期刊名称：International Journal of Hybrid Information Technology
印刷版ISSN：1738-9968
出版年度：2016
卷号：9
期号：12
页码：127-140
出版社：SERSC
摘要：Considering that the existing artificial bee colony (ABC) algorithm can not give simultaneously attention to evolution speed and solution quality, an improved ABC algorithm is proposed based on nelder-mead simplex method (NMSM) in this paper, and defined ad NMSM-ABC. In the process of iteration, the algorithm periodically passes the best individual vertex from NMSM operator into ABC, or migrates the optimal food source information from ABC to NMSM can get away from local minimum with aid of ABC. Hence, the proposed algorithm can realize cooperative co-evolution of the two so that the desired properties are obtained. In addition, the sensitivity analysis of the key parameter of NMSM-ABC is also conducted and the best value is suggested. Finally, Numerical experiments and comparisons on 6 benchmark functions are conducted with other ABC algorithms, and the results indicate that the proposed algorithm effectively overcomes the local minimum, and dramatically enhances the global searching ability and convergence speed, and is a good cooperative co-evolution algorithm.
