An innovative hybrid method combining grey wolf and marine predator optimization techniques for global and constrained problem-solving

Saptadeep Biswas, Binanda Maiti, Gyan Singh, Seyedali Mirjalili, Vladimir Simic, Dragan Pamucar, Uttam Kumar Bera

Abstract: The grey wolf optimizer (GWO) is a metaheuristic algorithm recognized for its effectiveness; however, it faces several limitations, such as a lack of diversity in its population, a tendency to prematurely converge on local optima, and insufficient convergence speed. To address these issues, we propose an innovative hybrid algorithm that combines the advantages of GWO with the marine predator algorithm (MPA), leading to the creation of the Hybrid Grey Wolf Marine Predator Algorithm (HGWMPA). By integrating the adaptive characteristics of MPA, this hybrid approach fosters a robust search mechanism that effectively balances exploration and exploitation. We performed comprehensive experimental assessments utilizing benchmark functions from the CEC competitions of 2014, 2017, 2020, and 2022. The findings indicate that the HGWMPA consistently surpasses numerous leading optimization methods, achieving an average rank of 1 across most benchmark functions. Specifically, HGWMPA secured top positions in 76.67% of functions in the CEC 2014 test suite, 70.00% in CEC 2017, 90.00% in CEC 2020, and 66.67% in CEC 2022, showcasing its robust performance across various benchmark scenarios. The experimental results reveal that HGWMPA excels in global exploration, local exploitation, convergence speed, and accuracy, achieving optimal or near-optimal solutions with minimal standard deviations. A detailed performance evaluation, employing the Wilcoxon rank-sum test and the MARCOS MCDM ranking technique, further confirms the competitive advantages of HGWMPA. The algorithm’s adaptability, characterized by the dynamic adjustment of parameters, enables an effective balance between exploration and exploitation, making it particularly suitable for a wide range of engineering design problems. Sensitivity analyses indicate that changes in population size, maximum iteration, and other parameter limits significantly influence the algorithm’s performance, providing valuable insights for enhancing their configurations. HGWMPA has been successfully applied to diverse engineering design challenges, demonstrating its versatility and effectiveness in minimizing costs while adhering to critical design constraints. This advancement in optimization techniques, represented by HGWMPA, integrates pioneering concepts from both GWO and MPA to effectively tackle real-world challenges.