Can lane-changing preferences of connected and autonomous vehicles induce spontaneous platoon formation?

Connected and Automated Vehicles (CAVs) achieve a headway gain when following other CAVs. This study defines two key factors: the CAV cluster factor and the headway gain. We analytically examined the capacity of a single-lane Manual Vehicle (MV)-CAV mixed flow related to these factors and the CAV market penetration rate. Furthermore, this paper reveals that this headway gain induces a preferential lane-changing behavior in CAVs, making them more inclined to choose to follow other CAVs. This preference promotes the spontaneous formation of CAV platoons. By analyzing four types of lane-changing scenarios, the study derives an expression for the CAV cluster factor in multi-lane MV-CAV mixed flows. Simulation experiments validated the theoretical findings and demonstrated that the CAV cluster factor in multi-lane mixed flows is primarily determined by the headway gain and the CAV market penetration rate. It exhibits independence from the specific longitudinal control laws of CAVs, the car-following behavior of MVs, and whether lane-changing decisions incorporate the speed of the leading vehicle. Although CAVs’ lane-changing preferences can contribute to the formation of larger spontaneous platoons, this effect remains limited. Under higher traffic densities, lane-changing opportunities are significantly diminished, which inhibit the influence of lane-changing preferences, causing the degree of CAV clustering in multilane flows to approach that observed in single-lane scenarios.