An Enhanced Opposition-Based Golden-Sine Whale Optimization Algorithm

Abstract

The Whale Optimization Algorithm is an ingenious method conceived by researchers, drawing inspiration from the feeding behavior of humpback whales. Characterized by its simple structure, limited parameters, high efficiency, and robust optimization capacity, WOA has been extensively applied across multiple domains to address various challenges. Nonetheless, it has been found that the algorithm demonstrates low global exploration capability, inadequate search precision, and susceptibility to local optima entrapment. Many enhancements have been suggested in the literature, with Opposition-Based Learning emerging as a particularly effective technique for improving the quality of the initial population. In the present study, we put forth the Enhanced Opposition-Based strategy, which integrates supplementary constraints into the existing Opposition-Based Learning framework, generating a more refined initial population. Furthermore, we introduce the Golden Sine Algorithm to modify the optimization approach of WOA, fostering an equilibrium between global exploration and exploitation abilities. In our evaluation, the proposed algorithm is assessed on nine classic benchmark functions with a dimensionality of 500, and compared with the original WOA, An enhanced whale optimization algorithm (eWOA), and the Elite Opposition-Based Golden-Sine Whale Optimization Algorithm (EGolden-SWOA). The results exemplify the superior performance of our proposed algorithm, underscoring its potential application in the optimization of truss structure design problems, the results indicate that ESWOA outperforms other enhanced algorithms, such as eWOA and EGolden-SWOA, in terms of its performance in engineering optimization. This signifies that ESWOA can be effectively applied to engineering optimization problems.