Modified Opposition Based Learning to Improve Harmony Search Variants Exploration

Harmony Search Algorithm (HS) is a well-known optimization algorithm with strong and robust exploitation process. HS such as many optimization algorithms suffers from a weak exploration and susceptible to fall in local optima. Owing to its weaknesses, many variants of HS were introduced in the last decade to improve its performance. The Opposition-based learning and its variants have been successfully employed to improve many optimization algorithms, including HS. Opposition-based learning variants enhanced the explorations and help optimization algorithms to avoid local optima falling. Thus, inspired by a new opposition-based learning variant named modified opposition-based learning (MOBL), this research employed the MOBL to improve five well-known variants of HS. The new improved variants are evaluated using nine classical benchmark function and compared with the original variants to evaluate the effectiveness of the proposed technique. The results show that MOBL improved the HS variants in term of exploration and convergence rate