DEGWO: a decision-enhanced Grey Wolf optimizer

The Grey Wolf optimizer (GWO) is an efficient meta-heuristic algorithm based on swarm intelligence, inspired by the hierarchical structure and hunting behavior of natural wolf packs. Due to straightforward algorithm flow lightweight and ease of implementation, GWO has been extensively applied to address optimization problems in various area. However, the original GWO suffers from slow convergence and a tendency to get trapped in local optimal solutions. In this paper, we propose an improved variant of GWO called Decision-Enhanced Grey Wolf optimizer (DEGWO), which introduces a weight assignment to the decisions made by three head wolves (α, β, δ) and establishes a decision value evaluation mechanism. Additionally, in order to prevent excessive reliance on α, β and δ that may lead to reduced population diversity and premature convergence issues, we incorporate dimension learning-based hunting and adaptive frequency perturbation mechanisms into DEGWO. A rigorous multiple analysis of comparisons on 24 well-known standard test functions with six state-of-the-art heuristics and five novel GWO variants, demonstrates that DEGWO exhibits superior capabilities in global exploration. Furthermore, to validate its applicability in other application domains, the proposed DEGWO algorithm was employed to optimize four simulation design problems.