Solving Optimization Problems using Hybrid Metaheuristics: Genetic Algorithm and Black Hole Algorithm

Abstract

During the last two decades, many optimization algorithms have been developed for solving optimization problems. These algorithms have inspired from an intelligent behaviour of a living species, or a natural phenomenon. Black Hole (BH) algorithm has been developed recently, it as a metaheuristic that is based on population imitates the black hole event in the universe, whereby circulating solution in the search space represents an individual star. Although the original BH has shown better performance on benchmark datasets, it does not possess exploration capabilities but performs a good local search. In this paper, a new hybrid metaheuristic based on the combination of BH algorithm and Genetic Algorithm (GA) is proposed. The type of the proposed hybrid algorithm is High level hybridization, when GA represent the initialization phase (Global Search), while BH represents the searching Phase (Local Search). The proposed GA-BH is examined based on several optimization problems. The results obtained showed that the proposed algorithm is better than the original BH and GA algorithm.