Guo Linyi, Gu Wei
School of Computer Science, Hubei University of Technology, Wuhan, 430074, China.
Sci Rep. 2024 Sep 6;14(1):20849. doi: 10.1038/s41598-024-71828-y.
The optical microscope algorithm (OMA) is a metaheuristic algorithm that draws inspiration from the magnifying functionality of optical microscopes. This study introduces an enhanced variant of OMA, termed PMSOMA, designed to mitigate the original version's limitations, notably its slow convergence rates and vulnerability to local optima. PMSOMA integrates a piecewise linear chaotic map to refine population initialization and augment diversity, alongside a sparse adaptive exploration mechanism to bolster search efficacy. The performance of PMSOMA was rigorously tested using a suite of 50 benchmark functions, the CEC2017 test suite, feature selection datasets, and three classical engineering challenges. The empirical findings confirm that PMSOMA surpasses both the original OMA and competing algorithms by delivering superior solutions, accelerating convergence, and demonstrating enhanced robustness in convergence.
光学显微镜算法（OMA）是一种元启发式算法，其灵感来源于光学显微镜的放大功能。本研究介绍了OMA的一种增强变体，称为PMSOMA，旨在减轻原始版本的局限性，特别是其收敛速度慢和易陷入局部最优的问题。PMSOMA集成了分段线性混沌映射以优化种群初始化并增加多样性，同时还采用了稀疏自适应探索机制来提高搜索效率。使用50个基准函数组成的测试集、CEC2017测试套件、特征选择数据集以及三个经典工程挑战对PMSOMA的性能进行了严格测试。实证结果证实，PMSOMA通过提供更优的解决方案、加快收敛速度以及在收敛过程中展现出更强的鲁棒性，超越了原始的OMA和其他竞争算法。
