A spectral characterization of \(H^{\infty}\)-optimal feedback performance and its efficient computation (Q1087510): Difference between revisions

The problem considered in this paper is that of giving an efficient computational scheme for the mixed-sensitivity optimization problem \[ \inf \{\| \left[ \begin{matrix} G-UZ\\ F\end{matrix} \right]\|: z\in H^{\infty}_{n\times m}\} \] where \(G\in H^{\infty}_{n\times m}\), \(F\in H^{\infty}_{m\times m}\), \(U\in H^{\infty}_{n\times n}\). The approach reduces this problem to that of computing the spectral radius of a self-adjoint Hankel and Toeplitz operator. The particular structure of this operator allows the computation of the spectral radius by computing the largest eigenvalue.