Interval observer design of dynamical systems with neural networks

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper proposes an interval observer design method to construct lower-bound and upper-bound of system state trajectories in run time. The developed interval observer consists of two auxiliary neural networks derived from the neural network in dynamical systems, and two observer gains to ensure the positivity and the convergence of the corresponding error dynamics. Particularly, if the neural network is driven by the output of the system, the developed approach contains a promising neural-network-free design feature. The developed method is validated with evaluations on an adaptive cruise control system with a neural network controller.

Original languageEnglish (US)
Title of host publicationHSCC 2021 - Proceedings of the 24th International Conference on Hybrid Systems
Subtitle of host publicationComputation and Control (part of CPS-IoT Week)
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9781450383394
DOIs
StatePublished - May 19 2021
Event24th ACM International Conference on Hybrid Systems Computation and Control, HSCC 2021, held as part of the 14th Cyber Physical Systems and Internet-of-Things Week, CPS-IoT Week 2021 - Virtual, Online, United States
Duration: May 19 2021May 21 2021

Publication series

NameHSCC 2021 - Proceedings of the 24th International Conference on Hybrid Systems: Computation and Control (part of CPS-IoT Week)

Conference

Conference24th ACM International Conference on Hybrid Systems Computation and Control, HSCC 2021, held as part of the 14th Cyber Physical Systems and Internet-of-Things Week, CPS-IoT Week 2021
Country/TerritoryUnited States
CityVirtual, Online
Period5/19/215/21/21

Keywords

  • dynamical systems
  • interval observer
  • neural networks
  • runtime monitoring

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Interval observer design of dynamical systems with neural networks'. Together they form a unique fingerprint.

Cite this