CAV Platoon Formation in Mixed Traffic: Analytical Control Design and Realistic Numerical Validation

Abstract: Truck platooning is a promising solution to reduce freeway congestion, engine energy consumption, and associated emissions. However, we still lack efficient local-level algorithms to form a truck platoon without jeopardizing the surrounding traffic flow’s safety and efficiency. Motivated by this research gap, this study developed a Sequential truck Platoon Formation (StPF) algorithm implemented on individual Connected and Autonomous Trucks (CAT) to locally form, grow, and merge platoons in mixed traffic. The StPF algorithm is built upon a novel Complex-Spring-Mass-Damper (SMD) Reactive Controller that drives a subject CAT to approach its target leader by modeling the interactions to the target leader and surrounding micro- and macro- traffic as a system of unique non-linear springs and dampers. Theoretical analysis of the complex-SMD controller verified that the platoon formation will be completed in a reasonable and practical spatial-temporal range and provided mathematical guidance on the parameter setting and leader selection for practical implementation. Moreover, the study designed the StPF Algorithm as a comprehensive protocol network. This enables StPF to coordinate platoon formation requests from multiple CATs in mixed traffic by self-organizing into independent clusters, conducting platoon formation in parallel under proper circumstances, while adapting to extreme scenarios and edge cases by seamlessly switching between complex-SMD and its variations. Numerical experiments using real-world traffic data identified that StPF is most effective at free-flow (Level of Service (LOS) A) and low traffic (LOS B) at a high CAT penetration rate (>7.5%) and confirmed the benefits to the road segment and the individual CAT due to platoon formation enabled by StPF.