time reconstruction technology realize intelligent transportation and provide remote, all-round, three-dimensional, visual monitoring, for intuitive understanding of intelligent transportation operation status.

2.1.2. Network Communication System

The V2X semi-physical simulation platform uses the indoor WiFi network with low latency to simulate the intelligent transportation communication network, and realizes communication between vehicle and vehicle (V2V), vehicle and infrastructure (such as traffic lights, speed limit signs, intelligent traffic supervision system) (V2I), vehicle and platform (V2N), pad and vehicle, pad and platform, and sand table and platform. All elements are linked to the same communication network to ensure the rapidity, security and low latency of data exchange of all important elements of intelligent transportation. The communication equipment is composed of a router, a switch and an on-board network card; the minimum requirement of the router is to support 4 gigabit network ports, the minimum requirement is to support 30 devices online at the same time, the required signal coverage is ≥60 m², and the sand table communication system can use 4G and/or 5G communication, so that it is closer to real traffic communication scenarios.

2.1.3. High-Precision Positioning System

There are 14 infrared cameras, as shown in Figure 2, installed 3 m above the smart transportation sand table, which can accurately provide the position of the sand table itself and the vehicles running on it by using dynamic capture technology; the positioning accuracy is less than 3 mm. The static map format of the sand table is a vector-directed graph; the slam map is in pgm format, containing vehicle accuracy, latitude, heading, speed and other information. The process of constructing the high-precision sand table map is as follows: the data collected by LiDAR and intelligent camera are fused to obtain the point cloud image data, the point cloud image data are preprocessed, and then the map coordinate system is established and the features in the map are extracted from the point cloud data to complete the map’s construction.

Figure 2. Indoor millimeter-level dynamic position capture system.

2.1.4. Intelligent Vehicle Terminal and Control Terminal

Intelligent miniaturized vehicles (vehicle terminals) are equipped with sensors such as LiDAR, intelligent cameras, and light sensing balls. The characteristics of these sensors are as follows: LiDAR requires measuring distance ≤ 12 m, measuring blind area ≤ 0.2 m, pitch angle error between ±1.5°, intelligent camera requires pixel ≥8 million, resolution is 3280 × 2464, car profile size is 25 mm × 24 mm × 8 mm, the operating system is ubuntu20.04, ROS version is Noetic. This intelligent miniaturized vehicle, shown in Figure 3, can realize autonomous driving functions such as path planning, tracking and active stopping. The integrated display device or a mobile phone can be used as the remote scheduling control terminal, as shown in Figure 4. The intelligent transportation facility terminal has traffic lights, ETC (Electronic Toll Collection), intelligent parking lot, etc.

Figure 3. Miniature intelligent connected vehicle.

Figure 4. Integrated display terminal.

2.1.5. Smart Transportation Sand Table

The area of the smart transportation sand table is about 80 M², and the main part includes the simulated road, LED information display and related traffic facilities, as shown in Figure 5. The simulated road is 1:15 scale with a single lane width of 15 cm. It has two types of road sections, straight and curved. The road surface is roughened to enhance tire adhesion. At the same time, the sand table has many road traffic facilities and scenarios such as ordinary urban roads, traffic signal plane intersections, no traffic signal plane intersections, no traffic signal city turntables, expressways, urban parking lots, and ETC, which can be used to carry out a variety of V2X technology-related experimental research.

Figure 5. Traffic simulation scenario of V2X semi-physical simulation platform.

2.1.6. Boundary Constraints and Real-World Challenge Simulation

To enhance the realism of virtual driving environments and the robustness of pedagogical experiments, this platform employs a multidimensional technical framework to establish boundary condition constraints. The system integrates high-precision motion capture cameras with SLAM-based topological mapping to monitor vehicle trajectories in real time. Upon detecting boundary violations, it instantly activates path replanning algorithms while dynamically constraining acceleration and steering angles through vehicle dynamics models, effectively preventing unrealistic “teleportation” phenomena. The platform ensures data continuity via dual-network redundant communication (WiFi + 5G) and local caching modules (Ubuntu 20.04 ROS nodes). The communication layer implements TCP packet fragmentation verification and QoS prioritization strategies, guaranteeing reliable transmission of safety-critical messages such as collision warnings. In case of sensor anomalies, the system autonomously switches to multi-sensor fusion mode, enabling real-time data compensation and recovery during LiDAR or camera failures.

To truly reproduce the uncertainty characteristics of the Internet of Vehicles communication environment, the platform deeply integrates NS3 network simulation modules to precisely inject controllable delays [18] (5–100 ms) and Rayleigh fading channel interference into V2V/V2I links, replicating urban multipath propagation and tunnel occlusion scenarios. Through dynamic adjustment of router bandwidth and interference source power spectral density, the system reconstructs real-world communication quality fluctuations. This mechanism interfaces with a sandbox LED display system to dynamically optimize traffic signal phases based on real-time channel states, requiring students to develop adaptive path planning algorithms that maintain optimal driving strategies under complex communication conditions.

2.2. Platform Logical Structure

Aiming at the requirements of teaching and technology research and development, the intelligent connected vehicle simulation experiment platform is constructed, and the sand table, intelligent hardware, intelligent miniaturized vehicle and integrated display are used as terminals. Through the indoor WiFi communication network, the intelligent traffic monitoring cloud platform and simulation system communicate with each other to form a closed-loop and logically complete simulation whole, as shown in Figure 6.

Figure 6. Logical structure of V2X semi-physical simulation platform system.

2.3. Comparative Analysis of V2X Simulation Tools

To clarify the differentiated advantages of the V2X semi-physical simulation platform in the experimental teaching of vehicle–road collaboration, this section systematically compares it with mainstream educational simulation tools such as CARLA, SUMO, and modular robot kits from six perspectives: cost, scalability, environmental fidelity, communication realism, pedagogical suitability, and learning outcomes, as shown in Table 1.

Table 1. Comparison between V2X communication platforms and mainstream simulation software.

Dimension	V2X HIL Simulation Platform	CARLA	SUMO	Modular Robotics
Cost	Moderate (hardware sandbox + commercial sensors, no high-performance GPU required)	High (GPU-dependent rendering, high hardware costs)	Low (pure software, open-source)	Low to moderate (hardware kits + basic sensors)
Scalability	High (supports dynamic scenario injection, multi-protocol extension, customizable road networks)	Moderate (relies on predefined 3D maps, requires programming for scenario expansion)	High (flexible traffic flow models but lacks vehicle control interfaces)	Low (limited by fixed kit modules)
Environmental Fidelity	High (1:15 physical sandbox + high-precision positioning, supports multi-sensor fusion)	Very high (high-fidelity 3D environment, precise physics engine)	Low (abstract traffic flow, no 3D visualization)	Low (simplified physics models, no traffic interaction)
Communication Realism	High (supports V2V/V2I/V2N protocols, simulates latency/interference)	Moderate (requires plugins for V2X simulation, no hardware link)	None (traffic flow simulation only)	Low (basic wireless communication, no V2X protocol support)
Pedagogical Suitability	High (layered experiment design, fault injection modules, teaching manuals)	Moderate (suited for algorithm research, lacks teaching resources and templates)	Moderate (suited for traffic engineering theory)	Moderate (suited for basic programming and control)
Learning Outcomes	Excellent (hardware–software synergy, intuitive understanding of full-chain vehicle–road collaboration)	Good (focuses on autonomous driving algorithms, requires advanced programming skills)	Moderate (macroscopic traffic analysis, lacks hands-on practice)	Moderate (basic robot control, no V2X scenarios)

An analysis of Table 1 shows that although CARLA can construct a high-fidelity 3D environment, it relies on predefined static maps and algorithm-driven models, making it inadequate for simulation of the impact of dynamic random events in real traffic on communication and decision-making. As a pure traffic flow simulation tool, SUMO performs well in macroscopic flow prediction. However, it lacks vehicle dynamics models and physical interaction interfaces, and is unable to reflect the delay response and sensor noise in real driving. Modular robot kits are limited by simplified physical models and solidified functional modules, making them inadequate for supporting complex scenarios such as multi-vehicle collaboration and V2X protocol verification. These tools generally face the problem of the lack of real physical traffic rules.

The V2X semi-physical simulation platform addresses the aforementioned limitations by integrating physical sand tables with digital twins. The sand table, combined with high-precision sensors such as LiDAR and infrared cameras and dynamic constraint models, can reproduce the physical interaction in real traffic scenarios. Through hands-on operation of physical vehicles and real-time data feedback, students gain an intuitive understanding of the nonlinear characteristics in vehicle–road collaboration systems—enabling them to develop core competencies for managing real traffic uncertainties in a controlled environment.

3. Teaching Method Design and Teaching Scenario

3.1. OBE Concept

OBE (Outcome-Based Education) is a student-centered educational approach that focuses on predefined learning outcomes or achievements as its core directive. It emphasizes that students should attain specific competencies and knowledge levels during the learning process, rather than merely completing a fixed number of courses or credit hours [19]. A vehicle–road collaborative experimental teaching course should be guided by specific educational outcomes, such as ensuring that students can master V2X communication protocols, understand the working principle of on-road collaborative systems, and have the ability to design and implement intelligent miniature vehicle path planning through course learning. The course design should include theoretical instruction, experiments and practical projects related to V2X technology. Teaching methods should promote these outcomes, for example through case studies, team projects and practical activities that allow students to explore the application of V2X technology in vehicle–road collaboration scenarios [20]. Assessment methods should correspond directly to learning outcomes [21], for example by examining students’ practical application in the project and their understanding of the design of vehicle–road collaborative experimental scenarios. Finally, based on students’ feedback and study results, the curriculum should be continuously improved to ensure that the teaching methods and content can effectively support students’ learning and growth in the field of V2X and vehicle–road collaboration, and strengthen students’ practical skills and innovation in this field.

3.2. BOPPPS Teaching Model

Teaching design is the key factor affecting the classroom teaching effect, and it is the basis of carrying out teaching activities and ensuring teaching quality [22]. BOPPPS, as an international universal teaching model, has strong flexibility and can be used in the development and design of teaching models in various fields [23,24]. The advantage of the BOPPPS model lies in its comprehensiveness and flexibility, which not only emphasize the clarity of learning objectives and the importance of assessment, but also focus on students’ active participation and continuous feedback on the learning process [25]. Through this structured approach to teaching, teachers can design and execute lesson plans more effectively, while also better meeting students’ learning needs and preferences [26]. Therefore, this paper combines the BOPPPS model with experimental teaching in the field of vehicle–road collaboration. A teaching mode suitable for the vehicle–road collaborative experimental course is constructed in six stages: Bridge-In, Objective, Pre-Assessment, Participatory Learning, Post-Assessment, Summary, as shown in Figure 7.

Figure 7. BOPPPS advanced teaching model.

3.2.1. Bridge-In

Use current novel industry cases or news reports to introduce the application of V2X technology in typical scenarios such as global path planning, automatic driving and traffic signal control in the field of vehicle–road collaboration, so as to arouse students’ interest in the course and their desire to explore.

3.2.2. Objective

Present clear learning objectives to enable students to understand key concepts and theoretical models in V2X such as path planning, environmental perception, decision control, data communication, traffic flow simulation and V2X safety, and master experimental verification methods.

3.2.3. Pre-Assessment

The students’ knowledge of V2X and vehicle–road collaboration technology is assessed through simple experiments or discussions of specific experimental topics, so as to provide a reference and basis for the follow-up teaching work.

3.2.4. Participatory Learning

Based on the OEB concept, the BOPPPS model is applied for vehicle–road collaboration V2X teaching. On the V2X semi-physical simulation platform, teachers and students can carry out experimental courses such as path planning experiments, environment perception, decision control, data communication, traffic flow simulation, V2X safety, etc., as shown in Table 2. Among these possibilities, experiments 1, 2 and 3 can be arranged for teaching in the sixth semester, and experiments 3, 4 and 5 can be arranged for teaching in the seventh semester.

Table 2. List of feasible experimental projects.

Serial Number	Experiment Name	Experiment Content	Experiment Purpose
Experiment 1	Data Communication Experiment	Based on different communication protocols, OBU of intelligent miniature vehicle, roadside equipment (RSU, MEC, Camera, radar), ETC and other equipment are connected to the cloud server for “human–vehicle–road” data exchange and communication, and the scheduled control functions and experiments are completed according to the communication content.	Enable students to master the communication protocols and data structure characteristics of the vehicle–road collaborative system. Cultivate students’ abilities in innovation, leadership and teamwork.
Experiment 2	Traffic flow simulation experiment	Set different traffic flows, obtain the information interaction between vehicles (real-time location sharing, speed coordination, obstacle avoidance decision-making, etc.) and traffic lights, V2X roadside units and other infrastructure through the cloud control platform, observe and analyze the changing rules of traffic flow, such as the formation and dissipation of congestion, vehicle speed distribution, etc.	To enable students to understand the changing rules of traffic flow in an intelligent traffic environment and the role of V2X technology in improving traffic safety and efficiency. Cultivate students’ innovative ability, hands-on ability and independent thinking ability.
Experiment 3	V2X Safety experiment	Through reasonable settings for different traffic scenarios and traffic signal periods, the safety performance of the autonomous driving module and V2X system, such as collision avoidance, emergency braking, vehicle stability and network security, is tested.	Enabling students to master the methods of independently designing experimental schemes, setting up experimental environments, and collecting and analyzing data in different traffic scenarios will help cultivate students’ hands-on ability and independent thinking ability.
Experiment 4	Environmental Awareness Experiment	Collect perception data from intelligent vehicle cameras, radar and other sensors and conduct data fusion to dynamically perceive the driving environment of the intelligent miniature vehicles and provide digital driving decision information for those vehicles.	Enable students to master the general method of sensor information fusion, and be familiar with the general AI deep learning model. Cultivate students’ innovative ability, hands-on ability and teamwork ability.
Experiment 5	Intelligent vehicle path planning experiment	Based on the high-precision map data of the V2X semi-physical simulation platform, different path planning algorithms are used to calculate a collision-free path from the starting point to the destination point under different cost function constraints, and the path optimization is carried out according to real-time traffic information and vehicle technical conditions.	Enable students to master the modeling and practical application ability of common path search algorithms such as A* and Dijkstra. Cultivate students’ abilities in innovation, leadership and teamwork.
Experiment 6	Vehicle decision control experiment	Based on multi-source sensor data fusion parameters, the expected driving path of the intelligent miniature car is generated, and the corresponding control signal is sent to the controller to ensure that the intelligent miniature vehicle can complete the automatic driving task under the premise of safety and efficiency.	Enable students to master the basic theories and methods of automatic driving decision planning and control, such as polynomial path planning theory, PID control technology, etc. Cultivate students’ abilities in innovation, leadership and teamwork.

3.2.5. Post-Assessment

Evaluate students’ understanding and learning effect of the corresponding intelligent transportation experiment project through questionnaires, simulated project presentations or experiments.

3.2.6. Summary

Summarize the main content of the course, emphasize the application of V2X technology in vehicle–road collaboration scenarios, and encourage students to think about how to apply the knowledge in the field of traffic science.

In this way, the BOPPPS model combined with the professional experiment content in the field of vehicle–road collaboration can build a comprehensive and interactive teaching framework, which is helpful to improve the learning effect of students’ professional knowledge and innovative thinking level in the field of V2X vehicle–road collaboration.

4. Design of Experimental Teaching Links

Global path planning plays a crucial role in autonomous driving, not only ensuring that the autonomous vehicle can safely and efficiently reach the end point from the starting point, but also optimizing the vehicle’s driving route, reducing driving time and energy consumption. In the research and teaching of autonomous driving technology, using the V2X semi-physical simulation platform to conduct global path planning experiments can help students better understand the algorithm principles and improve their ability to solve practical problems. In the environment of a V2X semi-physical simulation platform, this paper discusses the design of autonomous vehicle path planning experiments based on the OEB concept and the BOPPPS teaching model.

4.1. Teaching Preparation

Based on high-precision millimeter-level visual positioning, AR and high-precision map technologies, the cloud control center conducts fine speed guidance and regulation and scheduling for intelligent miniature vehicles to ensure autonomous obstacle avoidance and navigation functions. Through the network communication system, the vehicle network and the roadside system are effectively connected, so that the dynamic and static data of the road are integrated, and the interconnection between the vehicle, road, object and cloud can be realized. The ROS intelligent miniature vehicle positioning information, path planning information, vehicle status information, etc., can be displayed in real time on the integrated display terminal. It can effectively and vividly demonstrate the experimental process of path planning and stimulate students’ enthusiasm for learning.

4.2. Experimental Purposes and Requirements

The high-precision positioning system and the intelligent miniature vehicle with ROS system are used to carry out joint SLAM mapping of the smart transportation sand table. Figure 8 shows the high-definition (HD) electronic map of the smart transportation sand table. The Dijkstra path planning algorithm runs in the Ubuntu 20.04 environment to adjust the vehicle route in a timely fashion, from the 23rd node near the airport to the 25th node near the Tianjin Eye. Three different path planning strategies, namely the shortest distance, the fewest red traffic lights path and the shortest path passing through the highway, were explored, and the students drew experimental conclusions through analysis, mathematical modeling and programming. Finally, the students’ learning effect is evaluated through questionnaire survey and achievement analysis.

Figure 8. HD electronic map of V2X semi-physical simulation platform.

4.3. Teaching Process Design

4.3.1. Bridge-In (Stimulate Interest)

This paper introduces the application examples of path planning technology in automatic driving at home and abroad, discusses the impact of V2X technology on the safety and efficiency of automatic driving, and then leads to the experimental course theme of “Global path planning of intelligent vehicles based on a V2X semi-physical virtual simulation platform”.

4.3.2. Objective (Specific Aim)

The objective is to use the Dijkstra algorithm to find the optimal path in the road network of the smart transportation sand table from the 23rd node near the airport to the 25th node of the Tianjin Eye. The shortest distance path, the path with the fewest red lights, and the shortest path passing through the expressway are the three cost functions. The composite cost function for path P is given by Equation (1):

$$C(p)={\displaystyle \sum _{i=1}^n\left(\mathrm{\alpha }\times \frac{s_i}{s_{\max }}+\mathrm{\beta }\times \frac{1}{1+e^{-k(r_i-r_0)}}+\mathrm{\gamma }\times h_i\times {\mathrm{I}}_{highway}(i)\right)\times (1+\frac{\mathrm{\eta }}{1+e^{-\mathrm{\lambda }(t_i-t_{avg})}}})$$

(1)

In the formula, ${\displaystyle \frac{s_i}{s_{\max }}}$, $s_{\min }=\underset{\forall j\in E}{\min }(s_j)$ represent the normalized distance cost; $s_i$ indicates the actual length of the i-th road section; $s_{\min }$ denotes the shortest path distance from the origin to the destination in the network.

${\displaystyle \frac{1}{1+e^{-k(r_i-r_0)}}}$ denotes the penalty term associated with traffic signals, $r_i$ represents the number of red lights for section i, k represents the slope coefficient, $r_0$ represents the threshold.

$h_i\times {\mathrm{I}}_{highway}(i)\left\{\begin{array}{c}{h}_i,i\in Highway\\ 0,otherwise\end{array}\right.$ represents the preferred option for the expressway, $h_i$ represents the priority coefficient.

$1+{\displaystyle \frac{\mathrm{\eta }}{1+e^{-\mathrm{\lambda }(t_i-t_{avg})}}}$ denotes the adaptive weighting factor for dynamic tuning, $t_i$ is the actual real-time travel time measured and provided by V2X systems, $t_{avg}$ represents the historical mean, $\mathrm{\eta }$ is the adjustment intensity, and $\mathrm{\lambda }$ is the smoothing coefficient.

$\mathrm{\alpha },\mathrm{\beta },\mathrm{\gamma }$ are weight constraints, and they should follow Equation (2):

$$\mathrm{\alpha }+\mathrm{\beta }+\mathrm{\gamma }=1$$

(2)

To quantitatively evaluate the impact of V2X communication on the performance of path planning algorithms, this study designed a comparative experiment to analyze the differences in the effectiveness of the Dijkstra algorithm enhanced by V2X and the traditional independent path planning in solving the optimal path.

The experimental group utilized the V2X vehicle–road collaboration communication system. The improved Dijkstra algorithm based on real-time V2X data updated the dynamic traffic status information in real time through V2I and V2V communications. The control group, Non-V2X communication, adopted the traditional Dijkstra algorithm, with weights solely based on fixed path lengths and lacking real-time update capabilities.

The path planning of the vehicle under V2X communication and Non-V2X communication were collected separately, and the results are shown in Table 3. As the table shows, the group Non-V2X was unable to perceive congestion, and the actual driving speed was reduced to 0.8 m/s.

Table 3. Comparison of results based on V2X vs. Non-V2X communication.

Group	Path Length	Number of Red Lights Passed	Red Light Waiting Time	Road Type	Base Speed Limit	Dynamic Traffic Flow Impact	Total Travel Time
Non-V2X	9.3 m	3	7 + 4 + 5 = 16 s	Urban road	0.8 m/s	Real-time congestion	27.63 s
V2X	12.9 m	0	0 s	Highway	2.5 m/s	Unobstructed	5.16 s

The V2X group could dynamically calculate the optimal route. Due to the heavy traffic on the main roads and the presence of traffic lights, vehicles can choose to travel on the expressway. Although the route cost increases, the traffic flow is less and there is no need to wait for traffic lights, so the total travel time is within the ideal range. This is more in line with the expectations of vehicles in actual traffic environments. The actual driving path of the vehicle is as shown in Figure 9. The red part represents the path planning of the smart vehicle based on V2X communication, while the yellow part represents the path planning based on Non-V2X communication.

Figure 9. The actual driving path of the car.

Set the weights as $\mathrm{\alpha }=0.3,\mathrm{\beta }=0.4,\mathrm{\gamma }=0.3$, the normalized maximum path length $S_{\max }=14.2$ m; the slope coefficient k = 1, the penalty activation threshold $r_0=1$; the high-speed reward term $h_i=0.2$. By substituting the values into Equation (1), $C_{non-v2x}=1.253$; $C_{v2x}=0.412$. Through comparative analysis, we can see that the cost of the group Non-V2X is approximately three times that of the group with V2X communication, and the V2X advantage is obvious. The comparison results between the two groups are as shown in Figure 10. To visually demonstrate the performance superiority of V2X communication technology in path planning, the benefit index is used as an evaluation indicator in the figure, and a unified quantitative framework is used for the comparison of system performance. The benefit index maps the compound discounted value through a nonlinear transformation function into a positive evaluation indicator, and the relationship is expressed by Equation (3):

$$Index=\max (0,C_{\max }-C(P))$$

(3)

Figure 10. Comparison of multi-dimensional performance between V2X and Non-V2X.

In the formula, $C_{\max }=1$. The increase in values indicates the improvement in the overall performance of the planning scheme.

During the subsequent experiments, students can dynamically adjust the parameters. For instance, they can set $\mathrm{\beta }$ to 0 to ignore red lights and observe whether the car will choose the congested route, or increase the weight of $\mathrm{\gamma }$ to prioritize high-speed travel, analyzing the pros and cons of taking the highway bypass. This not only helps to stimulate students’ interest in learning, but also enables them to intuitively understand how V2X optimizes the overall cost through dynamic route selection, further deepening their understanding of the algorithm’s robustness.

4.3.3. Pre-Assessment (Diagnostic Practice)

By means of discussion, questions and simple examination, we will understand students’ grasp of prior knowledge of classical method of path planning, key points of the Dijkstra algorithm in the V2X environment, Python programming on an intelligent terminal PC, etc., so as to provide a basis for subsequent teaching implementation and dynamic adjustment.

4.3.4. Participatory Learning (Engaged Learning)

(1) Theoretical learning: Introduces the working principle of Dijkstra’s algorithm, including calculation processes such as initialization, node selection, neighbor update, marking as visited, checking completion conditions, repeated node selection, path reconstruction, etc. In a V2X vehicle–road collaborative environment, this process may involve real-time data, such as traffic conditions and signal light status, which are integrated into the cost calculation via V2X communication. Figure 11 shows the process of students visualizing the operating data (odometer data, IMU data, coordinate system information) of the ROS system in the car, using C++20 and Python (Version 3.9) programming languages.

Figure 11. Map data and vehicle trajectory information optimization.

(2) Practical operation: Students work in groups of five and carry out a detailed division of responsibilities. They use a high-precision positioning system, intelligent miniature vehicle and cloud control platform to program and operate on the integrated display terminal of the V2X semi-physical simulation platform or external PC, and apply the Dijkstra algorithm for path planning.

4.3.5. Post-Assessment (Post-Evaluation)

Evaluate the learning effectiveness according to the experimental results and students’ feedback, especially the understanding of the path selection strategy under different cost functions. Conduct simple written experiments, questions or discussions to assess students’ mastery of the core content of the course.

4.3.6. Summary (Outcome Patterns)

This paper reviews the key technical points of the Dijkstra algorithm in a V2X vehicular and road cooperation semi-physical simulation platform, discusses how to choose the appropriate cost function in different scenarios, and how V2X vehicular and road cooperation technology can improve the efficiency and safety of path planning. This will effectively improve students’ understanding of the basic concepts and principles of V2X, master the general methods of V2X path planning and optimization, and better cultivate their ability to solve practical problems and innovate.

5. Teaching Effect Demonstration and Data Analysis

5.1. Empirical Design of Teaching Effect

In order to better test the application effect of the V2X semi-physical simulation platform in teaching, this paper designs two evaluation methods: the subjective teaching effect and the objective teaching effect [27]. The subjective teaching effect evaluation is carried out by questionnaire survey, and the objective teaching effect evaluation is carried out by multi-index scoring and reliability analysis [28].

The objective teaching effect evaluation is divided into three parts: pre-class assessment, process-class assessment and post-class assessment, and the evaluation proportions are 10%, 40% and 50%, respectively. The evaluation system of the learning effect is shown in Figure 12, and the application effect of the semi-physical simulation platform in teaching is analyzed through quantitative comparison and investigation and analysis.

Figure 12. Objective teaching effect evaluation system.

5.2. Research Object

This study adopted a quasi-experimental design using the 2023 cohort (control group) and 2024 cohort (observation group) of automobile service engineering students as research subjects, with experimental samples constructed through stratified cluster sampling. Students in the observation group were required to meet two criteria: first, complete participation in both traditional road experiments during the sixth semester and V2X simulation platform instruction in the seventh semester; second, all participants were either competitors or volunteers in the 2023 World Intelligent Driving Challenge (WIDC) real-vehicle competition to ensure a consistent baseline understanding of vehicle–road collaboration technologies, with teaching effectiveness data statistically analyzed using SPSS software(Version 28.0). The control group consisted of all 2023 cohort students without exposure to the V2X platform. To control for potential selection bias, a pre-test verified no significant differences in baseline V2X theoretical assessment scores between groups, while the difference-in-differences (DID) method was employed to isolate the teaching platform’s effects from temporal confounding factors. Furthermore, course syllabi, teaching teams, and evaluation standards remained identical across both cohorts to minimize external interference.

5.3. Correlation Evaluation Method

5.3.1. Subjective Evaluation of the Correlation Between Simulation Platform and Teaching Effect

To validate the pedagogical effectiveness of the V2X semi-physical simulation platform, the research team developed a structured subjective evaluation system based on experimental data from 60 students in the observation group. The questionnaire employed a five-point Likert scale with the following options: Strongly Agree (SA), Agree (A), Neutral (N), Disagree (D), Strongly Disagree (SD) [29,30]. The survey comprised ten single-choice items (Q1–Q10), with each question offering these five standardized response options to precisely capture respondents’ authentic perceptions while minimizing measurement bias and subjective interference. The complete questionnaire content is presented in Table 4.

Table 4. Questionnaire design.

Items	Questions About the Platform and Learning Effectiveness	Answers
		SA	A	N	D	SD
Q1	Compared with traditional teaching methods, experimental courses based on the simulation platform enhance the understanding of new technologies in the field of vehicle–road collaboration
Q2	Compared with traditional teaching methods, the simulation platform stimulates students’ learning interest in vehicle–road collaboration technology
Q3	Compared with traditional teaching methods, the simulation platform improves students’ theoretical knowledge and application ability in the field of vehicle–road collaboration
Q4	Compared with traditional teaching methods, the simulation platform greatly increases the number of experimental projects that can be carried out
Q5	Compared with traditional teaching methods, the simulation platform significantly improves the safety of the teaching process
Q6	The hardware and software configuration of the simulation platform covers all the key knowledge points in the field of vehicle–road collaboration
Q7	Compared with traditional teaching methods, experimental courses based on simulation platforms make knowledge easier to understand and accept
Q8	Compared with traditional teaching methods, the teaching concept and course organization form based on the simulation platform are more reasonable and scientific
Q9	Compared with traditional teaching methods, students have a greater sense of participation, interaction and achievement in the teaching process
Q10	Compared with traditional teaching methods, the teaching results of experimental courses are more in line with pre-class teaching goal setting and effect orientation

Note: SA (Strongly Agree), A (Agree), N (Neutral), D (Disagree), SD (Strongly Disagree).

5.3.2. Evaluation of Knowledge Promotion Effect

Staged experiments were carried out throughout the process of learning before, during and after class, and the knowledge experiments in the three stages accounted for 10%, 40% and 50% of the total score, respectively. The students in the two groups were tested using the same paper, and the scores from each subdivided assessment result were collated to objectively evaluate the effect of the platform on the improvement of the students’ knowledge level.

5.4. Data Analysis

5.4.1. Questionnaire Analysis

The project team sent out a total of 60 questionnaires and recovered 58 questionnaires, with a recovery rate of 96.67%, meeting the requirements of questionnaire statistical analysis. For each questionnaire, 10 questions were reasonably designed according to the three assessment stages and the usefulness of the semi-physical simulation platform, and the subjective teaching results of using the semi-physical simulation platform were evaluated according to the recovered questionnaires. The valid data from the recovered questionnaires was analyzed, and the results are shown in Figure 13. The results show that Strongly Agree (SA) and Agree (A) account for 81.3% of the total samples. This indicates that students’ willingness to use the V2X semi-physical simulation platform for teaching is much higher than that of traditional experimental teaching. Teaching with the experimental platform can significantly improve the teaching effect of the vehicle–road collaborative experiment course.

Figure 13. Analysis of questionnaire data.

5.4.2. Correlation Analysis Between Simulation Platform and Objective Teaching Effect Improvement

This study constructs eight indicators related to the V2X semi-physical simulation platform and the improvement of the objective teaching effect, which are the following: experimental skills, innovation awareness, knowledge application, professionalism, safety awareness, self-presentation, teamwork and learning interest. These were used to analyze the improvement of the objective teaching effect of the above platform [31]. The results of eight indicators were obtained after cross-checking, as shown in Table 5. The study results of students in the observation group increased by 17.39% compared with those in the control group. Based on the data in Table 5, SPASS 26 was used for reliability tests to determine the variables for reliability analysis (i.e., scale items in the questionnaire) and ensure that the data types of these variables were suitable for reliability analysis. The results are shown in Table 6. Based on the analysis results of eight indicator features in the three stages, the Cronbach’s alpha coefficient of the total table of the observation group was 0.85, while the Cronbach’s alpha coefficient of the total table of the control group was 0.78, indicating that the scale reliability met the requirements [32]. It reflects a good effect of the simulation platform on the improvement of students’ professionalism.

Table 5. Study results table of eight indicators.

Features	Experimental Skills	Innovation Awareness	Knowledge Application	Professionalism	Safety Awareness	Self-Presentation	Teamwork	Learning Interest	Total Points
Control group
Pre-class grades	9.67	8.98	9.07	8.32	9.21	10.43	8.68	9.48	73.84
Process-class grades	9.23	8.52	11.35	8.96	10.10	8.73	9.87	8.67	75.43
Post-class grades	9.45	9.54	9.89	8.28	9.90	10.16	8.34	8.84	74.40
Total points	9.38	9.08	10.39	8.56	9.91	9.62	8.99	8.84	74.76
Observation group
Pre-class grades	11.43	11.98	10.12	12.42	10.28	10.99	12.13	12.36	91.71
Process-class grades	11.21	10.56	11.07	10.85	11.30	11.15	11.83	11.52	89.49
Post-class grades	10.03	11.77	10.40	11.68	10.67	10.74	10.07	10.22	85.58
Total points	10.64	11.31	10.64	11.42	10.88	10.93	10.98	10.95	87.76

Table 6. Platform reliability analysis table.

Features	Experimental Skills	Innovation Awareness	Knowledge Application	Professionalism	Safety Awareness	Self-Presentation	Teamwork	Learning Interest	Total Points
Scale mean after deleting items	9.78	9.95	10.45	10.20	9.78	10.11	8.96	9.66	Control group
	10.89	11.44	10.53	11.65	10.75	10.96	11.34	11.37	Observation group
Scale variance after deleting items	0.16	1.72	0.63	1.36	0.22	1.56	0.65	0.86	Control group
	0.57	0.59	0.24	0.62	0.27	0.04	1.24	1.16	Observation group
Revised item and total correlation	0.65	0.70	0.71	0.62	0.74	0.69	0.66	0.70	Control group
	0.72	0.82	0.76	0.83	0.79	0.70	0.74	0.81	Observation group
Square multiple correlation	0.74	0.68	0.65	0.72	0.70	0.61	0.67	0.69	Control group
	0.82	0.72	0.85	0.78	0.82	0.74	0.86	0.78	Observation group
Cronbach coefficient after deleting the item	0.74	0.64	0.77	0.83	0.77	0.69	0.72	0.86	Control group
	0.78	0.71	0.94	0.91	0.72	0.90	0.89	0.86	Observation group
Overall Cronbach coefficient of the scale	0.78								Control group
	0.85								Observation group

Based on students’ learning outcomes, statistical analysis methods were employed to examine whether the differences between the control group and observation group in pre-course, mid-course, post-course scores, and total scores were statistically significant. The dataset included academic records of 60 students from each group, covering three assessment phases (pre-course, mid-course, post-course) and weighted total scores. To comprehensively evaluate between-group differences, independent sample t-tests were conducted to quantify the practical significance of observed differences.

Preliminary analysis of Table 5 revealed that the observation group outperformed the control group across all assessment phases and total scores. However, statistical significance testing was required to validate these differences. The independent samples t-tests revealed statistically significant differences between groups across all assessment phases: pre-course scores showed t = −7.89 with p < 0.001, mid-course scores demonstrated t = −4.12 with p < 0.001, post-course scores yielded t = −3.45 with p = 0.001, and total scores produced t = −6.78 with p < 0.001. All p-values were below the 0.05 significance threshold, indicating that the observation group’s superior performance across all measurement points was statistically significant with substantial practical importance.

Further analysis of total score distribution characteristics showed the observation group’s mean total score was significantly higher than the control group’s. The observation group exhibited slightly greater score variability (SD = 4.38) compared to the control group (SD = 3.91). The boxplot visualization in Figure 14 clearly demonstrated that the observation group had higher interquartile ranges and median values than the control group, with no overlapping outliers, providing additional support for the statistical test results.

Figure 14. Comparison of total scores between groups.

In conclusion, this study confirmed through the independent sample t-test that the observation group was significantly superior to the control group before, during, after class and in the total score, and the effect size analysis indicated that the differences were of practical significance.

The project team weighted the assessment results of each stage of the 2023 and 2024 student vehicle collaborative experiment courses to obtain the final results. The detailed evaluation results are shown in Table 5 and Table 6. In order to further highlight the differences between the control group and the observation group in eight characteristic indicators, an in-depth comparative analysis was performed on the assessment results of each characteristic of the two groups of students in the intelligent transportation experiment course, as shown in Figure 15. The results show that the total scores and eight characteristic scores of the observation group are completely enveloped and significantly better than those of the control group, which fully indicates that teaching with the V2X semi-physical simulation platform shows significant advantages in improving the teaching effect compared with traditional teaching methods. The eight characteristic scores of the observation group were evenly distributed and all were between 10 and 11 points, which reflected that the V2X semi-physical simulation platform had a comprehensive and obvious effect on the cultivation and improvement of students’ comprehensive quality.

Figure 15. Comparative analysis of experiment scores of each feature of the experimental course.

5.5. Cross-Scenario Universality and Feasibility Analysis of Large-Scale Application

The V2X semi-physical simulation platform, with its modular hardware architecture and open-source technology ecosystem, has successfully demonstrated the underlying logic for large-scale application in cross-institutional and cross-scenario promotion and verification. Through the shared model of regional education alliances, it can flexibly adapt to the resource-differentiated demands of various institutions. Additionally, the platform’s dynamic parameter configuration function can quickly generate six types of scenarios, including urban intersections and highways, with an experimental reproduction rate of over 80%, fully validating its scenario generalization capability. For student groups with different professional backgrounds, the platform provides hierarchical experiment templates and visualization tools, reducing the pass rate difference between engineering and non-engineering students to 9%, significantly enhancing the inclusiveness of teaching. The open-source resource sharing mechanism further empowers scenario expansion. These features collectively indicate that the platform, through a closed loop of “configurable scenarios–adaptive teaching–open collaboration”, has a universal foundation across the education ecosystem, providing a prerequisite for large-scale promotion.

Based on the above promotion foundation, the platform has achieved the feasibility of large-scale institutional deployment through a collaborative design of low-cost scalable architecture, low-threshold operation and maintenance, and synchronized technology iteration. At the hardware level, all platform components adopt industrial-grade standard parts, supporting bulk purchasing and phased expansion. At the operation and maintenance level, containerized deployment and automated diagnostic tools will greatly enhance maintenance efficiency, and teachers with standardized training can manage multi-node environments, significantly reducing human resource input. The continuous contribution of the open-source community ensures the platform’s technological cutting-edge status and avoids duplicate investment.

6. Conclusions and Prospects

This study thoroughly examines the effectiveness of the V2X vehicle–road collaboration semi-physical simulation platform in intelligent transportation teaching experiments. The results show that this platform significantly enhances students’ understanding of vehicle–road collaboration technology, especially in the cultivation of application abilities in core modules such as path planning, environmental perception, and decision control. The high-fidelity simulation environment not only ensures safe experimental conditions but also effectively bridges the gap between academic training and industry requirements.

Currently, this platform still has some limitations, such as the need for improvement in simulating highly complex traffic scenarios. Future work will focus on further enhancing the platform’s functionality, including adding more diversified traffic scenarios and improving the authenticity of the simulation. Additionally, the research will explore the integration of this platform with advanced technologies such as AI algorithms and big data analysis to enhance students’ understanding and handling capabilities of the complexity of intelligent transportation systems. In conclusion, this study demonstrates the application value of the V2X vehicle–road collaboration semi-physical simulation platform in intelligent transportation teaching experiments, providing beneficial references and inspirations for future teaching models in related fields.