Optimal Choice of Parameters for Firefly Algorithm

Abstract

The firefly algorithm (FA) is a novel metaheuristic algorithm inspired by the social behavior of fireflies and the phenomenon of bioluminescent communication. The effects of the major parameters on FA were systematically investigated based on some benchmark functions. The step size factor, absorption coefficient and population size all have significant impact on the performance of FA. The absorption coefficient and step size factor have optimal values in practical application, and improper choice of these factors will lead to bad results. Increasing population size can improve the solution quality, although the computing time will be longer. In this paper, the characteristics of FA parameters are discribed and guidelines for determining parameter values are given.