Analysis of different methods for combining and improving solutions with evolutionary metaheuristics in flow-shop production problems

Abstract

In recent years, there has been growth in the development of algorithms for solving optimization problems in flow-shop production environments involving different methods of combining and improving solutions. On the other hand, it is well known that, for a good performance of an evolutionary algorithm, it is essential to adjust its parameters and methods to a given problem. In this line, and considering makespan and total tardiness as evaluation criteria, this paper analyzes several methods of combination and improvement of the solutions, with a genetic algorithm and a scatter search algorithm, to evaluate their impact on the diversity of the solutions generated and on the convergence of both metaheuristics, when solving the permutation flow-shop scheduling problem with an instance of 50 jobs and 10 machines. The analysis of the results seeks to improve the understanding of both the behavior of these metaheuristics and the combination and improvement methods considered, so that practical guidelines for real applications can be obtained.