Lyapunov stability-based optimal control of following vehicles on a highway

Azzedine Yahiaoui

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

The application of modern control methods to automated driving of following vehicles on a highway can make the advanced transportation systems much more safe and efficient. This paper focuses on a computational approach to optimal control and stability analysis of the adjacent vehicles following in a single traffic lane of a highway in order to ensure the functional safety and assistance of automated driving systems. These vehicles are entirely automated as they have the ability to follow a line path at a high speed and reduce the distance that separates them while adjusting their responses automatically without intervention from humans (i.e. drivers). To meet such requirements for driving safety, this paper explores a new approach to string stability of following vehicles that travel in a straight line of a highway, using optimal control and Lyapunov stability theories. Consequently, sufficient conditions for string stability of a vehicle traffic using continuous-time Lyapunov function and matrix Riccati equation are presented. The search for a quadratic Lyapunov function is then formulated as a convex optimisation problem in terms of a linear invariant system towards a quadratic optimal criterion. The analysis of these methods for minimising the performance criterion is also given. Finally, the paper ends with a conclusion and perspective for future work.

Original languageEnglish
Pages (from-to)196-211
Number of pages16
JournalInternational Journal of Vehicle Autonomous Systems
Volume14
Issue number2
DOIs
Publication statusPublished - 1 Nov 2018

Keywords

  • Control system analysis
  • Lyapunov function
  • Matrix riccati equation
  • Safe driving of vehicles
  • String stability analysis

Fingerprint

Dive into the research topics of 'Lyapunov stability-based optimal control of following vehicles on a highway'. Together they form a unique fingerprint.

Cite this