A Detailed Comparison of Two New Heuristic Algorithms Based on Gazelles Behavior

Abstract

In this study, Mountain Gazelle Optimization (MGO) and Gazelle Optimization Algorithm (GOA) algorithms, which have been newly proposed in recent years, were examined. Although MGO and GOA are different heuristic algorithms, they are often considered the same algorithms by researchers. This study was conducted to resolve this confusion and demonstrate the discovery and exploitation success of both algorithms. While MGO developed the exploration and exploitation ability by being inspired by the behavior of gazelles living in different groups, GOA model was developed by being inspired by the behavior of gazelles in escaping from predators, reaching safe environments and grazing in safe environments. MGO and GOA were tested on 13 classical benchmark functions in seven different dimensions and their success was compared. According to the results, MGO is more successful than GOA in all dimensions. GOA, on the other hand, works faster than MGO. Additionally, MGO and GOA were tested on three different engineering design problems. While MGO was more successful in the tension/compression spring design problem and welded beam design problems, GOA achieved better results in the pressure vessel design problem. The results show that MGO improves the ability to explore and avoid local traps better than GOA. MGO and GOA are also compared with three different heuristic algorithms selected from the literature (GSO, COA, and ZOA). According to the results, MGO has shown that it can compete with new algorithms in the literature. GOA, on the other hand, lags behind comparison algorithms.