City Research Online

Machine learning modelling, optimisation and thermal compensation of photonic waveguides

Chugh, S. (2020). Machine learning modelling, optimisation and thermal compensation of photonic waveguides. (Unpublished Doctoral thesis, City, University of London)

Abstract

Nanophotonic devices has led to many interesting applications in optical sensing, fibre lasers, fibre amplifiers, optical signal processing, and many others. Modelling and optimisation of such devices depends upon the numerical methods employed for modal analysis, such as finite difference method, finite element method, beam propagation method, and others to compute various optical properties including effective index, power confinement, mode effective area, dispersion, confinement loss, etc. One of the aims of this dissertation is to develop a finite element based time domain technique, similar to finite difference time domain method, that can have varying mesh resolutions for spatial discretisation of the computational domain. Parallel programming has been employed to speed up the simulations.

However, various design parameters of the optical devices are generally optimised before fabrication. This becomes an iterative process of trying and testing different design parameters which may require significant time and computer resources when dealing with complex optical structures. In this research work, the power of artificial intelligence techniques has been employed to quickly estimate the various properties of different photonics devices (slot, strip, and directional coupler waveguides) and photonic crystal fibre. An in-house code using a machine learning (an application of artificial intelligence) regression approach has been developed. Accuracy of these techniques are described by comparing their outputs with the actual outputs. PyTorch and Python programming languages are extensively used during the development of machine learning approach.

Publication Type: Thesis (Doctoral)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments: Doctoral Theses
Doctoral Theses > School of Mathematics, Computer Science and Engineering Doctoral Theses
School of Mathematics, Computer Science & Engineering > Computer Science > Software Reliability
Date Deposited: 24 Aug 2020 14:28
URI: https://openaccess.city.ac.uk/id/eprint/24806
[img] Text - Accepted Version
This document is not freely accessible until 31 May 2023 due to copyright restrictions.

To request a copy, please use the button below.

Request a copy

Export

Downloads

Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login