An introduction to the shift realization for finite dimensional continuous time systems (Q4955140)

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 1457722
Language Label Description Also known as
default for all languages
No label defined
    English
    An introduction to the shift realization for finite dimensional continuous time systems
    scientific article; zbMATH DE number 1457722

      Statements

      An introduction to the shift realization for finite dimensional continuous time systems (English)
      0 references
      0 references
      12 June 2000
      0 references
      continuous time systems
      0 references
      shift realization
      0 references
      Hankel map
      0 references
      time-domain realization
      0 references
      unstable systems
      0 references
      \textit{P. A. Fuhrmann} [``Algebraic system theory: An analyst's point of view'', J. Franklin Inst. 301, 521-540 (1976; Zbl 0332.93001)] constructed the so-called shift realization in terms of rational models, which provides also a minimal realization for matrix-valued proper rational functions. Recall that a rational function \(G(s)\), \(s\in {\mathbb C},\) is called proper if \(G\) has a finite limit at infinity. The frequency domain realization problem is, given a proper rational matrix-valued function \(G\), to represent \(G\) in the form \(G(s)=C(sI-A)^{-1}B+D\), \(s\in {\mathbb C}\backslash \sigma (A)\), where for some finite dimensional space \(X={\mathbb K}^n\) (here \({\mathbb K}\) equals \({\mathbb R}\) or \({\mathbb C}\)), \(A:X\rightarrow X\), \(B:{\mathbb K}^m\rightarrow X\), \(C:X\rightarrow {\mathbb K}^p\), \(D:{\mathbb K}^m\rightarrow {\mathbb K}^p\), i.e., as the transfer function of a continuous time-invariant linear system \[ \dot{x}(t)=Ax(t)+Bu(t),\quad y(t)=Cx(t)+Du(t),\quad x(t_0)=x_0 \] (\(t\in {\mathbb R}\), \(t\geq t_0)\). Such a realization \((A,B,C,D)\) is called minimal if \(X\) has a minimal possible dimension. In the paper under review, the author constructs the time-domain counterpart of Fuhrmann's shift realization. He introduces the notion of a Hankel map which is a generalization of the notion of a Hankel operator on \(L^2\) spaces. If \(M(t)\) is a continuous \(p\times m\) matrix-valued function on \({\mathbb R}_+\) the map \[ H_M:u(\cdot)\mapsto y(\cdot)=\int_{-\infty }^0M(\cdot -\tau)u(\tau) d\tau \] from the space of piecewise continuous \(m\)-dimensional vector-valued functions on \({\mathbb R}_-\), with a compact support, to the space of continuous \(p\)-dimensional vector-valued functions on \({\mathbb R}_+\), is called the Hankel map with symbol \(M\). The time-domain realization problem under consideration is under which conditions can a function \(M\) from \(C^{p\times m}({\mathbb R}_+)\) be written as the impulse response of a linear system \((A,B,C,D)\), i.e. \(M(t)=Ce^{tA}B\) for \(t\geq 0\). The main result of the paper asserts that the necessary and sufficient condition for this is that a Hankel map has a finite rank. The author constructs a minimal realization for an arbitrary finite-rank Hankel map \(H_M\), in which the state space \(X\) is the range of \(H_M\) (which is a subspace in \(C^{p\times 1}({\mathbb R}_+)\)), and \(A\) is the restriction on \(X\) of the left shift operator in \(C^{p\times 1}({\mathbb R}_+)\). Such a left shift realization has been extensively studied in the infinite dimensional situation for the continuous-time case, however with the additional stability assumptions. The author's result allows to consider also unstable systems. It is also shown, using the Laplace transform as a state space transformation, how to derive Fuhrmann's rational model for a transfer function from the time domain continuous-time shift realization.
      0 references

      Identifiers

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