Spectral Reconstruction and Bayesian Model Framework for Characterization of Long Period Fiber Gratings

Abstract

The use of Long-Period Fiber Gratings (LPFGs) as sensors has been thoroughly researched, given the multitude of parameters these structures can monitor by themselves (such as temperature, strain, curvature) and the potential for combination with other materials that allow for monitoring of parameters such as humidity, pH and chemical concentration, at a low price and with easy fabrication processes available. This interest has increased the need for the development of interrogation systems for these sensors, particularly in the C-band spectral region. Given the cost and physical limitations (such as size and weight) of traditional solutions like Optical Spectrum Analyzers (OSA), the development of low-cost approaches for LPFG spectral analysis became an important topic that needed further development. The development of a simple curve fitting routine for LPFG spectra is reported in this article, along with a framework for automatic detection of certain physical phenomena such as corrosion and the presence of chemical species, among others.