Hybrid step size based cuckoo search

Abstract

Cuckoo Search is a metaheuristic optimization method which uses obligate brood parasitic conduct of cuckoos. Successive moves (steps) of a cuckoo are basically drawn from a process which uses biased random walk and follows power law step-length distribution with a heavy tail. Therefore, quick convergence and precision cannot be assured in most of the cases. To resolve the convergence and accuracy problem this paper proposes a novel cuckoo search method which uses Lévy flight and Gauss distribution in alternate iteration along with variable Pa and a. The experimental outcomes validate that the proposed method outperforms the existing methods.