On the Optimality of Reduced-Order Models for Band Structure Computations: A Kolmogorov -Width Perspective
This paper establishes that reduced-order models for phononic, acoustic, and photonic band structure computations achieve exponential convergence rates determined by spectral gaps, as proven via Kolmogorov -width analysis of holomorphic eigenpairs and spectral projectors, thereby providing a sharp optimality benchmark that validates the effectiveness of greedy algorithms and existing methods like RBME.