Solving Multilocal Optimization Problems with Parallel Stretched Simulated Annealing

This work explores the use of parallel computing to solve multilocal optimization problems with Stretched Simulated Annealing (SSA), a method that combines simulated annealing with a stretching function technique. Several approaches to the parallelization of SSA are explored, based on different strategies for the refinement of the initial feasible region in subregions and its allocation to the processors involved. The parallel approaches, collectively named as PSSA (Parallel SSA), make viable what would otherwise be unfeasible with traditional sequential computing: an efficient search of the subregions that allows to find many more optima in a reasonable amount of time. To prove the merits of PSSA, several experimental metrics and numerical results are presented for a set of benchmark problems.