Spectral Demodulation of Tilted Fiber Bragg Grating Based Sensor for Temperature, Salinity and Pressure Measurement Using Convolutional Neural Network

This paper proposes a temperature-salinity-pressure sensor based on titled fiber Bragg grating, using convolutional neural network for spectral demodulation instead of conventional wavelength detection. Prediction results exhibit an error of ~0.0547℃, ~0.6506‰, ~0.006MPa, respectively.