Connected and Automated Vehicle Platoon Formation Control via Differential Games

TL;DR

This paper develops a differential game framework for connected and automated vehicle platoon control, deriving Nash equilibria under various information topologies and validating the models through simulations.

Contribution

It introduces a novel differential game approach for CAV platooning, deriving closed-form solutions for specific topologies and analyzing general cases.

Findings

Closed-form Nash equilibria for PF and TPF topologies

Effective control strategies verified by simulations

Framework adaptable to various information structures

Abstract

In this study, the connected and automated vehicles (CAVs) platooning problem is resolved under a differential game framework. Three information topologies are considered here. Firstly, Predecessor-following (PF) topology is utilized where the vehicles control the distance with respect to the merely nearest predecessor via a sensor link-based information flow. Secondly, Two-predecessor-following topology (TPF) is exploited where each vehicle controls the distance with respect to the two nearest predecessors. In this topology, the second predecessor is communicated via a Vehicle-to-vehicle (V2V) link. The individual trajectories of CAVs under the Nash equilibrium are derived in closed-form for these two information topologies. Finally, general information topology is examined and the differential game is formulated in this context. In all these options, Pontryagin's principle is employed to investigate the existence and uniqueness of the Nash equilibrium and obtain its corresponding trajectories. In the general topology, we suppose numerical computation of eigenvalues and eigenvectors. All these approaches represent promising and powerful analytical representations of the CAV platoons under the differential games. Simulation experiments have verified the efficiency of the proposed models and their solutions.