Presentation
18 April 2022 Source localization based on deep learning of phononic modes
Author Affiliations +
Abstract
This work introduces the source localization application using a phononic crystal (PC) array. The PC band structure and the eigen-modes are analyzed and utilized for detecting the angle of arrival. The eigen-modes, as the basis functions of the scattering wave, possess strong angle-dependent features, naturally suitable for developing source localization algorithms. An artificial neural network is trained with randomly weighted eigen-modes to achieve deep learning of the modal features and angle dependence. The trained neural network can then accurately identify the incident angle of an unknown scattering signal, with minimal side lobe levels and suppressed main lobe width.
Conference Presentation
© (2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Weidi Wang, Amir Ashkan Mokhtari, Ankit Srivastava, and Alireza V. Amirkhizi "Source localization based on deep learning of phononic modes", Proc. SPIE PC12048, Health Monitoring of Structural and Biological Systems XVI, PC120480L (18 April 2022); https://doi.org/10.1117/12.2612432
Advertisement
Advertisement
KEYWORDS
Source localization

Scattering

Artificial neural networks

Calcium

Neural networks

Sensors

Switches

RELATED CONTENT

Sensor calibration methods: performance study
Proceedings of SPIE (August 01 1990)
On focal plane array feature extraction using a 3 D...
Proceedings of SPIE (November 01 1991)
Neural networks for web-process inspection
Proceedings of SPIE (September 02 1993)
Object discrimination using neural networks
Proceedings of SPIE (September 16 1992)

Back to Top