Inverse Design of Ultra-Wideband Frequency Selective Surface Using a Graph based Conditional Variational Autoencoder (G-CVAE) integrated with a Physics Informed Neural Network (PINN)

This work proposes the inverse design of bandstop Frequency Selective Surface using a Graph based Conditional Variational Autoencoder (G-CVAE) integrated with a Physics-Informed Neural Network (PINN). This inverse design involves the prediction of FSS geometry that exhibits ultra-wide stopband chara...

Celý popis

Podrobná bibliografie

Vydáno v:	IEEE journal on multiscale and multiphysics computational techniques s. 1 - 10
Hlavní autoři:	V, Bharathi, Ramanujam, Krishnamurthy, Ramanujam, Parthasarathy
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	IEEE 2025
Témata:	5G and beyond 5G applications Accuracy Autoencoders Computer architecture Conditional Variational Autoencoder Feature extraction Frequency Selective Surface Geometry Inverse design Microprocessors Optimization Physics-Informed Neural Network Shielding Effectiveness Training Vectors
ISSN:	2379-8815, 2379-8815
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!