Event-triggering $${H}_{\infty }$$-based observer combined with NN for simultaneous estimation of vehicle sideslip and roll angles with network-induced delays

Boada, Maria Jesus L.; Boada, Beatriz L.; Zhang, Hui · 2021 · Crossref

DOI: 10.1007/s11071-021-06269-7

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 critical need for accurate estimation of vehicle sideslip and roll angles, which are essential for lateral stability and rollover prevention in advanced driver-assistance systems and autonomous vehicles. Since these variables cannot be directly measured by cost-effective sensors in production vehicles, the authors propose a robust estimation method that accounts for vehicle nonlinearities, parameter uncertainties, and network-induced delays. The study specifically targets networked control systems (NCS) where communication bandwidth limitations necessitate efficient data transmission strategies. The methodology combines an event-triggered $H_\infty$-based observer with neural networks (NN) and linear-parameter-varying (LPV) techniques. The vehicle dynamics are modeled using an LPV framework to handle nonlinearities, with uncertainties in tire cornering stiffness addressed through bounded parameter variations. Neural networks generate "pseudo-measures" of the sideslip and roll angles using data from standard sensors (GPS, IMU, steering wheel). To reduce network load, an event-triggered communication scheme is implemented, transmitting data packets only when a specific error threshold is violated, thereby avoiding unnecessary transmissions while managing time delays. The observer design ensures robust stability and $H_\infty$ performance, minimizing the effect of disturbances and delays on estimation accuracy. The study validates the proposed observer through both simulation and experimental results. The design guarantees asymptotic stability under the event-triggered scheme and satisfies the $H_\infty$ performance index, ensuring that the estimation error remains within acceptable bounds despite network delays and parameter uncertainties. The integration of NN allows the system to handle strong nonlinearities, while the LPV approach manages time-varying characteristics like longitudinal speed. The event-triggered mechanism successfully reduces data transmission frequency without compromising the quality of the state estimates. The significance of this work lies in its practical applicability to modern vehicle architectures. By utilizing existing low-cost sensors and reducing communication overhead, the proposed method offers a viable solution for real-time state estimation in resource-constrained NCS. The ability to simultaneously estimate sideslip and roll angles with robustness against delays and uncertainties enhances the reliability of stability control systems, contributing to improved vehicle safety and the broader adoption of autonomous driving technologies.

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.

• situational awareness