On inversion of convolution integral operators on a finite interval

From MaRDI portal





The Gohberg-Heinig inversion formula \[ T^{-1}\varphi(t)=\varphi(t) -\int\limits_0^\tau \gamma(t,s)\varphi(s)\,ds\, ,\qquad 0 \leq t\leq\tau \] for a convolution operator \[ T\varphi(t):=f(t)-\int\limits_0^\tau k(t-s)f(s)\,ds\, ,\qquad k\in L^1_{m\times m}[-\tau,\tau]\, . \] on a finite interval \(0\leq t\leq\tau\) is computed explicitly for the case when the kernel function is given by \(k(t)=Ce^{tA}B\); here \(A\), \(B\) and \(C\) are constant matrices of the size \(n\times n\), \(n\times m\) and \(m\times n\) respectively. It is proved that the operator \(T\) is invertible if and only if the following \(n\times n\) matrix \(M\) is invertible \[ M:=I_n-\int\limits_0^\tau e^{-sA}BCe^{sA}\,ds\, , \] where \(I_n\) is the identity \(n\times n\) matrix. Moreover, the kernel \(\gamma(t)\) in the inversion formulae (1) is written explicitly with the help of the matrices \(A\), \(B\), \(C\) and \(M\). For the entire collection see [Zbl 1031.00013].











This page was built for publication: On inversion of convolution integral operators on a finite interval

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4464534)