Inverse Design of Bend-Resistant Photonic Crystal Fibers Based on DNN With Feature Analysis

In this paper, a bend-resistant photonic crystal fiber inverse design using a deep neural network (DNN) with feature analysis is proposed for higher efficiency. First, a DNN model is trained to design structures of the bend-resistant PCF according to the given optical performance. Then, Shapley addi...

Celý popis

Podrobná bibliografie

Hlavní autoři:	Dan Yang, Zhenzhen Ma, Hong Liu, Jiaqi Liu, Bin Xu
Médium:	Artigo
Jazyk:	Inglês
Vydáno:	IEEE 2026-01-01
Edice:	IEEE Photonics Journal
Témata:	Photonic crystal fiber inverse design DNN SHAP XAI
On-line přístup:	https://ieeexplore.ieee.org/document/11271800/
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!