Distributed Cooperative Control and Optimization of Connected Automated Vehicles Platoon Against Cut-in Behaviors of Social Drivers

TL;DR

This paper presents a cooperative control framework for connected automated vehicle platoons that accounts for unpredictable lane changes of human-driven vehicles, enhancing safety and efficiency in mixed traffic environments.

Contribution

It introduces a three-stage dynamic platoon management framework utilizing C-V2X communication, trajectory estimation, and cooperative control to handle social driver cut-in behaviors.

Findings

Effective trajectory prediction of human-driven vehicles.

Improved safety and efficiency in mixed traffic platoons.

Validation through simulation cases.

Abstract

Connected automated vehicles (CAVs) have brought new opportunities to improve traffic throughput and reduce energy consumption. However, the uncertain lane-change behaviors (LCBs) of surrounding vehicles (SVs) as an uncontrollable factor significantly threaten the driving safety and the consistent movement of a group of platoon CAVs. How to ensure safe, efficient, and fuel economic platoon control poses a key challenge faced by researchers in complex traffic environments. This study proposes a dynamic platoon management and cooperative driving framework for a mixed traffic flow consisting of multiple CAVs and possible human-driven vehicles (HDVs) as the SVs on unsignalized roads. In the proposed framework, the leader CAV of the platoon provides a high-level automatic driving decision to the follower CAVs by developing an optimal trajectory estimation of the HDVs while distributed observers and tracking controllers are properly implemented by the follower CAVs. Specifically, the proposed framework consists of three stages. At the observation stage, the cruising information of all the SVs will be collected by the leader CAV through the Cellular-Vehicle-to-X (C-V2X) infrastructure, while an automatic decision-making driving assistance system (ADMDSS) is constructed to determine the driving states of the platoon. When the HDVs approach the communication range of the platoon, in the prediction stage, the trajectories of the HDVs as the target vehicles will be estimated and the reference trajectory planning for the leader CAV and the cooperative controller design for the follower CAVs will be respectively activated by using C-V2X infrastructure. Simulation cases are presented to illustrate the effectiveness of the proposed approaches.