Vehicle Lateral Dynamics Estimation using Switched Unknown Inputs Interval Observers: Experimental Validation

Ifqir, Sara; Oufroukh, Naima Ait; Ichalal, Dalil; Mammar, Said · 2018 · Crossref

DOI: 10.23919/acc.2018.8431486

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 addresses the challenge of robustly estimating vehicle lateral dynamics—specifically yaw rate and lateral velocity—in the presence of parameter uncertainties and unknown inputs. Existing estimation methods often rely on deterministic analyses that assume precise knowledge of vehicle parameters, which is unrealistic in field conditions where factors like road friction, load variations, and tire saturation cause significant uncertainties. The authors propose a Switched Unknown Input Interval Observer (SUIIO) to estimate the upper and lower bounds of the state vector, treating road curvature as an unknown input and cornering stiffness as bounded uncertainties. The methodology employs a bicycle model for vehicle lateral dynamics combined with a vision system measurement model. To account for variations in longitudinal velocity, the system is represented as a switched uncertain linear system. The observer design utilizes Multiple Quadratic Input-to-State Stable (ISS) Lyapunov functions. Sufficient conditions for the existence of the observer are derived and formulated as Linear Matrix Inequalities (LMIs). This approach ensures that the observer remains stable and provides bounded estimates despite the presence of unknown inputs and parameter variations, without requiring exact knowledge of the uncertain parameters. Experimental validation was conducted using real data from a prototype vehicle equipped with sensors for yaw rate, steering angle, longitudinal speed, and a vision system. The results demonstrate that the proposed SUIIO successfully estimates the upper and lower bounds of the vehicle's lateral dynamics. The estimation scheme proved effective in handling unknown inputs and parameter uncertainties, providing accurate bounds for the state variables even under varying road adhesion conditions and longitudinal speeds. The significance of this work lies in providing a systematic design methodology for interval observers tailored to switched systems with unknown inputs, a problem area with limited existing literature. By offering a robust estimation technique that does not rely on precise parameter knowledge, this approach enhances the reliability of state estimation for autonomous vehicles operating in unstructured environments. The use of LMIs for observer design facilitates practical implementation, and the experimental validation confirms the method's effectiveness in real-world scenarios, contributing to the development of safer and more robust autonomous driving systems.

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