Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

DS: Fachverband Dünne Schichten

DS 10: Thin Film Application

DS 10.12: Talk

Wednesday, March 20, 2024, 18:00–18:15, A 060

Developing neural networks for designing optical thin films — •Elena Stoyanova, Krassimir Panayotov, Tihomir Tenev, and Ilko Miloushev — Institute of Solid State Physics, Bulgarian Academy of Sciences, 72 Tzarigradsko chaussee, Sofia 1784, Bulgaria

Almost every optical device is made up of a number of optical surfaces that guide and modify light as it passes through them. Optical coatings are produced as thin-film multilayers of different materials, using specialized deposition processes. Coatings are applied to optical components to guide and modify light as it passes through them and to use them at desired wavelength regions. Artificial Intelligence as neuron networks is included in the process of construction the thin films. Neural networks can learn and simulate complicated, nonlinear connections between input and output data. A computer program on Python is written, which investigates different layered structures in the spectral region between 0.2 *m and 2 *m. Film parameters values are given.

Keywords: Thin Films design; Artificial Intelligence; Neural Networks