Improved Cuckoo Search Algorithm Based on Firefly Mechanism

Abstract

Cuckoo Search (CS) algorithm has the characteristics of few parameters, easy realization and strong exploration ability. However, CS algorithm relies heavily on the random step in the process of exploration and lacks communication with excellent individuals in the process of optimization, which leads to slow convergence speed of CS algorithm. Firefly algorithm (FA) achieves the goal of optimizing by constantly approaching individuals who are better than themselves. Inspired by the firefly algorithm, Our search algorithm introduces the firefly mechanism into the CS algorithm to increase the information exchange between particles to improve the convergence speed of the algorithm. CS-FA has no random term of FA, because cuckoo search algorithm has powerful global exploring ability. The simulation results show that our algorithm CS-FA has better optimization ability and faster convergence speed than CS algorithm, FA and also PSO algorithm.