Machine Learning Approach to Analyze the Different Optical Properties of FBGs
Keyword(s):
Abstract A generalized machine learning (ML) approach is proposed and demonstrated to analyse the various optical properties such as effective refractive index, bandwidth, reflectivity and wavelength of the Fiber Bragg gratings (FBGs). For this purpose, three commonly used FBG variants namely conventional, π phase-shifted and chirped FBG have been taken into consideration. Furthermore, the reflected spectra of those types of FBGs were predicted using a common tool. An exact spectrum was able to reproduce using this proposed model. This simple and fast-training feed-forward artificial neural network can predict the output for unknown device parameters along with the non-linear and complex behaviour of the spectrum minutely.
2018 ◽
Vol 55
(5)
◽
pp. 1248-1260
◽
2018 ◽
Vol 24
(07)
◽
pp. 1
◽
2020 ◽
Vol 34
(02)
◽
pp. 1693-1700
◽
2020 ◽
Vol 13
(28)
◽
pp. 2849-2857
Keyword(s):