Lyapunov exponents of nilpotent Itô systems with random coefficients. (Q1766002)

From MaRDI portal





scientific article; zbMATH DE number 2138896
Language Label Description Also known as
English
Lyapunov exponents of nilpotent Itô systems with random coefficients.
scientific article; zbMATH DE number 2138896

    Statements

    Lyapunov exponents of nilpotent Itô systems with random coefficients. (English)
    0 references
    0 references
    0 references
    25 February 2005
    0 references
    The stochastic differential equation \[ \begin{aligned} \text{d}v_ {t} = \begin{pmatrix} 0 & a(x_ {t}) \\ 0 & 0 \end{pmatrix} v_ {t}\,\text{d}t &+ \varepsilon ^ 2 \begin{pmatrix} a_ {11}(x_ {t}) & a_ {12}(x_ {t}) \\ a_ {21}(x_ {t}) & a_ {22}(x_ {t}) \end{pmatrix} v_ {t}\,\text{d}t \\ &+ \varepsilon \sum ^ {r}_ {\alpha =1} \begin{pmatrix} b^ \alpha _ {11}(x_ {t}) & b^ \alpha _ {12}(x_ {t}) \\ b^ \alpha _ {21}(x_ {t}) & b^ \alpha _ {22}(x_ {t}) \end{pmatrix} v_ {t}\,\text{d}W^ \alpha _ {t} \tag{1} \end{aligned} \] is considered, where \((W^ 1,\dots ,W^ {r})\) is a standard \(r\)-dimensional Wiener process and \(x\) is a diffusion process on a finite-dimensional \(\sigma \)-compact manifold \(M\), independent of the Wiener process. Suppose that the generator of the process \(x\) is elliptic, has smooth coefficients and there exists an invariant probability measure \(\mu \) for \(x\). Let the functions \(a\) and \(c(\cdot ) = \sum ^ {r}_ {\alpha =1} | b^ \alpha _ {21}| ^ 2\) be smooth and \(\mu \)-integrable, \(\bar a = \int _ {M} a\,\text{d}\mu \neq 0\) and \(\bar c = \int _ {M} c\,\text{d}\mu >0\). Under some additional technical hypotheses upon the functions \(a\), \(a_ {ij}\), \(b^ \alpha _ {ij}\) and the operator \(G\), it is shown that the top Lyapunov exponent \(\lambda _ \varepsilon \) of (1) satisfies \(\lambda _ \varepsilon = \varepsilon ^ {2/3}| \bar a| ^ {2/3} \bar c^ {1/3}\lambda _ 0 + O(\varepsilon ^ {4/3})\) as \(\varepsilon \to 0\), where \(\lambda _ 0\) is the top Lyapunov exponent of the equation \[ \text{d}v_ {t} = \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix} v_ {t}\,\text{d}t + \begin{pmatrix} 0 & 0 \\ 1 & 0 \end{pmatrix} v_ {t}\,\text{d}W_ {t}. \]
    0 references
    0 references
    stochastic differential equations
    0 references
    nilpotent systems
    0 references
    Lyapunov exponents
    0 references

    Identifiers