Improved Alpha-Guided Grey Wolf Optimizer

Abstract

Grey wolf optimizer (GWO) is a new meta-heuristic swarm intelligence algorithm, which has shown promising performance in solving optimization problems. In order to improve the convergence speed of GWO, an alpha-guided GWO (AgGWO), in which the evolving process of the population is guided by the update direction of alpha (best solution), is proposed. However, in the AgGWO, its evolutionary guidance mechanism makes the algorithm more likely to fall into the local optimal solution and the fixed value of theta may not be suitable for all problems optimization. To overcome these shortcomings and simplify its structure, an improved AgGWO (IAgGWO) is proposed in this paper. In the IAgGWO, the update direction of alpha is used to guide the evolving process of alpha, beta (second best solution), and delta (third best solution), and A and C are the coefficient scalars instead of coefficient vectors in the original algorithm. Therefore, a mutation operator is introduced to further enhance the exploration. The advantageous performance of the IAgGWO is validated by comparisons with other four algorithms on 35 benchmark functions and the engineering problem of two-stage operational amplifier design.