With the growing demand for the integrated development of vehicle intelligence and connectivity, China has taken the lead in proposing the Vehicle-Road-Cloud integration development path, aiming to accelerate the deep integration and exploration of vehicle intelligence and connectivity. Grounded in systems engineering methodology, the paper adopts the Vehicle-Road-Cloud Integrated System (VRCIS) as its research framework. First, it explores the benefit of VRCIS development and evaluates the feasibility of application scenarios to address current technical problems. Secondly, it classifies the application scenarios, based on the connected capabilities of the VRCIS and summarizes the table of connected scenarios, by reviewing international research and practices. Thirdly, the paper analyzes the three developmental stages of cooperative connected autonomous driving and generates corresponding functional scenarios based on the summarized connected scenarios. Finally, the development path for the Vehicle-Road-Cloud integrated scenarios is proposed, guiding future technological evolution and implementation.

The technology of Simultaneous Localization And Mapping（SLAM）is the hot and difficult point of current research. It is one of the key technologies to realize unmanned driving. The paper introduction is centered on the principle, basic structure, sensors used and map types of SLAM and analyzes the advantages, disadvantages and scope of application of various fusion methods. Secondly, according to the different fusion methods used, it introduces the current status of research at home and abroad, and points out its worthwhile affirmation and continuation of the technology as well as shortcomings. Finally, it summarizes the current problems of multi-sensor fusion unmanned SLAM research, analyzes the research difficulties and puts forward suggestions on the possible development direction of future research, with the aim of providing references for the development of unmanned driving.

With the development trend of the “New Four Modernizations” in the automotive industry, the demand for intelligent and personalized cockpit is increasing. Through data-driven technology, scenarios can be generated unsupervised in intelligent cockpits to enhance user experience. Construct an unsupervised scenario generation model using Apriori association rule algorithm and multi-indicator automatic filtering mechanism. The results indicate that this technology can automatically generate adaptive scenario based on environmental data and user behavior without manual intervention, demonstrating broad application potential in enhancing flexibility and personalized service levels in intelligent cockpit.

Combing the existing engine product development test standards systematically, bechmarking the advanced test standards and test methods, analyzing multi-dimensional user scenario from the user driving purposes, road topology, temperature, humidity, user driving characteristics,then establish the engine test verification system facing the user complex use environment and working condition,all of the above is for improving the quality of engine products. Engine product development tests cover all user working conditions, including extreme working condition test, limit size test, extreme environment test, enhanced acceleration test, inspection compliance test, and user simulation test. For the special characteristics of products, development tests including the parts level, the system level, the machine level and the vehicle level are used to achieve the coverage of user scenarios of more than 95%, and ensure the high quality requirements of high reliability and low failure rate of the engine.

Driving style recognition plays a crucial role in enhancing personalized driving experiences and optimizing energy utilization in smart connected vehicles. Considering the relationship between different road environments and driving styles, a cascade delivery framework is designed to fully utilize real-world natural driving data and segment the data into events with distinct physical meanings. Using driver IDs as pseudo-labels, an XGBoost model learns differences in driving styles, identifying the key features and weights critical for recognition. Following the principles of a hybrid expert system, the WK-means algorithm clusters driving styles under varying conditions, ultimately generating driving scores to evaluate driver performance. The statistical analysis of the clustered data shows that this method effectively recognizes drivers with diverse driving styles, which lays the foundation for the further development of intelligent networked vehicle technology.

The application of digital projection technology promotes the intelligent development of automobile headlights. In order to explore the new generation of car headlight technology, this article outlines the development history of automotive headlights. By studying the current status of digital projection headlights, three digital projection headlight technology solutions, DLP, Mirco LED, and LCD display, are compared and analyzed. The development trend of array projection headlights is also discussed, in order to provide reference for related research.

To address the research gap in subjective evaluation systems for in-vehicle recommendation systems, this study proposes a comprehensive and scientifically grounded evaluation framework based on key features such as richness, interaction design, experience design, and response speed. The study explores assessment criteria across different technical backgrounds, usage scenarios, and recommendation algorithms. The results indicate that this evaluation system can effectively guide the design and optimization of in-vehicle recommendation systems, providing a basis for enhancing driving experience and user satisfaction.

To address the design optimization requirements of MacPherson strut independent suspension systems, a dynamic model of this suspension system is constructed based on multibody dynamics theory, considering its unique geometric configuration and parameter variation patterns. The motion characteristics are revealed through the derivation of differential equations. The results demonstrate that the established mathematical model can accurately characterize the dynamic relationships among various components during the suspension’s motion, thereby validating the effectiveness of the parametric modeling approach. This study provides a theoretical foundation and computational basis for performance prediction, structural optimization of MacPherson strut independent suspension systems, and the evaluation of their impact on overall vehicle dynamic characteristics.

In order to analyze the crucial role of the Steer-By-Wire (SBW) system in intelligent driving and its integrated application with the chassis control system, this paper focuses on the architecture and functions of the SBW system, including the fully redundant hardware architecture and variable steering ratio function. It also explores the vehicle application of the system, such as the calibration of the variable steering ratio and the matching of road-feel feedback. The paper also proposes an optimal basic road-feel feedback function scheme, as well as processes and methods for vehicle application suitable for the variable steering ratio and road-feel feedback of the SBW system, providing technical support for the development and engineering application of the SBW system.

Firstly, this paper reviews the definitions of the key states of lithium-ion batteries, including State of Charge (SOC), State of Power (SOP), State of Function (SOF), State of Energy (SOE), State of Health (SOH), Remaining Useful Life (RUL), State of Temperature (SOT) and State of Safety (SOS), and analyzes their coupling relationships. Then, it classifies and elaborates the methods for joint estimation of battery double states. In the future, multi-state joint estimation can further improve the estimation accuracy. Advanced sensor technologies, such as fiber-optic sensors, can more accurately measure the internal state quantities of batteries. At present, battery group state estimation is mostly focused on individual cells, and it is necessary to further explore the joint estimation at the battery module and group levels. Given the nonlinear characteristics of lithium-ion batteries, machine learning can achieve higher estimation accuracy with relatively low complexity. With the development of big data and cloud technologies, new-type battery state estimation will become a trend.