Adaptive Multi-Population Artificial Bee Colony Algorithm Based on Fitness Landscape Analysis
Xinyu Zhou,Xiaocui Zhang,Weifeng Gao,Hui Wang,Yong Ma
DOI: https://doi.org/10.1016/j.asoc.2024.111952
IF: 8.7
2024-01-01
Applied Soft Computing
Abstract:In recent years, the multi-population mechanism stands out as an effective method to improve the artificial bee colony (ABC) algorithm. However, many existing multi-population ABC variants still exhibit some deficiencies, such as a fixed number of subpopulations and a single type of population partition method. Hence, in this work, a new multi-population ABC variant, called ABC-AMP, is proposed by designing an adaptive multi-population mechanism. For ABC-AMP, to determine the number of subpopulations and the type of population partition method, the fitness landscape analysis technique is first used to identify the problem features. Then, based on the identified features, the number of subpopulations and the most suitable type of population partition methods are determined adaptively. Meanwhile, to preserve the search experience, an improved dual-elite search strategy is crafted in the scout bee phase. To evaluate the performance of ABC-AMP, extensive experiments are conducted on two renowned test suites (CEC2013 and CEC2017) and three practical optimization problems. In comparison with four multi-population ABC variants, five other state-of-the-art ABC variants and four non-ABC variants, the comparative results demonstrate that ABC-AMP outperforms its competitors, showcasing a competitive performance.