Driver steering assistance for lane departure avoidance

Enache, N. Minoiu; Netto, M.; Mammar, S.; Lusetti, B. · 2008 · OpenAlex-citations

DOI: 10.1016/j.conengprac.2008.10.012

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper presents the design and experimental validation of a steering assistance system intended to prevent unintended lane departures caused by driver inattention, fatigue, or illness. Motivated by the significant proportion of fatal crashes attributed to running off the road, the study addresses the challenge of integrating automation with conventional steering columns without causing intrusive conflicts with the driver. The system is designed to intervene only when necessary, ensuring the vehicle remains within lane borders while converging toward the centerline, all while maintaining passenger comfort through bounded control inputs and vehicle dynamics. The methodology employs a fourth-order linear "bicycle" vehicle model coupled with a second-order model for an electrically powered steering column. The control strategy is built upon Lyapunov theory and Linear Matrix Inequalities (LMI) optimization. A switching strategy determines whether control is assigned to the driver or the assistance system based on driver torque thresholds and the vehicle’s position relative to a mathematically defined "normal driving" zone. This zone is characterized by a polyhedral set in the state space, ensuring that state variables and lateral offsets remain within safe bounds. The control law is designed to guarantee asymptotic stability, minimize overshoot relative to the lane center, and bound the assistance torque. Robustness against variations in vehicle speed and road adhesion is achieved through gain scheduling and matrix polytope techniques. Experimental tests were conducted on a prototype passenger vehicle equipped with sensors for side slip angle, yaw rate, steering angle, and driver torque, as well as a vision-based lane detection system. Two activation strategies were evaluated. The results demonstrated that the assistance system successfully kept the vehicle within the lane borders during simulated lapses of attention. Specifically, the front wheels remained within the guaranteed overshoot limits (±1.76m) and did not cross the actual lane boundaries (±1.75m). The system exhibited good robustness, with closed-loop poles indicating stable dynamics across the tested speed range of 18–22 m/s. While both activation strategies were deemed appropriate, the second strategy showed superior reactivity in cases of rapid drifting. The significance of this work lies in providing a theoretical framework for handling driver-assistance interactions in switched systems, ensuring bounded dynamics and safety. By using LMI-based optimization, the study guarantees specific performance metrics, such as limited wheel displacement and torque, which are critical for passenger comfort and safety. The successful experimental validation confirms that such systems can effectively prevent lane departures due to inattention without requiring complex infrastructure, offering a viable approach for active safety systems in conventional vehicles.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• lane positioning
• steering pattern

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model