STABILIZATION BY NOISE REVISITED

被引：53

作者：

ARNOLD, L

机构：

[1] Institut für Dynamische Systeme, Universität Bremen, Bremen

来源：

ZEITSCHRIFT FUR ANGEWANDTE MATHEMATIK UND MECHANIK | 1990年 / 70卷 / 07期

关键词：

D O I：

10.1002/zamm.19900700704

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

Given a d × d matrix A and a d × k matrix B. We investigate how the stability behavior of ẋ = (A + BG(t)) × can be improved by a k × d matrix‐valued white noise process G(t). The analysis is based on the theory of Lyapunov exponents (exponential growth rates) and uses an averaging principle. It generalizes a result by the author and co‐workers for the case B = identity, in which the top Lyapunov exponent can be brought arbitrarily close to trace A/d by cleverly choosing G(t). Copyright © 1990 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim

引用

页码：235 / 246

页数：12

共 28 条

[1]

[Anonymous], 1987, DIFFUSIONS MARKOV PR

[2] LYAPUNOV EXPONENTS - A SURVEY [J].