Hierarchical Event-Triggered Control for CAV Platoons Using Adaptive Neural Network

In this paper, a hierarchical event-triggered platoon control framework is developed for connected and automated vehicle (CAV) platoons with unknown dynamics and communication resource limitation. Firstly, a two-layer model that consists of an upper-layer model and a lower-layer model is introduced to depict the longitudinal kinematic and dynamic characterizes of CAVs in the platoon. Secondly, a double-layer event-triggered platoon control design framework is proposed for the considered CAV platoon. In the upper-layer, an event-triggered platoon controller is designed to generate desired platoon trajectory for CAV. A compensation signal is properly defined such that the platoon controller allows the spacing policy of the CAVs switching among constant spacing, constant time-headway, and variable time-headway policies. In the lower-layer, an adaptive neural network sliding-mode controller is developed to force the CAVs to track desired trajectory. An event-triggering condition that made up of data transmission error and vehicle consensus error is designed to determine data transmission sequence, such that a balance can be achieved between data transmission and platoon control performance. Finally, theoretical analysis is presented to illustrate CAV state stability and platoon string stability, and simulation studies are carried out to further demonstrate feasibility and effectiveness of the scheme.