Experimental Demonstration of Peak Wavelength Measurement of Multiplexing Fiber Bragg Gratings Using Convolutional Neural Network

We propose a peak detection method for measuring fiber Bragg gratings (FBGs) using convolutional neural network (CNN) to improve the performances of wavelength division multiplexing. In multiplexing, each FBG occupies certain range; therefore, number FBGs that can be installed is limited by band light source. To address this issue, methods overlapping multiple same within single occupied range have been studied. This contributes improving limit multipoint manifold. However, results in complex reflectance spectra making it difficult accurately measure wavelengths individual conventional methods. Therefore, we developed process CNN, which suitable identifying unique feature data. Each was characterized spectral shape assigning different full width at half maximum each. introduced noise-additive learning, well-known data augmentation increases tolerance variations experimental signal. As result, standard deviation significantly improved 2.8 pm and strain measurements with three were successfully demonstrated. The CNN model first solve problem arbitrary changes. Furthermore, found applicable combined multiplexing either identical or wavelengths.