An Improved Cuckoo Search Based on Starling Block Behavior

Cuckoo Search is a new metaheuristic algorithm. However, has some shortcomings like low convergence precision and easily relapsing into the local optima. In this paper, we introduce the collective behaviour of the starling group into CS algorithm, which can improve the precision and the searching range. In order to balance the convergence speed and MSE, a new variable step-size algorithm has been presented. Simulation results prove that compared with the basic CS algorithm, the new algorithm can avoid the local optimum and attain higher convergence precision.