Platoon Control for Cyber–Physical Vehicle Systems With Intermittent Communication

Vehicle platoon control is one of the promising technologies to improve the performance of transportation systems. However, communication connections among vehicles in cyber–physical vehicle systems (CPVS) are often intermittent due to environmental factors and limitations of physical equipment. Moreover, the system models of the vehicles are generally unknown in practice, making it challenging to implement platoon control for CPVS. To address these challenges, we investigate the challenge platoon control problem for CPVS with intermittent communication, where the system models of both the leader vehicle and the follower vehicles are unknown. A data-driven-based learning algorithm is first designed to obtain the unknown leader model. Then, based on the learned leader model, a finite-time distributed observer is proposed to estimate the leader vehicle state in finite time under intermittent communication. Furthermore, with the unknown system models, a data-driven-based controller gain learning algorithm is proposed to learn the controller gain. Based on the learned controller gain, adaptive decentralized tracking controllers are designed to perform platoon control for CPVS. Finally, the effectiveness of our result is examined by a simulation example.