Modified bat algorithm with cauchy mutation and elite opposition-based learning

Metaheuristics can be used to solve optimization complex problems because they offer approximate and acceptable solutions. In recent years, nature has been a source of inspiration for many computer scientists when proposing new metaheuristics such as the algorithms inspired by swarm intelligence. They are based on the behavior of animals that live in groups such as birds, fishes and bats. In this context, Bat Algorithm (BA) is a recent metaheuristic inspired by echolocation of bats during their flights. However, a problem that this algorithm faces is the loss of the ability to generate diversity and, consequently, the chances of finding the global solution are reduced. This paper proposes a modification to the original BA using two methods known as Cauchy mutation operator and Elite Opposition-Based Learning. The new variant aims generate diversity of the algorithm and increases its convergence velocity. It was compared to the original BA and another variant found in the literature. For this comparison, the proposed variant used four benchmark functions, during 30 independent runs. After the experiments, the superiority of the new variant is highlighted when the results are compared to the original BA.