Home // VEHICULAR 2023, The Twelfth International Conference on Advances in Vehicular Systems, Technologies and Applications // View article

A Neural Network-Based Estimation of Tire Self-Aligning Torque

Authors:
Younesse El Mrhasli
Bruno Monsuez
Xavier Mouton

Keywords: Tire Self-Aligning Torque, Estimation, Neural Network, Simulation.

Abstract:
Virtual sensing has attracted the interest of car makers and automotive service providers, owing to its cost-effective advantages, capacity to extract valuable insights from car data and its significance in enhancing the reliability of Advanced Driving Assistance Systems (ADAS). For instance, accurate virtual sensing of tire forces and torques can help adapt and improve the control strategies embedded in the vehicle's active safety systems. This paper deals with tire Self-Aligning Torque (SAT) estimation, an inherent parameter for identifying the limits of the vehicle at an early stage to prevent skidding. We present a data-driven approach to estimate the right and left front SATs, using a Neural Network (NN) model. The estimator takes directly existing in vehicle signals and does not rely on expensive and unpractical sensors, which makes it cost-efficient and fast. Simulation results based on a high-fidelity vehicle model show a good performance of the chosen NN to estimate the SATs while considering combined slip and road friction change.

Pages: 5 to 11

Copyright: Copyright (c) IARIA, 2023

Publication date: March 13, 2023

Published in: conference

ISSN: 2327-2058

ISBN: 978-1-68558-061-2

Location: Barcelona, Spain

Dates: from March 13, 2023 to March 17, 2023