Abstract:Firefly Algorithm (FA) is an optimization algorithmbased on swarm intelligence which mimics the social behavior of fireflies basedon the flashing and attraction characteristics of fireflies With the aim toaddress the disadvantages of the firefly algorithm of slow convergence speedand ease of falling into the local optimum in the later period of the evolutionprocess

the firefly algorithm is improved herein Two kinds of heuristicinformation are proposed into the algorithm to guide the convergence of thealgorithm The first one takes the current global best as the heuristicinformation referencing the “global optimal” idea in particle swarmoptimization

therefore

an algorithm called FAGO (Firefly Algorithm based onGlobal Optimization) is formed The second one is called FABE (Firefly Algorithmbased on Bayesian Estimation) using the optimal moving direction calculated byBayesian estimation as heuristic information The improved algorithms in thisstudy are applied to numerical simulations of several classical test functionsand compared with traditional FA and some other′s research are carried out Thesimulation results show that the proposed algorithms can well accelerate theconvergence speed and improve the convergence accuracy