Distributed Model Predictive Control for Heterogeneous Vehicle Platoons Under Unidirectional Topologies

Zheng, Yang; Li, Shengbo Eben; Li, Keqiang; Borrelli, Francesco; Hedrick, J. Karl · 2016 · OpenAlex-citations

DOI: 10.1109/tcst.2016.2594588

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of controlling heterogeneous vehicle platoons under unidirectional communication topologies where the desired set point is not universally known to all vehicles. While vehicle platooning offers benefits in traffic efficiency and safety, existing control strategies often rely on linear dynamics, ignore input constraints, or assume all vehicles have access to the leader’s state. The authors propose a Distributed Model Predictive Control (DMPC) algorithm that handles nonlinear vehicle dynamics, input constraints, and arbitrary unidirectional topologies without requiring global knowledge of the desired equilibrium. The methodology models the platoon as a group of dynamically decoupled but spatially constrained nodes. Each vehicle possesses nonlinear longitudinal dynamics, including engine, driveline, and aerodynamic effects, subject to torque constraints. The communication structure is represented by a directed graph, where information flows from preceding to downstream vehicles. The DMPC algorithm assigns each node a local open-loop optimal control problem based solely on information from its immediate neighbors and any direct connection to the leader. The cost function penalizes deviations from the desired equilibrium (if known), input effort, and discrepancies between predicted and assumed trajectories of neighbors. Crucially, the algorithm employs an equality-based terminal constraint that forces the terminal state of each node to equal the average of its neighbors’ assumed states, ensuring consensus. The study proves that asymptotic stability of the DMPC scheme is achieved under explicit sufficient conditions regarding the weights of the cost functions. By using the sum of local cost functions as a Lyapunov candidate, the authors demonstrate that the system converges to the desired formation even when only nodes directly connected to the leader know the target speed and spacing. Simulations involving passenger cars validate the effectiveness of the proposed DMPC, showing successful coordination and stability across various unidirectional topologies, such as predecessor-following and two-predecessor-leader following structures. The significance of this work lies in its ability to provide a robust control framework for realistic platoon scenarios where communication is limited and vehicle dynamics are nonlinear. Unlike previous methods that require centralized computation or global state knowledge, this distributed approach scales effectively and accommodates heterogeneous vehicle parameters. The findings highlight the critical role of communication topology in stabilizing the platoon and offer a practical solution for implementing autonomous vehicle platoons using existing vehicle-to-vehicle communication technologies.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• teleoperation remote driving