Optimal parameter settings for bat algorithm

Abstract

Bat algorithm is a swarm intelligent algorithm inspired by the echolocation behaviour of bats. Recently, there are many research works focus on this algorithm, however, most of them are with different parameter settings when dealing with un-constrained numerical problems. In this paper, one optimal parameter setting for bat algorithm is discussed with orthogonal experimental design, and tested on CEC 2013 competition problems. To test the performance, the optimal setting is used to compare with other four bat algorithms. Simulation results show the validity.