Numerical laplace transformation and inversion in the analysis of physical systems (Q3691027)

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 3915586
Language Label Description Also known as
default for all languages
No label defined
    English
    Numerical laplace transformation and inversion in the analysis of physical systems
    scientific article; zbMATH DE number 3915586

      Statements

      Numerical laplace transformation and inversion in the analysis of physical systems (English)
      0 references
      0 references
      0 references
      1984
      0 references
      numerical inversion
      0 references
      quadrature
      0 references
      contour integration
      0 references
      sampling
      0 references
      ill-posed problem
      0 references
      Let F(s) be the Laplace transform of a function f(t), and assume that F(s) has no singularities in the right-half of the complex plane. Numerical Laplace transform analysis requires a discrete transform pair. This specifies a set of sampling points \(\{s_ 1,s_ 2,s_ 3,...\}\) in the complex frequency domain, a set of sampling instances \(\{t_ 1,t_ 2,t_ 3,...\}\) in the time domain, and a functional relationship (a matrix) between the column matrices \([F(s_ i)]\), \([f(t_ i)]\). Formulation of a discrete transform pair consists of finding some basis for choosing the sampling points and for determining values for the matrix coefficient. The object is to produce a transform pair which may be relied upon to give results of acceptable accuracy with a minimum computation in the overall analysis. This is an ill-posed problem and no claim to uniform superiority of a particular discrete transform pair can be made. The standard discrete transform pair is obtained by quadrature approximation of the continuous transform pair \(\{\) f(t),F(s)\(\}\) using the Bromwich contour and uniform integration meshes. The number of samples required is based on a specification of a time duration \(T_{\max}\) for which the response is required, and the band width \(f_{\max}\) needed to adequately resolve the function f(t). The number of samples required in either domain is then given by \(n=2f_{\max}T_{\max}\). The present paper presents a modified method, applicable to the transient analysis of physical systems, which is able to produce results of comparable accuracy to the standard method, but which achieves the specifications for \(f_{\max}\) and \(T_{\max}\) with fewer samples than are necessary with the standard method. The modification consists of using sampling instants whose separation increases systematically with the passage of time. The new discrete transform pair is heuristically developed from the standard discrete pair. Convergence and stability of the method are not addressed.
      0 references

      Identifiers