Automated Vehicle Platooning: A Two-Stage Approach Based on Vehicle-Road Cooperation

Within the advancing realm of intelligent transportation systems, vehicle platooning stands out as an innovative approach to improve road efficiency and reduce energy consumption. However, forming platoons swiftly and accurately under the fluctuating conditions of dynamic traffic, while optimizing the interests of platoon members, presents considerable challenges. Acknowledging the constraints of prior research, which often relies on ideal hypothetical setups, this study unveils a practical two-phase platooning methodology within the Vehicle-Road Cooperation (VRC) framework, making it aptly suited for real-world platooning instances. In the initial phase, the Random Forest algorithm combined with Fuzzy C-Means (FCM) clustering is employed to classify vehicles into uniform platoons, corroborated by data from the Kaggle dataset. Following this, platoon leadership is elected via a distributed voting mechanism based on Raft consensus, leveraging the Simple Tree Matching (STM) algorithm to evaluate the extent of path alignment among vehicles. The method's effectiveness in forming platoons efficiently and synchronizing the interests of their members is verified through experiments in dynamic traffic environments, demonstrating its potential as a solid foundation for intelligent transportation systems.