.png){kind=icon}
.png){kind=icon}
Return
Efficient exploration of high-dimensional topological spaces for phononic crystals bandgap customization
.png){kind=icon}
PRE
AI
C
A
X
C
Z
DOI：10.1016/j.aei.2026.104740.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• A novel inverse design framework for phononic crystals (PnCs) is proposed integrating VAE, Transformer-LSTM, and genetic algorithms. • A large-scale database of 13,000 2D PnCs structures was constructed using high-order polar coordinate parameterization and automated FEM. • Transformer-LSTM exhibits outstanding predictive accuracy for bandgap boundaries with an error under 1% and R2 of 0.992. • Latent space genetic optimization enables efficient search for PnCs structures with customized bandgaps in the 420–520 Hz range. • 3D-printed experimental prototypes confirm up to 95% vibration attenuation in the bandgap.
Keywords:
Phononic crystals
Inverse design
Variational Autoencoder
Transformer-LSTM
Genetic algorithm
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
9.9
Papers:
4.0K
Citations:
1.7W
Organization
Cited Papers