Answers are grounded in this paper and cite their sources.
A class of (block) rational Krylov-subspace-based projection methods for solving the large-scale continuous-time algebraic Riccati equation (CARE) 0 = R(X) := AHX + XA + CHC - XBBHX with a large, sparse A, and B and C of full low rank is proposed. The CARE is projected onto a block rational Krylov subspace ⱼ spanned by blocks of the form (AH - sₖI)⁻¹CH for some shifts sₖ,\ k = 1, …, j. The considered projections do not need to be orthogonal and are built from the matrices appearing in the block rational Arnoldi decomposition associated to Kⱼ. The resulting projected Riccati equation is solved for the small square Hermitian Yⱼ. Then the Hermitian low-rank approximation Xⱼ = ZⱼYⱼZⱼH to X is set up where the columns of Zⱼ span Kⱼ. The residual norm \|R(Xⱼ )\|F can be computed efficiently via the norm of a readily available 2p × 2p matrix. We suggest reducing the rank of the approximate solution Xⱼ even further by truncating small eigenvalues from Xⱼ. This truncated approximate solution can be interpreted as the solution of the Riccati residual projected to a subspace of Kⱼ. This gives us a way to efficiently evaluate the norm of the resulting residual. Numerical examples are presented.
No takes yet. Share an insight, caveat, or question.
Bertram et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: