Safety Monitoring and Alert for Neural Network-Enabled Control Systems

Lan, J. (2023) Safety Monitoring and Alert for Neural Network-Enabled Control Systems. In: 22nd IFAC World Congress (IFAC 2023), Yokohama, Japan, 09-14 Jul 2023, pp. 9436-9441. (doi: 10.1016/j.ifacol.2023.10.237)

[img] Text
293875.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.



This paper considers the safety monitoring and enhancement for neural network-enabled control systems with disturbance and measurement noise. A robustly stable interval observer is designed to generate sound lower and upper bounds of the system state. The obtained interval is used to monitor the runtime system state and predict the one-step ahead future system trajectory, providing system safety monitoring and alert. The simulation results of a numerical example and an adaptive cruise control system demonstrate efficacy of the observer in runtime system monitoring and its potentials in detecting sensor faults and enhancing system safety.

Item Type:Conference Proceedings
Additional Information:This work was supported by a Leverhulme Trust Early Career Fellowship under Award ECF-2021-517.
Keywords:Safety, neural network, observer, fault detection, intelligent autonomous vehicles.
Glasgow Author(s) Enlighten ID:Lan, Dr Jianglin
Authors: Lan, J.
College/School:College of Science and Engineering > School of Engineering > Autonomous Systems and Connectivity
Published Online:22 November 2023
Copyright Holders:Copyright © 2023 The Authors
First Published:First published in IFAC-PapersOnLine 56(2):9436-9441
Publisher Policy:Reproduced under a Creative Commons license

University Staff: Request a correction | Enlighten Editors: Update this record

Project CodeAward NoProject NamePrincipal InvestigatorFunder's NameFunder RefLead Dept
314249Decarbonising Machine Learning for Safe and Robust Autonomous SystemsDezong ZhaoLeverhulme Trust (LEVERHUL)ERC-2021-517ENG - Autonomous Systems & Connectivity