引用本文:申晓宁,黄遥,游璇,等.基于解空间反向跳跃和信息交互强化的新型混合蛙跳算法[J].控制与决策,2021,36(1):105-114
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览次   下载 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于解空间反向跳跃和信息交互强化的新型混合蛙跳算法
申晓宁1,2,3, 黄遥1, 游璇1, 王谦1
(1. 南京信息工程大学自动化学院,南京210044;2. 南京信息工程大学江苏省大气环境与装备技术协同创新中心,南京210044;3. 南京信息工程大学江苏省大数据分析技术重点实验室,南京210044)
摘要:
种群多样性和信息交互的深度与方式对混合蛙跳算法的爬山能力、探索能力和开发能力有着深远影响.针对混合蛙跳算法易于陷入局部最优、收敛速度慢和寻优精度差等缺点,提出一种基于解空间反向跳跃和信息交互强化的新型混合蛙跳算法.首先,增加子群次优解与次劣解的信息交互,促进子群内部信息的利用,引入反向跳跃思想改进局部更新机制,降低迭代后期劣解产生概率,提升空间开发能力;然后,借鉴2-opt方法实现局部最优解变异,增加子群的多样性;最后,采用各局部最优解交叉的方式加深子群间的交互深度,同时利用反向跳跃机制防止种群同化.采用23个单峰、多峰和固定维度下的复杂多峰函数作为测试集进行仿真实验,结果表明所提出算法具有更优的搜索性能,能够有效提高种群多样性,防止算法早熟收敛,且能够适应不同类型的函数优化问题.
关键词:  混合蛙跳算法  种群多样性  信息交互  反向跳跃  局部更新  函数优化
DOI:10.13195/j.kzyjc.2019.0719
分类号:TP301.6
基金项目:国家自然科学基金项目(61502239,51705260);江苏省自然科学基金项目(BK20150924);江苏省青蓝工程项目.
A new shuffled frog leaping algorithm based on reverse leaping in solution space and information interaction enhancement
SHEN Xiao-ning1,2,3,HUANG Yao1,YOU Xuan1,WANG Qian1
(1. School of Automation,Nanjing University of Information Science and Technology,Nanjing210044,China;2. Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology,Nanjing University of Information Science and Technology,Nanjing210044,China;3. Jiangsu Key Laboratory of Big Data Analysis Technology,Nanjing University of Information Science and Technology,Nanjing210044,China)
Abstract:
The diversity of population and the depth and way of information interaction have a profound impact on the ability of the shuffled leaping frog algorithm in climbing, exploring and development. Aiming at the shortcomings of the shuffled leaping frog algorithm, such as the inclination to local optimum, the slowness of convergence speed and the poor optimization accuracy, a new shuffled frog leaping algorithm based on reverse leaping in solution space and information interaction enhancement is proposed. Firstly, the information exchange between sub-inferior solutions and sub-optimal solutions is added to promote the utilization of information within sub-groups. In order to reduce the probability of inferior solutions in the later iteration period and to enhance the spatial development ability, the idea of reverse leaping is introduced to improve the local update mechanism. Then, the 2-opt method is used to realize the local optimal variation of sub-groups and increase the diversity of sub-groups. Finally, the local optimal crossover method is adopted to deepen the depth of interaction among subpopulations, and the reverse leaping mechanism is used to prevent population assimilation. The simulation experiments are carried out with a test sets includes 23 unimodal, multimodal and fixed-dimension complex functions. The results of simulation show that the proposed algorithm has better search performance, improves the diversity of population effectively, prevents premature convergence of the algorithm, and can adapt to different types of function optimization problems.
Key words:  shuffled frog leaping algorithm  population diversity  information interaction  reverse leaping  local update  function optimization

用微信扫一扫

用微信扫一扫