Prioritized Multi-Step Decision-Making Gray Wolf Optimization Algorithm for Engineering Applications

Abstract

This article introduces the Prey-Movement Strategy Gray Wolf Optimizer (PMS-GWO), an enhanced version of the Gray Wolf Optimizer (GWO) designed to improve optimization efficiency through a novel multi-step decision-making process. By integrating adaptive exploration–exploitation strategies, PMS-GWO dynamically manages leadership roles, balances local and global searches, and introduces a prey escape mechanism, significantly improving solution diversity. Comparative analysis across 23 benchmark functions demonstrates PMS-GWO's superior performance, achieving up to 28.6% faster convergence and a 55.5%–93.8% increase in solution accuracy compared to the standard GWO. Notably, PMS-GWO enhances computational efficiency by 21.7%–27.4% and shows a 168.8% improvement in solution accuracy for the complex Michalewicz function over the baseline GWO. Visual convergence speed analysis, evidenced by a rapid fitness value decline within 100 iterations, reveals PMS-GWO's quickest convergence time of 0.02 s among tested algorithms. Furthermore, a comparison of runtime for several algorithms, including PMS-GWO, MMCCS-GWO, CC-GWO, MGWO, and GWO, clearly indicates that PMS-GWO achieves the lowest runtime of 2.364 s, significantly faster than CC-GWO and MGWO, which both exceed 5 s. This visual representation highlights the computational efficiency of PMS-GWO compared to other algorithms. PMS-GWO also outperforms advanced GWO variants like MMSCC-GWO, MGWO, and CCS-GWO, particularly in complex optimization landscapes, highlighting its adaptability and effectiveness for real-world applications in energy systems and engineering design. The multi-step decision-making process implemented in PMS-GWO is critical to achieving these improved convergence and diversity metrics.