Levy mutated Artificial Bee Colony algorithm for global optimization
Rajasekhar Anguluri, Ajith Abraham, Millie Pant · 2011
This paper proposes an improved version of Artificial Bee Colony (ABC) algorithm with mutation based on Levy Probability Distributions. The Levy distribution has a peculiar property of generating an offspring farther away from its parent which depends on internal parameter α compared to that of Gaussian mutations, this property enables in finding out most optimal solutions to the problems than that of conventional methods. The proposed algorithm is tested on 7 standard benchmark functions and on a set of non-traditional problems suggested in the special session of CEC'2008. Analysis and comparison of results with other state of art optimization algorithms like GA and PSO, shows the superiority of improved mutation, especially on high dimensional problems. This paper finally investigates the performance of proposed algorithm on the frequency-modulated sound wave synthesis problem, a real world problem in the field on communication engineering.