Further results on the construction of invariant families of sets for linear systems with delay

R.H. Gielen, S.V. Rakovic, M. Lazar

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Citation (Scopus)
1 Downloads (Pure)

Abstract

Almost all synthesis methods for the construction of invariant sets for delay difference equations (DDEs) suffer either from computational complexity, i.e., those based on the Krasovskii approach, or come with considerable conservatism, i.e., those based on the Razumikhin approach. Only recently, the notion of an invariant family of sets was introduced in order to allow for a suitable trade¿off between conceptual generality and computational simplicity. In this paper, the construction of such families is further elaborated. In particular, a parametrization of families of sets is proposed that leads to tractable synthesis algorithms while preserving several advantages over the Razumikhin approach. For polyhedral sets, the corresponding synthesis algorithms can be implemented via an algebraic recursion and parametric linear programming. Thus, the invariant family of sets becomes a computationally tractable set invariance notion for DDEs, the advantages of which are illustrated via several examples.
Original languageEnglish
Title of host publicationProceedings of the 10th IFAC Workshop on Time Delay Systems, 22-24 June 2012, Boston, Massachussetts
Pages254-259
DOIs
Publication statusPublished - 2012
Event10th IFAC Workshop on Time Delay Systems (TDS 2012), June 22-24, 2012, Boston, MA, USA - Northeastern University, Boston, MA, United States
Duration: 22 Jun 201224 Jun 2012
http://www.coe.neu.edu/Research/IFACTDS2012/main.htm

Workshop

Workshop10th IFAC Workshop on Time Delay Systems (TDS 2012), June 22-24, 2012, Boston, MA, USA
Abbreviated titleTDS 2012
Country/TerritoryUnited States
CityBoston, MA
Period22/06/1224/06/12
Internet address

Fingerprint

Dive into the research topics of 'Further results on the construction of invariant families of sets for linear systems with delay'. Together they form a unique fingerprint.

Cite this