Active Excitations for Maximum Friction Coefficient Estimation

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

  • Nicolas Lampe
  • Karl-Philipp Kortmann
  • Clemens Westerkamp
  • Hans-Georg Jacob

Organisationseinheiten

Externe Organisationen

  • Hochschule Osnabrück
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des SammelwerksIEEE Intelligent Vehicles Symposium (IV 2023)
ISBN (elektronisch)979-8-3503-4691-6
PublikationsstatusVeröffentlicht - 2023

Publikationsreihe

Name IEEE Intelligent Vehicles Symposium
ISSN (Print)1931-0587
ISSN (elektronisch)2642-7214

Abstract

For optimizing advanced driver assistance systems (ADAS) and implementing autonomous driving, knowledge of vehicle dynamics and the perception of the vehicle's environment is required. A crucial parameter influencing vehicle dynamics is the maximum friction coefficient between tires and road. Since this coefficient cannot be measured practically without high technical effort, model-based estimation algorithms are used. However, estimating the maximum friction coefficient is only possible with sufficient vehicle dynamic excitation, as this coefficient is then observable. Since maneuvers with sufficient excitation are rare during normal driving, in this paper, different levels of active excitations are used to enable observability and estimation of the maximum friction coefficient during maneuvers with insufficient vehicle dynamic excitation. First, a vehicle dynamic model is presented and analyzed regarding the observability during active excitations. Second, model-based estimation using an unscented Kalman filter (UKF) is implemented for the test vehicle and the UKF parameters are tuned for active excitations. Finally, model-based maximum friction coefficient estimation using onboard vehicle sensors is enabled by using active excitations. The experimental results show that is possible to estimate the maximum friction coefficient with a low error as well as a low credibility for maneuvers with insufficient vehicle dynamic excitation by using active excitations.

ASJC Scopus Sachgebiete

Zitieren

Active Excitations for Maximum Friction Coefficient Estimation. / Lampe, Nicolas; Kortmann, Karl-Philipp; Westerkamp, Clemens et al.
IEEE Intelligent Vehicles Symposium (IV 2023). 2023. ( IEEE Intelligent Vehicles Symposium).

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Lampe, N, Kortmann, K-P, Westerkamp, C & Jacob, H-G 2023, Active Excitations for Maximum Friction Coefficient Estimation. in IEEE Intelligent Vehicles Symposium (IV 2023). IEEE Intelligent Vehicles Symposium. https://doi.org/10.1109/IV55152.2023.10186603
Lampe, N., Kortmann, K.-P., Westerkamp, C., & Jacob, H.-G. (2023). Active Excitations for Maximum Friction Coefficient Estimation. In IEEE Intelligent Vehicles Symposium (IV 2023) ( IEEE Intelligent Vehicles Symposium). https://doi.org/10.1109/IV55152.2023.10186603
Lampe N, Kortmann KP, Westerkamp C, Jacob HG. Active Excitations for Maximum Friction Coefficient Estimation. in IEEE Intelligent Vehicles Symposium (IV 2023). 2023. ( IEEE Intelligent Vehicles Symposium). doi: 10.1109/IV55152.2023.10186603
Lampe, Nicolas ; Kortmann, Karl-Philipp ; Westerkamp, Clemens et al. / Active Excitations for Maximum Friction Coefficient Estimation. IEEE Intelligent Vehicles Symposium (IV 2023). 2023. ( IEEE Intelligent Vehicles Symposium).
Download
@inproceedings{ac6ce6d08ba0483da23b0e8f170b5582,
title = "Active Excitations for Maximum Friction Coefficient Estimation",
abstract = "For optimizing advanced driver assistance systems (ADAS) and implementing autonomous driving, knowledge of vehicle dynamics and the perception of the vehicle's environment is required. A crucial parameter influencing vehicle dynamics is the maximum friction coefficient between tires and road. Since this coefficient cannot be measured practically without high technical effort, model-based estimation algorithms are used. However, estimating the maximum friction coefficient is only possible with sufficient vehicle dynamic excitation, as this coefficient is then observable. Since maneuvers with sufficient excitation are rare during normal driving, in this paper, different levels of active excitations are used to enable observability and estimation of the maximum friction coefficient during maneuvers with insufficient vehicle dynamic excitation. First, a vehicle dynamic model is presented and analyzed regarding the observability during active excitations. Second, model-based estimation using an unscented Kalman filter (UKF) is implemented for the test vehicle and the UKF parameters are tuned for active excitations. Finally, model-based maximum friction coefficient estimation using onboard vehicle sensors is enabled by using active excitations. The experimental results show that is possible to estimate the maximum friction coefficient with a low error as well as a low credibility for maneuvers with insufficient vehicle dynamic excitation by using active excitations.",
keywords = "Kalman filtering, active excitation, maximum friction coefficient, observability analysis, parameter estimation",
author = "Nicolas Lampe and Karl-Philipp Kortmann and Clemens Westerkamp and Hans-Georg Jacob",
note = "Funding Information: ACKNOWLEDGMENT The authors would like to thank the Dr. J{\"u}rgen and Irmgard Ulderup foundation for funding this project and ZF Friedrichshafen AG for their support during the test drives. In addition, the authors would like to thank Mohamed Elerian for his contribution on the conduct of his master thesis.",
year = "2023",
doi = "10.1109/IV55152.2023.10186603",
language = "English",
isbn = "979-8-3503-4692-3",
series = " IEEE Intelligent Vehicles Symposium",
booktitle = "IEEE Intelligent Vehicles Symposium (IV 2023)",

}

Download

TY - GEN

T1 - Active Excitations for Maximum Friction Coefficient Estimation

AU - Lampe, Nicolas

AU - Kortmann, Karl-Philipp

AU - Westerkamp, Clemens

AU - Jacob, Hans-Georg

N1 - Funding Information: ACKNOWLEDGMENT The authors would like to thank the Dr. Jürgen and Irmgard Ulderup foundation for funding this project and ZF Friedrichshafen AG for their support during the test drives. In addition, the authors would like to thank Mohamed Elerian for his contribution on the conduct of his master thesis.

PY - 2023

Y1 - 2023

N2 - For optimizing advanced driver assistance systems (ADAS) and implementing autonomous driving, knowledge of vehicle dynamics and the perception of the vehicle's environment is required. A crucial parameter influencing vehicle dynamics is the maximum friction coefficient between tires and road. Since this coefficient cannot be measured practically without high technical effort, model-based estimation algorithms are used. However, estimating the maximum friction coefficient is only possible with sufficient vehicle dynamic excitation, as this coefficient is then observable. Since maneuvers with sufficient excitation are rare during normal driving, in this paper, different levels of active excitations are used to enable observability and estimation of the maximum friction coefficient during maneuvers with insufficient vehicle dynamic excitation. First, a vehicle dynamic model is presented and analyzed regarding the observability during active excitations. Second, model-based estimation using an unscented Kalman filter (UKF) is implemented for the test vehicle and the UKF parameters are tuned for active excitations. Finally, model-based maximum friction coefficient estimation using onboard vehicle sensors is enabled by using active excitations. The experimental results show that is possible to estimate the maximum friction coefficient with a low error as well as a low credibility for maneuvers with insufficient vehicle dynamic excitation by using active excitations.

AB - For optimizing advanced driver assistance systems (ADAS) and implementing autonomous driving, knowledge of vehicle dynamics and the perception of the vehicle's environment is required. A crucial parameter influencing vehicle dynamics is the maximum friction coefficient between tires and road. Since this coefficient cannot be measured practically without high technical effort, model-based estimation algorithms are used. However, estimating the maximum friction coefficient is only possible with sufficient vehicle dynamic excitation, as this coefficient is then observable. Since maneuvers with sufficient excitation are rare during normal driving, in this paper, different levels of active excitations are used to enable observability and estimation of the maximum friction coefficient during maneuvers with insufficient vehicle dynamic excitation. First, a vehicle dynamic model is presented and analyzed regarding the observability during active excitations. Second, model-based estimation using an unscented Kalman filter (UKF) is implemented for the test vehicle and the UKF parameters are tuned for active excitations. Finally, model-based maximum friction coefficient estimation using onboard vehicle sensors is enabled by using active excitations. The experimental results show that is possible to estimate the maximum friction coefficient with a low error as well as a low credibility for maneuvers with insufficient vehicle dynamic excitation by using active excitations.

KW - Kalman filtering

KW - active excitation

KW - maximum friction coefficient

KW - observability analysis

KW - parameter estimation

UR - http://www.scopus.com/inward/record.url?scp=85167968511&partnerID=8YFLogxK

U2 - 10.1109/IV55152.2023.10186603

DO - 10.1109/IV55152.2023.10186603

M3 - Conference contribution

SN - 979-8-3503-4692-3

T3 - IEEE Intelligent Vehicles Symposium

BT - IEEE Intelligent Vehicles Symposium (IV 2023)

ER -