MPC Strategies for Cooperative Guidance of Autonomous Vehicles

Bertrand, Sylvain; Marzat, Julien; Piet-Lahanier, Hélène; Kahn, Arthur; Rochefort, Yohan · 2014 · OpenAlex-citations

DOI: 10.12762/2014.al08-11

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 presents a comprehensive framework for the cooperative guidance of autonomous vehicle fleets using Model Predictive Control (MPC). The research addresses the complexity of missions such as large-area surveillance and multi-target tracking, which are often difficult for single vehicles or human operators. By splitting tasks among multiple vehicles, the authors aim to improve feasibility and efficiency. The study focuses specifically on designing reactive and distributed control laws, where each vehicle computes its own control inputs based on shared state information, assuming perfect, delay-free communication among the fleet. The methodology relies on a distributed MPC approach where each vehicle solves an optimization problem over a finite prediction horizon. The system models vehicles as point masses moving in a 2D plane, with dynamics defined by position, velocity, and heading, subject to constraints on speed and acceleration. The core of the strategy is a composite cost function $J_i$ for each vehicle, comprising navigation, safety, and control effort terms. The navigation cost regulates speed and guides vehicles toward waypoints using reference trajectories and balls, while also maintaining fleet cohesion through a specific cost term that penalizes deviations from desired inter-vehicle distances. The safety cost prevents collisions with obstacles and other vehicles by penalizing predicted distances that fall below safe thresholds. Additionally, a control cost limits energy consumption by penalizing excessive control inputs. The paper details the mathematical formulation of these costs, including specific weighting coefficients and hyperbolic tangent functions to manage fleet formation and avoidance behaviors smoothly. It also discusses the computational aspects of solving the resulting constrained nonlinear programming problem, noting that while global optimization methods exist, numerical techniques like Sequential Quadratic Programming are preferred for real-time implementation due to computational constraints. The authors propose an efficient method for selecting optimal costs within limited computation time. The significance of this work lies in providing a unified MPC-based solution for common cooperative control challenges, including collision avoidance, formation flying, and area exploration. By defining specific cost functions that balance mission objectives with safety and coordination, the framework ensures safe collaboration among autonomous vehicles. The performance of the proposed approach is validated through simulation results, demonstrating its effectiveness in managing complex cooperative missions for Unmanned Aerial Vehicles (UAVs). This contribution offers a robust, distributed control strategy that enhances the autonomy and reliability of vehicle fleets in demanding operational environments.

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