Joint estimation of vehicle states and spatio-temporal tyre–road friction using sequential Monte Carlo

Verfasst von

Björn Volkmann, Paul Kerzel, Jan Hendrik Ewering, Daniel Weber, Julian King, Thomas Seel, Simon F.G. Ehlers

Abstract

Accurately estimating and predicting tyre–road friction is essential for optimising safety-critical control strategies and task planning in advanced driver assistance and autonomous driving systems. Recent approaches propose the integration of camera and vehicle dynamics-based estimation methods. These hybrid approaches achieve more reliable estimation but cannot retain and recall this information for a location. They are also unable to quantify uncertainty in tyre–road friction with respect to changing road conditions. This paper proposes a novel scheme for estimating tyre–road friction under unknown road conditions. A 2D map of the environment is combined with a vehicle dynamics model to enable joint estimation of environment and vehicle state, including tyre–road friction. A tailored sequential Monte Carlo scheme enables online updating of map information and allows for recall upon revisiting a location. Results show that uncertainty in tyre–road friction can be reduced during successive trials of emergency braking. As a result, braking distance predictions become more accurate, with an RMSE reduction of 4.95 %, averaged across three road types.

Details

Organisationseinheit(en)
Institut für Mechatronische Systeme
Externe Organisation(en)
ZF Friedrichshafen AG
Typ
Artikel
Journal
Vehicle system dynamics
ISSN
0042-3114
Publikationsdatum
25.03.2026
Publikationsstatus
Elektronisch veröffentlicht (E-Pub)
Peer-reviewed
Ja
ASJC Scopus Sachgebiete
Fahrzeugbau, Sicherheit, Risiko, Zuverlässigkeit und Qualität, Maschinenbau
Elektronische Version(en)
https://doi.org/10.1080/00423114.2026.2645983 (Zugang: Offen )