Grey Wolf Optimizer

The grey wolf optimizer is a population-based metaheuristic algorithm for numerical optimization that models the social leadership hierarchy and cooperative hunting behavior of grey wolves in the wild. It ranks the candidate solutions in its population into four roles named after the wolf pack hierarchy, alpha, beta, delta and omega, treating the three best candidates as alpha, beta and delta and having every other candidate, the omega wolves, update its position toward a weighted combination of those three; hunting proceeds through three loosely modeled phases, tracking and approaching prey, encircling it and then attacking, with mathematical coefficients that shrink over the course of a run so the search shifts gradually from broad exploration toward focused exploitation around the best solutions found. Seyedali Mirjalili and colleagues introduced the algorithm in 2014, and it has since been applied to engineering design, economic dispatch, logistics and network routing problems, among the large family of nature-inspired metaheuristics that followed particle swarm optimization and ant colony optimization.