Non-linear filtering of rare events with large signal-to-noise ratio (Q3776314)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 4036826
Language Label Description Also known as
default for all languages
No label defined
    English
    Non-linear filtering of rare events with large signal-to-noise ratio
    scientific article; zbMATH DE number 4036826

      Statements

      Non-linear filtering of rare events with large signal-to-noise ratio (English)
      0 references
      1987
      0 references
      nonlinear filtering
      0 references
      small noise diffusions
      0 references
      global boundedness
      0 references
      iven a generalized random field
      0 references
      generalized Gaussian field with independent increments
      0 references
      Lévy-Baxter theorem
      0 references
      spectral density
      0 references
      0 references
      The paper considers the problem of estimating the signal given by \[ dx^{\epsilon}=b(x^{\epsilon})dt+\epsilon \sigma (x^{\epsilon})dW_ 1,\quad x^{\epsilon}(0)=x_ 0, \] given observations from \[ dy^{\epsilon}=h(x^{\epsilon})dt+\epsilon dW_ 2,\quad y^{\epsilon}(0)=0, \] in a finite interval [0,T], under global boundedness and Lipschitz conditions on b, \(\sigma\) and h. The robust form of the iven a generalized random field X satisfying the equation \(PX=f\), where P is a symmetric, positive differential operator of order 2p with constant coefficients and f is a generalized Gaussian field with independent increments. It is assumed that f is a zero mean process with covariance operator P. Under some weak assumptions on the operator P an analogue of the Lévy- Baxter theorem for the process X is proved. As its consequence one obtains the following: Let \(q=4p\). Then all coefficients of the operator P are uniquely determined by the trajectories of X in any arbitrarily small subdomain of T and thus in case of X stationary its spectral density is determined also.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references