Implementation of Residual Tandem Neural Networks for Photonic Inverse Design

<p>Deep-learning approaches can greatly benefit the modeling and design of nanophotonic and optical structures. Traditional full-wave simulations are time and resource-intensive, which can act as a bottleneck in photonic design. On the other hand, deep-learning approaches for designing the response of nanophotonic geometries can be computationally inexpensive and produce accurate and efficient results. In this project, we specifically investigate the case of optical forces near meta-structures. We propose using an inverse design approach with residual blocks to account for the deep nature of this architecture and inherently address the non-uniqueness problem. A tandem approach, which consists of two interconnected models, is used, with the predictive model (the model that takes in a metastructure geometry and outputs a spectrum) acting as the ground truth. Region of Interest (ROI) modeling is also used to target particular regions in the optical spectrums that are of interest and produce highly targeted meta-surface geometries. This baseline architecture can be modified to fit many nanophotonic and optical structures. We report successful results in modeling optical forces near nanophotonic and optical structures using an inverse tandem design approach and region of interest modeling, which can be valuable for the future of the nanophotonic and optical field.</p>

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Electrical Engineering and Computer Science

Year dc:date.available

2025

Author dc:creator

• Zahraii, Ponthea A

Contributors dc:contributor

• Nasim Mohammadi Estakhri, Ph.D.
• Nooshin Mohammadi Estakhri, Ph.D.
• Mohamed Allali, Ph.D.
• Yuxin Wen, Ph.D.

• Inverse Design
• Photonic Metastructures
• Residual Neural Networks
• Deep Learning
• Tandem Neural Networks
• Optical Simulation
• Artificial Intelligence and Robotics
• Electromagnetics and Photonics

Repository record dc:identifier

https://digitalcommons.chapman.edu/eecs_theses/6

OAI identifier oai:identifier

oai:digitalcommons.chapman.edu:eecs_theses-1006