Simultaneous Temperature and Strain Measurement in Fiber Bragg Grating via Wavelength-Swept Laser and Machine Learning
Byeong Kwon Choi, Ji Su Kim, Soyeon Ahn, Sung Yoon Cho, Min Su Kim, Jaehyun Yoo, Min Yong Jeon
Abstract
Measuring temperature and strain using fiber Bragg grating (FBG) simultaneously has been a difficult task in recent years. Creating a dataset to differentiate temperature and strain via machine learning (ML) techniques necessitates collecting voluminous data rapidly. This study proposes a method to simultaneously measure the temperature and strain of a single FBG using ML while employing a wavelength-stabilized wavelength-swept laser (WSL). We analyzed peak intensities and sidelobe positions in the temporal domain under different temperatures (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0~^{\circ }$ </tex-math></inline-formula>C–<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$55~^{\circ }$ </tex-math></inline-formula>C) and strains (0–<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$300~\mu \varepsilon $ </tex-math></inline-formula>) of FBG, amassing approximately 46000 datasets comprising six variables. ML was performed using Python’s Scikit-learn, with the dataset split into training (75%) and testing (25%) subsets. Temperature and strain could be measured simultaneously with approximately 95% accuracy, suggesting the potential of ML in accurately predicting temperature and strain using FBG responses.