Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-14318
For citation please use:
Main Title: Robust numerical methods for robust control
Author(s): Benner, Peter
Byers, Ralph
Mehrmann, Volker
Xu, Hongguo
Type: Research Paper
URI: https://depositonce.tu-berlin.de/handle/11303/15545
http://dx.doi.org/10.14279/depositonce-14318
License: http://rightsstatements.org/vocab/InC/1.0/
Abstract: We present numerical methods for the solution of the optimal $H_{\infty}$ control problem. In particular, we investigate the iterative part often called the $\gamma$-iteration. We derive a method with better robustness in the presence of rounding errors than other existing methods. It remains robust in the presence of rounding errors even as $\gamma$ approaches its optimal value. For the computation of a suboptimal controller, we avoid solving algebraic Riccati equations with their problematic matrix inverses and matrix products by adapting recently suggested methods for the computation of deflating subspaces of skew-Hamiltonian/Hamiltonian pencils. These methods are applicable even if the pencil has eigenvalues on the imaginary axis. We compare the new method with older methods and present several examples.
Subject(s): H∞ control
algebraic Riccati equation
CS decomposition
Lagrangian subspaces
skew-Hamiltonian
Hamiltonian pencil
Issue Date: 27-Feb-2004
Date Available: 17-Dec-2021
Language Code: en
DDC Class: 510 Mathematik
MSC 2000: 93B40 Computational methods
93B36 H∞-control
Series: Preprint-Reihe des Instituts für Mathematik, Technische Universität Berlin
Series Number: 2004, 06
ISSN: 2197-8085
TU Affiliation(s): Fak. 2 Mathematik und Naturwissenschaften » Inst. Mathematik
Appears in Collections:Technische Universität Berlin » Publications

Files in This Item:
bbmx3_ppt.pdf
Format: Adobe PDF | Size: 345.24 kB
DownloadShow Preview
Thumbnail
bbmx3_ppt.ps
Format: Postscript | Size: 656.75 kB
Download

Item Export Bar

Items in DepositOnce are protected by copyright, with all rights reserved, unless otherwise indicated.