Evolved opposition-based Mountain Gazelle Optimizer to solve optimization problems

A recently established swarm-based algorithm, namely, Mountain Gazelle Optimizer (MGO) which draws inspiration from social structure and hierarchy of wild mountain gazelles is competitive for solving optimization problems. However, the MGO has some drawbacks: when dealing with higher dimensions, early iterations could become stuck in suboptimal search area. It would be difficult for the MGO to abandon the local optimal solution if the early best solutions neglect the relevant search space. Therefore, to overcome these limitations, this paper offers an Evolved Opposition-based Learning (EOBL) mechanism which helps the algorithm to jump out of the local optima while accelerating the convergence speed. This novel mechanism is incorporating with MGO to propose Evolved Opposition-based Mountain Gazelle Optimizer (EOBMGO). The experiments are conducted with CEC2005 and CEC2019 benchmark functions, along with seven engineering challenges to examine the performance of the proposed EOBMGO. Furthermore, the statistical tests, like the t-test and Wilcoxon rank-sum test, are verified and demonstrate that the proposed EOBMGO outperforms the existing top-performing algorithms. The outcomes indicated that the proposed technique may be seen as an efficient and successful approach for complex optimization challenges.

Automated summary

This paper introduces an Evolved Opposition-based Learning mechanism for the Mountain Gazelle Optimizer (EOBMGO) to overcome its limitations in dealing with higher dimensions and local optima. Experimental results and statistical tests demonstrate that EOBMGO outperforms existing algorithms, making it an efficient approach for complex optimization challenges.