Simulated annealing: in mathematical global optimization computation, hybrid with local or global search, and practical applications in crystallography and molecular modelling

TL;DR

This paper explores simulated annealing (SA) in global optimization, detailing its implementation, hybridization with local/global search methods, and applications in crystallography and molecular modeling, especially prion fibrils.

Contribution

It provides a detailed study of SA's practical implementation and introduces hybrid optimization methods, enhancing global search efficiency in scientific applications.

Findings

Hybrid SA methods improve global search effectiveness.

Application of SA in crystallography and molecular modeling.

Insights into SA's role in studying prion amyloid fibrils.

Abstract

Simulated annealing (SA) was inspired from annealing in metallurgy, a technique involving heating and controlled cooling of a material to increase the size of its crystals and reduce their defects, both are attributes of the material that depend on its thermodynamic free energy. In this Paper, firstly we will study SA in details on its practical implementation. Then, hybrid pure SA with local (or global) search optimization methods allows us to be able to design several effective and efficient global search optimization methods. In order to keep the original sense of SA, we clarify our understandings of SA in crystallography and molecular modeling field through the studies of prion amyloid fibrils.