Evolutionary Algorithms Applied to Landau-Gauge Fixing

Abstract

Current algorithms used to put a lattice gauge configuration into Landau gauge either suffer from the problem of critical slowing-down or involve an additional computational expense to overcome it. Evolutionary Algorithms (EAs), which have been widely applied to other global optimisation problems, may be of use in gauge fixing. Also, being global, they should not suffer from critical slowing-down as do local gradient based algorithms. We apply EA's and also a Steepest Descent (SD) based method to the problem of Landau Gauge Fixing and compare their performance.