Simulate annealing for group structure optimization

It is challenging to select an appropriate group structure for multigroup neutron transport problem. Many group structures were chosen long ago, and the reasoning behind the creator's choices may be unknown. In this work, we apply the simulated annealing optimization method to develop improved group structures for a few test problems. Simulated annealing spans a large solution space before narrowing in on an optimal solution. Our solution space, however, is too large and too inconsistent for true optimization. Instead, we find potentially optimal group structures; ones that yield more accurate solutions than our standard group structures. We allow these group structures to be problem-dependent, with the view of developing a fast method to identify new group structures on-demand for different applications. The group structures used here outperform our current group structures of comparable size for the problems tested. The intent is for these optimized group structures to be used in a machine learning algorithm to help users choose a group structure based on problem characteristics, such as geometry and isotopics. (authors)