Platooning Strategy for Connected and Autonomous Vehicles: Transition from Light Traffic

Abstract

This study presents a strategy for platoon formation and evolution of connected and autonomous vehicles (CAVs) in free-flow traffic. The proposed strategy is based on swarm intelligence, which describes bird flocking, fish schooling, and so on, in natural and artificial systems. In this concept, CAVs behave according to some rules to move together as a platoon without collisions. The rules are expressed by a spring–mass–damper system: CAV platoon formation and evolution are controlled by the spring constant and damping coefficient. Valid domains of these control parameters were derived on the basis of physical vehicle properties (e.g., bounded acceleration and deceleration) for realistic control. Furthermore, various relationships—maximum (in which the spring constant was set at its maximum for the given flow), quadratic, and cubic—between the control parameters and traffic flow were examined with simulations to obtain insight into desirable control parameter settings. The results suggest that the most efficient platooning can be achieved by the maximum relationship between the spring constant and flow with critical damping. However, the cubic relationship coupled with overdamping is more desirable in low-flow states to allow more freedom for vehicles to change lanes.