At the international student conference Poster 2024, organized by our faculty, Bc. Aneta Furmanová, a student under Dr. Viktor Hruška, was awarded first prize in the Natural Sciences section. In her work, she combined the theory of acoustic wave propagation in periodic structures with machine learning methods. The basis was the numerical solution of the transmission of a quasi-plane wave through a pipe with a periodically varying cross-section using Floquet-Bloch theory. Subsequent symbolic regression describes the dispersion relation of wave transmission using analytical and interpretable formulas. These can otherwise only be obtained for a few special cases of waveguide geometry. The application of the obtained relationships can directly lead to the design of acoustic dampers. In addition, locally periodic structures can also occur as basic elements of more complex acoustic metamaterials. In the future, this work will serve not only to expand practical knowledge about wave propagation in this type of structure but also as a springboard for other wave systems with inhomogeneities, which are by no means limited to acoustics.
