Abstract

In this article, a self-adaptive global mine blast algorithm (GMBA) is proposed for numerical optimization. This algorithm is designed in a novel way, and a new shrapnel equation is proposed for the exploitation phase of mine blast algorithm. A theoretical study is performed, which proves the convergence of any typical shrapnel piece; a new definition for parameters values is defined based on the performed theoretical studies. The promising nature of newly designed exploitation idea is verified with the help of multiple numerical experiments. A state-of-the-art set of benchmark problems are solved with the proposed GMBA, and the optimization results are compared with seven state-of-the-art optimization algorithms. The experimental results are statistically validated by using Wilcoxon signed-rank test, and time complexity of GMBA is also calculated. It has been justified that the proposed GMBA works as a global optimizer for constrained optimization problems. As an application to the newly developed GMBA, an important data clustering problem is solved on six data clusters and the clustering results are compared with the state-of-the-art optimization algorithms. The promising results claim the proposed GMBA as a strong optimizer for data clustering application.

中文翻译：

---

用于数值优化的自适应全局排雷算法

摘要

本文提出了一种自适应的全球排雷算法（GMBA），用于数值优化。以新颖的方式设计了该算法，并为矿井爆炸算法的开发阶段提出了新的弹片方程。进行了理论研究，证明了任何典型弹片的收敛性。在进行的理论研究的基础上定义了参数值的新定义。借助多个数值实验，验证了新设计的开采理念的广阔前景。提出的GMBA解决了一组最新的基准测试问题，并将优化结果与七种最新的优化算法进行了比较。实验结果通过Wilcoxon符号秩检验进行了统计验证，并且还计算了GMBA的时间复杂度。已经证明，提出的GMBA可作为约束优化问题的全局优化程序。作为对最新开发的GMBA的应用，解决了六个数据集群上的重要数据聚类问题，并将聚类结果与最新的优化算法进行了比较。令人鼓舞的结果声称，拟议中的GMBA作为数据聚类应用程序的强大优化器。