Synthesis of fiber Bragg grating parameters using hybrid optimization algorithm

Abstract

A simple, fast and accurate method is proposed to reconstruct the parameters of fiber Bragg grating (FBG) from its reflectivity. The proposed method is based on hybrid of Tabu Search (TS) and Nelder-Mead (NM) Simplex algorithm. This method allows wide coverage of the parameters' solution space and minimizes the risk of trapping in local minima. Tabu Search explores the solution space and Nelder-Mead Simplex accelerates the convergence to a minimum. Comparisons are made with several methods such as Simulated Annealing (SA), Tabu Search and Nelder-Mead Simplex Algorithm applied to FBG parameters reconstruction.