Suluova Hamid Furkan, Pham Duc Truong
Department of Mechanical Engineering, The University of Birmingham, Birmingham B15 2TT, UK.
Biomimetics (Basel). 2024 Oct 18;9(10):634. doi: 10.3390/biomimetics9100634.
Based on bee foraging behaviour, the Bees Algorithm (BA) is an optimisation metaheuristic algorithm which has found many applications in both the continuous and combinatorial domains. The original version of the Bees Algorithm has six user-selected parameters: the number of scout bees, the number of high-performing bees, the number of top-performing or "elite" bees, the number of forager bees following the elite bees, the number of forager bees recruited by the other high-performing bees, and the neighbourhood size. These parameters must be chosen with due care, as their values can impact the algorithm's performance, particularly when the problem is complex. However, determining the optimum values for those parameters can be time-consuming for users who are not familiar with the algorithm. This paper presents BA, a Bees Algorithm with just one parameter. BA eliminates the need to specify the numbers of high-performing and elite bees and other associated parameters. Instead, it uses incremental k-means clustering to divide the scout bees into groups. By reducing the required number of parameters, BA simplifies the tuning process and increases efficiency. BA has been evaluated on 23 benchmark functions in the continuous domain, followed by 12 problems from the TSPLIB in the combinatorial domain. The results show good performance against popular nature-inspired optimisation algorithms on the problems tested.
基于蜜蜂觅食行为，蜜蜂算法（BA）是一种优化元启发式算法，已在连续域和组合域中得到广泛应用。原始版本的蜜蜂算法有六个用户选择的参数：侦察蜂数量、高性能蜜蜂数量、顶级或“精英”蜜蜂数量、跟随精英蜜蜂的觅食蜂数量、被其他高性能蜜蜂招募的觅食蜂数量以及邻域大小。必须谨慎选择这些参数，因为它们的值会影响算法的性能，尤其是在问题复杂时。然而，对于不熟悉该算法的用户来说，确定这些参数的最佳值可能很耗时。本文提出了一种只有一个参数的蜜蜂算法BA。BA无需指定高性能和精英蜜蜂的数量以及其他相关参数。相反，它使用增量k均值聚类将侦察蜂分成组。通过减少所需的参数数量，BA简化了调优过程并提高了效率。BA已在连续域的23个基准函数上进行了评估，随后在组合域中对来自TSPLIB的12个问题进行了测试。结果表明，在测试的问题上，BA相对于流行的自然启发式优化算法具有良好的性能。
