Reconstructing arbitrary strain distributions within fiber gratings by evolutionary programming

An innovative algorithm based on the evolutionary programming (EP) method is developed for recovering distributions of axial strain along a fiber grating from its reflection spectral response. The proposed method exhibits a number of attractive features that prove to be effective for solving the inverse problems. The basics of EP are reviewed and the detailed programming procedures of the proposed algorithm are presented. By combining EP and the matrix method for calculating the reflection spectrum of a FBG, we obtain a new method for the distributed sensing. The numerical simulations show good agreements between the original and the reconstructed strain profiles.