Model reference-based vehicle lateral control for lane departure avoidance

Neto, André Benine; Scalzi, Stefano; Mammar, Saïd; Netto, Mariana; Lusetti, Benoit · 2014 · Crossref

DOI: 10.1504/ijvas.2014.063044

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 practical implementation of a lane departure avoidance assistance system for passenger vehicles, addressing the safety issue of unintended lane departures caused by driver inattention or fatigue. The authors propose a dynamic state feedback controller that incorporates an internal model of road curvature to ensure the vehicle converges to the center of the lane with zero lateral offset, even on curved roads. The control strategy is designed to fully take over steering responsibility during emergency situations, based on ergonomic studies suggesting that shared control is ineffective when drivers lack time to react. The methodology utilizes a linear single-track vehicle model to capture lateral dynamics, assuming small wheel sideslip angles. To handle road curvature as a disturbance, the authors integrate an internal model consisting of double integrators of the lateral offset into the state vector. Controller synthesis is performed using Lyapunov theory and Bilinear Matrix Inequalities (BMI). This optimization minimizes the reachable invariant set of the vehicle states while satisfying constraints for bounded control inputs (to prevent actuator damage and tire saturation) and pole clustering (to enhance damping characteristics). The activation strategy relies on measuring driver steering torque; the system activates if the torque is below 5 N·m and the vehicle approaches the lane boundary, and deactivates if torque exceeds 2 N·m. The controller was validated through simulations in the CarSim environment and tested on a prototype vehicle on a test track. Simulation results confirmed the controller’s ability to reject disturbances and maintain stability under conditions sharper than those specified in the design phase, including lateral tire force saturation. Practical experiments on the test track demonstrated successful lane keeping performance. The computed state feedback gains ensured asymptotic stability, with eigenvalues located within the desired conic sector for damping. The system successfully maintained the vehicle within the lane despite road curvatures up to the design limit of 0.005 m⁻¹. The significance of this work lies in its robust control synthesis that guarantees stability and performance bounds through invariant set analysis, rather than relying solely on nominal performance. By integrating road curvature into the control model, the system achieves zero steady-state error on curves, a common limitation in simpler controllers. The successful transition from simulation to a physical prototype validates the practical applicability of BMI-based control synthesis for advanced driver assistance systems, offering a reliable solution for preventing accidents caused by unintended lane departures.

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