Cuckoo Search Algorithm based Dynamic Parameter Adjustment Mechanism for Solving Global Optimization Problems

摘要

Cuckoo search algorithm CSA is a recent optimization algorithm of swarm intelligence, which has demonstrated powerful outcomes on many optimization issues. nevertheless, it has some limitations such as stuck in local optima and premature convergence especially When solving complicated problems of optimization. Also, the CSA parameters are static during generations time which lead to stuck in local optima and couldn't find the best solutions. In this paper we proposed an improved standard cuckoo search algorithm based dynamic parameter adjustment mechanism called (CSDPA). The CSDPA presents two equation to update the parameters values of steps size and discovery probability during search process. The experiments are tested on ten conventional benchmark functions. Outcomes demonstrate the new CSDPA approach is outperform of the CSA and another CSA variants.