Broadband Acoustic Metamaterial Design via Machine Learning

Acoustic metamaterials are engineered microstructures with special mechanical and acoustic properties enabling exotic effects such as wave steering, focusing cloaking. In this research, we develop a new machine learning framework for predicting optimal metastructures planar configurations of scatterers specific functionalities. Specifically, implement by combining probabilistic generative modeling deep propose two models: conditional variational autoencoder (CVAE) supervised (SVAE) model. As an application the method, here design cloak considering minimization total scattering cross-section (TSCS) set cylindrical obstacles. We work sets objects confined in region space streamline minimal TSCS, demonstrating broadband cloaking effect at discrete wavenumbers. After establishing artificial neural networks that capable TSCS based on location cylinders, discuss our inverse algorithms, autoencoders Gaussian process, arrangements given TSCS. show results up to eight cylinders efficiency other advantages approach.