A novel arithmetic optimization algorithm based on chaotic maps for global optimization

Abstract

Chaotic maps are effective in developing evolutionary algorithms (EAs) to avoid local optima and speed convergence. Because of this capability of chaotic maps, these maps have been hybridized with various optimization algorithms. In this study, a new optimization method based on the combination of chaotic maps and arithmetic optimization algorithm (CAOA) is proposed. AOA represents the behavior of four basic arithmetic operators. Therefore, AOA performs the optimization process in a wide search space with its mathematical model. Therefore AOA has an effective convergence capability. In addition, ten different chaotic maps are applied on the AOA. In this study, 7 scenarios were created with chaotic maps, taking into account different phases of AOA. The proposed CAOA is tested on eighteen benchmark problems. CAOA produces successful and promising results in solving optimization problems compared to original AOA algorithm. The proposed CAOA is also compared with the original AOA. The superior aspects of CAOA over AOA are discussed. CAOA convergence performance is also discussed. The statistical significance of the proposed hybrid algorithm is tested with the Wilcoxon sign-rank method.