Optimization of Hamiltonian algorithm for fiber alignment by simulation and experiments

Abstract

We aim to optimize the Hamiltonian algorithm for fiber alignment by simulation and experiments. We develop (1) a three-step optimization method to adjust two relevant parameters in the Hamiltonian algorithm in order to effectively make the alignment process faster and more stable, and (2) two methods to terminate Hamiltonian searching in order to preserve successful alignments under the influence of noise. We demonstrate the effectiveness of all these practical methods through experiments using an alignment system. © 2005 Society of Photo-Optical Instrumentation Engineers.