A Reference Governor approach for Lateral Control of Autonomous Vehicles - Université Grenoble Alpes
Communication Dans Un Congrès Année : 2021

A Reference Governor approach for Lateral Control of Autonomous Vehicles

Résumé

This paper presents a novel approach on the union of lane tracking/change via the use of a Reference Governor (RG). This is achieved by feeding a modification of the reference during lane-change, called "virtual" reference, to the predesigned lane-tracking controller. The suggested methodology ensures reference is smooth enough so that the tracking controller does not overshoot and sustains fast convergence to the adjacent lane. The virtual reference is obtained as the solution of a convex optimization problem so that future closedloop states are enforced to belong in an admissible set. The efficiency of the proposed control-scheme is illustrated in highfidelity simulation results, for the case of an automated Renault Zoe. Consequently, the lane-tracking controller, a H ∞ yaw-rate tracker for which robustness certificates are presented, is able to perform a lane change maneuver when the RG is used.
Fichier principal
Vignette du fichier
ITS_2021.pdf (815.19 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03280581 , version 1 (07-07-2021)
hal-03280581 , version 2 (22-07-2021)

Identifiants

  • HAL Id : hal-03280581 , version 1

Citer

Dimitrios Kapsalis, Olivier Sename, Vicente Milanes, John Jairo Martinez Molina. A Reference Governor approach for Lateral Control of Autonomous Vehicles. 24th IEEE International Conference on Intelligent Transportation - ITSC2021, Sep 2021, Indianapolis, United States. ⟨hal-03280581v1⟩
95 Consultations
249 Téléchargements

Partager

More