The Chaos-Based Approaches for Actual Metaheuristic Algorithms

Abstract

Along with rapid developments in computational technologies, evolutionary/heuristic/metaheuristic algorithms have frequently become used in many applications to solve optimization problems. Nowadays, new algorithms are being developed and improvements have been made to existing algorithms. In this study, chaos-based modifications have been proposed for recently introduced metaheuristic algorithms: Backtracking Search (BS), Grey Wolf Optimizer (GWO) and Vortex Search (VS), and the algorithms have been analyzed by detailed comparisons. The proposed approaches are based on generating new values through chaos maps, rather than some random numbers normally used in the algorithms, to improve their solutions. In addition, some modifications are performed to the structural operations of the algorithms used in the optimization process by taking advantage of chaos-based values. The performances of the algorithms are evaluated by considering two metrics: convergence rates and statistical results. Experiments demonstrated that the performance of the algorithms with the proposed modifications based on the chaos approach, are better than, or at least comparable to, the original algorithms.

Keywords

• metasezgisel algoritmalar
• kaos haritaları
• kaotik diziler
• metaheuristic algorithms
• chaotic maps
• chaotic sequences