The representation and characterization of Drazin inverses of operators on a Hilbert space (Q2568389): Difference between revisions

Given a bounded linear operator \(T\) acting on a Hilbert space \(H\) with null space \(N(T)\) and range \(R(T)\), denote by \(T^{D}\) the Drazin inverse of \(T\) (\(TT^{D}=T^{D}T\), \(T^{D}TT^{D}= T^{D}\), \(T^{n+1}T^{D}=T^{n}\)); \(T^{D}\) exists iff there is the smallest non-negative integer \(k\) such that \(N(T^{k})=N(T^{k+1})\) and \(R(T^{k})=R(T^{k+1})\). The main result of the paper under review is that, in this case, if \( T \)=\( \left[\begin{matrix} T_{11} & T_{12}\\O & T_{22}\end{matrix} \right] \) then \(T^{D}= \left[\begin{matrix} T_{11}^{-1} & \Sigma_{i=0}^{k-1}T_{11}^{i-k-1}T_{12}T_{22}^{k-i-1}\\O & O \end{matrix}\right]\) with respect to the space decomposition \(H=R(T^{k})\bigoplus R(T^{k})^{\perp}\). It is the unique operator \(X=I- TT^{D}\) such that \(T^{k}X=O\), \(XT^{k}=O\), \(X^{2}=X\) and \(T^{k}+X\) is invertible.