Weighted distance grey wolf optimization with immigration operation for global optimization problems

Abstract

The proposed algorithm presents a solution to improve the grey wolf optimizer performance using weighted distance and immigration operation. The weight distance is used for the omega wolves movement is defined from fitness value of each leader (alpha, beta and delta). The traditional grey wolf algorithm has only one pack and has opportunity to trap in local optimum so the wolves in our proposed algorithm have more pack and have migrated between them. When the amount of pack has more than to predefine some pack will be eliminated. The experimental results are evaluated by a comparative with the traditional grey wolf optimizer (GWO) algorithm, particle swarm optimization (PSO) and differential evolution (DE) algorithm on 9 well-known benchmark functions. The experimental results showed that the proposed algorithm is capable of efficiently to solving complex optimization problems.