| 引用本文： |
-
郑洪清,谢聪,周永权.一种改进的樽海鞘群算法[J].广西科学,2022,29(2):287-292. [点击复制]
- ZHENG Hongqing,XIE Cong,ZHOU Yongquan.An Improved Salp Swarm Algorithm[J].Guangxi Sciences,2022,29(2):287-292. [点击复制]
|
|
| 摘要: |
| 针对基本樽海鞘群算法(Salp Swarm Algorithm，SSA)在求解复杂函数时存在求解精度差和易陷入局部最优等缺陷，提出一种改进的樽海鞘群算法(Improved Salp Swarm Algorithm，ISSA)。首先，在领导者位置引入随机维度以拓展种群多样性；其次，改变追随者方式，即在算法前期以较大概率执行差分进化操作，进一步增强种群多样性，在算法后期较大概率执行黄金正弦算法，较好地平衡了算法的全局搜索和局部勘探能力。通过23个基准函数测试表明，本研究改进算法在收敛速度、计算精度和稳定性方面优于基本樽海鞘群算法和黄金正弦算法(Golden Sine Algorithm，Gold-SA)，同时与其他改进樽海鞘群算法相比，该算法也具有一定优势。 |
| 关键词: 樽海鞘群算法 函数优化 黄金正弦算法 差分进化策略 基准函数 |
| DOI：10.13656/j.cnki.gxkx.20220526.008 |
| 投稿时间：2021-03-09 |
| 基金项目:国家自然科学基金资助项目(61463007)和广西自然科学基金项目(2021GXNSFBA220080)资助。 |
|
| An Improved Salp Swarm Algorithm |
|
ZHENG Hongqing1, XIE Cong2, ZHOU Yongquan3
|
| (1.Institute of Education, Guangxi Vocational Normal University, Nanning, Guangxi, 530007, China;2.Guangxi Agricultural Vocational and Technical University, Nanning, Guangxi, 530007, China;3.School of Artificial Intelligence, Guangxi University for Nationalities, Nanning, Guangxi, 530006, China) |
| Abstract: |
| The Salp Swarm Algorithm (SSA) has the defects of poor solution accuracy and easy to fall into local optimum when solving complex functions.To solve this problem,an Improved Salp Swarm Algorithm (ISSA) is proposed.Firstly,a random dimension is introduced into the leader position to expand the population diversity.Secondly,the mode of followers is changed,that is,the differential evolution operation is carried out with a high probability in the early stage of the algorithm to further enhance the population diversity,and the golden sine algorithm is carried out with a high probability in the later stage of the algorithm,which better balances the global search and local exploration ability of the algorithm.23 benchmark function tests showed that the improved algorithm is better than the basic salp swarm algorithm and the golden sine algorithm in terms of convergence speed,calculation accuracy and stability.Meanwhile,the improved algorithm also has certain advantages compared with other improved salp swarm algorithms. |
| Key words: salp swarm algorithm function optimization golden sine algorithm differential evolution strategy benchmark functions |
