Investigation of the Most Effective Meta-Heuristic Optimization Technique for Constrained Engineering Problems

One of the most common areas of meta-heuristic search (MHS) algorithms is optimization problems. In addition, the performance of only a few of the hundreds of MHS algorithms in the literature is known for constrained engineering design problems. The reason for this is that in most of the studies in which MHS algorithms have been developed, only classical benchmark problems are used to test the performance of the algorithms. Besides, applying MHS techniques to engineering problems is a costly and difficult process. This clearly demonstrates the importance of investigating the performance of new and powerful MHS techniques in engineering problems. In this paper, we investigate the search performance of the most recent and powerful MHS techniques in the literature on constrained engineering problems. In experimental studies, 20 different MHS techniques and five constrained engineering problems most commonly used in the literature have been used. Wilcoxon Runk Sum Test was used to compare the performance of the algorithms. The results show that the performance of MHS algorithms in classical benchmark problems and their performance in constrained engineering problems do not exactly match.