Sliding Mode Control for Robust Path Tracking of Automated Vehicles in Rural Environments

Jose Matute, Sergio Diaz, Ali Karimoddini*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Achieving robust path tracking is essential for efficiently operating autonomous driving systems, particularly in unpredictable environments. This paper introduces a novel path-tracking control methodology utilizing a variable second-order Sliding Mode Control (SMC) approach. The proposed control strategy addresses the challenges posed by uncertainties and disturbances by reconfiguring and expanding the state-space matrix of a kinematic bicycle model guaranteeing Lyapunov stability and convergence of the system. A state prediction is integrated into the developed SMC to mitigate response time delays. Furthermore, the controller integrates adaptive mechanisms to adjust time-varying parameters within the control formulation based on longitudinal velocity, thereby enhancing path-tracking performance and reducing chattering phenomena. The effectiveness of the proposed approach is comprehensively evaluated through simulations and experiments encompassing challenging driving scenarios characterized by high-curvature paths, varying altitudes, and sensor disturbances, typical in rural driving environments. Results demonstrate that disturbances have varying impacts depending on the type of sensor affected. Real-world tests validate these findings, offering practical insights for automated vehicle path-tracking implementation.

Original languageEnglish
Pages (from-to)1314-1325
Number of pages12
JournalIEEE Open Journal of Vehicular Technology
Volume5
DOIs
Publication statusPublished - 2024

Keywords

  • Automated Vehicles
  • Path Tracking
  • Robustness
  • Rural Environments
  • Sliding Mode Control
  • Vehicle Control

Fingerprint

Dive into the research topics of 'Sliding Mode Control for Robust Path Tracking of Automated Vehicles in Rural Environments'. Together they form a unique fingerprint.

Cite this