A gain-based lower bound algorithm for real and mixed μ problems

Peter J Seiler Jr, Andrew Packard, Gary Balas

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

A new lower bound algorithm for real and mixed μ problems is presented. The basic idea of this algorithm is to use a related worst-case gain problem to compute the real blocks and, if the block structure is mixed, the standard power iteration to compute the complex blocks. Numerical tests indicate that the algorithm is fast and provides good lower bounds for both real and mixed μ problems of small to moderate size.

Original languageEnglish (US)
Pages (from-to)493-500
Number of pages8
JournalAutomatica
Volume46
Issue number3
DOIs
StatePublished - Mar 1 2010

Keywords

  • Robust stability
  • Structured singular value
  • Uncertain linear systems

Cite this