An Adaptive Constrained Path Following Control Scheme for Autonomous Electric Vehicles

  • Yuhang Zhang
  • , Weida Wang
  • , Wei Wang
  • , Chao Yang*
  • , Yuanbo Zhang
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

To ensure driving safety, autonomous electric vehicles need to follow the planned path accurately, which depends on the vehicle path following control strategy. It is widely acknowledged that the model-based control strategies have excellent potential on path following, but their effect is seriously affected by the parametric uncertainty. To solve this problem, an adaptive constrained path following control scheme considering the influence of parametric uncertainty is proposed for autonomous electric vehicles. Firstly, an adaptive feedback control law and its update law are proposed to deal with the variation of tire cornering stiffness during vehicle path following process. Secondly, a constraint function for lateral displacement error during path following process is designed to further improve driving safety. And then, the closed-loop stability of the proposed control scheme is proved. Finally, the validation is implemented by simulation and experiment, the results show that the proposed control scheme is effective under different working conditions. The maximum lateral displacement error is reduced to 0.0297m by using the proposed scheme in the experiment. The proposed scheme might provide a theoretical reference for control practice of autonomous vehicles.

Original languageEnglish
Pages (from-to)3569-3578
Number of pages10
JournalIEEE Transactions on Vehicular Technology
Volume71
Issue number4
DOIs
Publication statusPublished - 1 Apr 2022

Keywords

  • adaptive constraint control
  • Autonomous vehicles
  • parametric uncertainty
  • path following

Fingerprint

Dive into the research topics of 'An Adaptive Constrained Path Following Control Scheme for Autonomous Electric Vehicles'. Together they form a unique fingerprint.

Cite this