A Modified Fast Marriage in Honey Bee Opjtimization Algorithm.pdf

A Modified Fast Marriage in Honey Bee Optimization Algorithm M.T Vakil-Baghmisheh Mina Salim ICT Research Center, Faculty of Electrical Computer Engineering, University of Tabriz, Tabriz, Iran mvakil@tabrizu.ac.ir mina.salim@ Abstract—In this paper we present a modified version of an existing honey bee optimization algorithm: the modified fast marriage in honey bee optimization (MFMBO). Then we compare performances of this new algorithm and three existing bee algorithms, i.e. the artificial bee colony (ABC), the queen bee (QB), and the fast marriage in honey bee optimization (FMBO) on four benchmark functions for various numbers of variables up to 100. The obtained results show that the modified algorithm is faster than the others in most cases. Especially for Griewank and Schwefel functions by increasing precision of answer and number of variables, the difference between the speeds of the FMBO and MFMBO becomes prominent. In general the MFMBO algorithm is quite competitive with the other mentioned algorithms. Keywords- bee algorithms, fast marriage in honey bee optimization, artificial bee colony algorithm, queen bee algorithm I. INTRODUCTION Using evolutionary algorithms has become conventional for function optimization purposes, examples are genetic algorithms [1], ant colony optimization (ACO) algorithms[2], particle swarm optimization (PSO) algorithms [3], simulated annealing (SA) algorithm [4], and bee systems [5] , just to call a few. There are different algorithms based on the behavior of honey bees. First in 1997 Sato and Hagiwara presented the bee system [5] which was a modified version of genetic algorithms, but most of bee algorithms have been presented after 2003. In 2003 the discrete queen bee algorithm is introduced by Jung [6] which could be considered a variant of genetic algorithms, with some main differences as explained in section II-B. In 2005 Yang introduced the virtual bee algorithms [10] and in 2