Enhancing firefly algorithm using generalized opposition-based learning

Firefly algorithm has been shown to yield good performance for solving various optimization problems. However, under some conditions, FA may converge prematurely and thus may be trapped in local optima due to loss of population diversity. To overcome this defect, inspired by the concept of opposition-based learning, a strategy to increase the performance of firefly algorithm is proposed. The idea is to replace the worst firefly with a new constructed firefly. This new constructed firefly is created by taken some elements from the opposition number of the worst firefly or the position of the brightest firefly. After this operation, the worst firefly is forced to escape from the normal path and can help it to escape from local optima. Experiments on 16 standard benchmark functions show that our method can improve accuracy of the basic firefly algorithm.