Stochastic versions of the LaSalle theorem (Q1284433): Difference between revisions

The famous LaSalle theorem [see \textit{J. R. LaSalle}, J. Differ. Equations 4, 57-65 (1968; Zbl 0159.12002)] for locating limit sets of nonautonomous systems is generalized to the case of ordinary stochastic differential equations driven by \(m\)-dimensional Brownian motion under the hypothesis of local Lipschitz continuity and at most linear-polynomial growth of drift and diffusion parts. The proofs are carried out by a lemma due to \textit{R. Sh. Liptser} and \textit{A. N. Shiryaev} [ Martingale theory. Vyp. 38. Moskva: `Nauka'. (1986; Zbl 0654.60035)] on asymptotics of semimartingales, the well-known Kolmogorov-Centsov continuity theorem [\textit{I. Karatzas} and \textit{S. E. Shreve}, Brownian motion and stochastic calculus. Graduate Texts in Mathematics, 113. New York etc.: Springer-Verlag. (1991; Zbl 0734.60060; 1988; Zbl 0638.60065)], Burkholder type estimates and the uniform continuity of stochastic \(L^p\)-integrable martingales, and of course, by the help of Ito's formula. Thus, attracting deterministic sets, an estimation of exponential and polynomial growth rates can be found for stochastic differential equations. A series of examples illustrates the immense power of stochastic versions of LaSalle's theorem and striking importance of that author's work.