A comparative analysis of meta-heuristic approaches for facility layout design problem: a case study for an elevator manufacturer

Abstract

In this study, a facility layout problem having NP-hard problem characteristic is attempted to be solved by using two different meta-heuristic approaches-Genetic Algorithm (GA) and Simulated Annealing (SA)-and a hybrid approach-Genetic Algorithm/Simulated Annealing (HGASA). The case study is completed for a company which can be seen as a small or a medium size enterprise. First, parameter values of GA and SA are determined by testing for various combinations of them. Then, the algorithms are run for one hundred times. The results of the algorithms are compared based on their fitness values and calculation time requirements using the paired-t test, mean and standard values. The results show that SA performs better than the others in terms of the fitness values and the time requirements. In this study, we also test the performance of our GA, SA and HGASA methodologies using some of the well-known test problems from the literature. We obtain very close results to those in literature.