Design and Implementation of Tilted FBG for Concurrent Temperature and Humidity Measurement using Machine Learning

Abstract

Tilted Fiber Bragg Grating (TFBG) coated with humidity and temperature responsive material is proposed for simultaneous measurement of environmental parameters. Different types of materials are analyzed to optimize the coating material on the grating structure for the measurement of humidity and temperature simultaneously. The optimized coating material is used to fabricate the sensor and experimentally investigated under different humidity and temperature conditions. To further enhance the performance, machine learning algorithms such as Gaussian Progress Regression, Random Forest, K-Nearest Neighbor, AdaBoost, Gradient Boosting algorithm were trained with the spectrum data to estimate the environmental parameters simultaneously. Gradient boosting algorithm best suited for the estimation with better performance metrics of absolute error of 0.6509, RMS of 2.7412 and MAPE of 1.312 for monitoring temperature and humidity simultaneously.