KLT-Based Algorithm for Sub-Picometer Accurate FBG Tracking With Coarse Wavelength Sampling

Abstract

An algorithm based on Karhunen–Loeve transform (KLT) for accurate tracking of fiber Bragg grating (FBG) sensors is presented. The routine is based on main eigenvalue analysis of the KLT applied to the Fourier transform of the FBG spectrum. The algorithm achieves sub-picometer accuracy, down to 4.9 fm, when applied to coarse (156 pm) wavelength sampling, providing accuracy superior to correlation-based algorithms. Experimental validation has been carried out in a climatic chamber. The proposed algorithm can be applied to any off-the-shelf white-light setup for FBG interrogation, and is compatible with 100–1-kHz operation.