Dimensionality reduction for efficiency human and computer codesign in integrated photonics - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Dimensionality reduction for efficiency human and computer codesign in integrated photonics

Résumé

Computer design of photonic integrated chip results usually into more efficient solutions, but the key geometrical features that constitute the core of the design, and explain most of its performance and the logic behind them, are difficult to identify. Taking as example densely packed arrays of waveguides, we explain how the knowledge of the physics governing these systems allows the construction of a new representation basis that has much less degrees of freedom than the direct space representation. The analysis of the results of a subsequent Genetic Algorithm optimization reveals new innovative strategy that mediate the shortcomings of human designing.
Fichier principal
Vignette du fichier
Sheveleva_Dim_Red (1).pdf (3.4 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03607339 , version 1 (25-11-2022)

Identifiants

Citer

Anastasiia Sheveleva, Christophe Finot, Pierre Colman. Dimensionality reduction for efficiency human and computer codesign in integrated photonics. SPIE Photonics Europe 2022, SPIE, Apr 2022, Strasbourg, France. pp.1214808, ⟨10.1117/12.2620510⟩. ⟨hal-03607339⟩
16 Consultations
17 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More