Highly Dense FBG Temperature Sensor at Multi-core Fiber Calibrated with Artificial Neural Network

Fiber Bragg grating (FBG) sensors are widely used for distributed temperature and strain sensing. Here we demonstrate a novel approach to calibrate densely inscribed FBG sensor at central core of multi-core fiber. Long-term memory neural networks trained on sparsely inscribed FBG sensors at peripherical cores were used to predict the positions of their reflectance peaks from complex spectrum obtained from densely inscribed FBG sensor. Our studies show that deep learning algorithms is an effective tool to increase the spatial resolution of the FBG temperature and strain sensors.