A chaotic grey wolf optimizer for constrained optimization problems

Bio-inspired algorithms have become popular due to their capability of finding good solutions for complex optimization problems in an acceptable computational time. The Grey Wolf Optimizer is a nature-inspired, population-based metaheuristic that simulates the social hierarchy and the hunting strategy observed in a grey wolf pack. Although the Grey Wolf Optimizer has been successfully applied to solve different optimization problems, it may suffer from premature convergence and get stuck in local optima. In order to overcome these drawbacks, this paper proposes a chaotic version of the Grey Wolf Optimizer that differs from the original algorithm and previously published modified versions because it does not add a chaotic variable in the parameters that control the execution of the algorithm. Instead, the proposed model uses a chaotic variable to define the wolves in the pack that will be used to guide the hunting process in each iteration of the algorithm. Numerical experiments using 20 benchmark functions are carried out. The performance of the proposed model is compared with the performance of the original Grey Wolf Optimizer and other well-known algorithms, namely the Particle Swarm Optimization, the Genetic Algorithm, the Symbiotic Organisms Search, and the Teaching-Learning Based Optimization. Nine chaotic maps reported in the literature are tested. The results show that the proposed algorithm has a very competitive performance, and the Chebyshev map presented the best performance among the chaotic maps simulated. The proposed algorithm can be integrated into other modified versions of the Grey Wolf Optimizer in a straightforward way.