| | |
Summary: Quasi-Monte Carlo Methods in Robust Control
Design
Peter F. Hokayem, Chaouki T. Abdallah
, Peter Dorato
EECE Department, MSC01 1100
1 University of New Mexico
Albuquerque, NM 87131-0001, USA
{hokayem,chaouki,peter}@eece.unm.edu
Abstract
Many practical control problems are so complex that traditional anal-
ysis and design methods fail to solve. Consequently, in recent years prob-
abilistic methods that provide approximate solutions to such 'difficult'
problems have emerged. Unfortunately, the uniform random sampling
process usually used in such techniques unavoidably leads to clustering
of the sampled points in higher dimensions. In this paper we adopt the
quasi-Monte Carlo methods of sampling to generate deterministic samples
adequately dispersed in the sample-space. Such approaches have shown
to provide faster solutions than probabilistic methods in fields such as
Financial Mathematics.
1 Introduction
|
