计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (4): 48-51.DOI: 10.3778/j.issn.1002-8331.2009.04.014

• 研究、探讨 • 上一篇    下一篇

改进的群搜索优化算法

张雯雰1,滕少华1,李丽娟2   

  1. 1.广东工业大学 计算机学院,广州 510006
    2.广东工业大学 建设学院,广州 510006
  • 收稿日期:2008-01-10 修回日期:2008-03-31 出版日期:2009-02-01 发布日期:2009-02-01
  • 通讯作者: 张雯雰

Improved Group Search Optimizer algorithm

ZHANG Wen-fen1,TENG Shao-hua1,LI Li-juan2   

  1. 1.Faculty of Computer,Guangdong University of Technology,Guangzhou 510006,China
    2.Faculty of Construction,Guangdong University of Technology,Guangzhou 510006,China
  • Received:2008-01-10 Revised:2008-03-31 Online:2009-02-01 Published:2009-02-01
  • Contact: ZHANG Wen-fen

摘要: 群搜索优化(Group Search Optimizer,GSO)算法是一种新的群集智能优化算法,适宜于解决多模态高维问题。对GSO算法进行了一些改进,简化了计算过程,提高了优化性能。主要在两个方面进行改进,一是在迭代过程中,控制允许变异的维的数量,使之从多到少变化,以提高收敛速度。二是用随机数来确定生成个体新位置所用的一组随机值的正负数比例,避免正负数比例趋于固定,增加随机性。经过6个常用测试函数测试及与其他文献结果对比后可知,在低维情况下,此算法与GA、EP、ES、PSO、GSO算法相比有较好的整体收敛性能,高维时,此算法与GA、PSO、GSO比较,收敛性能有明显优势。

关键词: 搜索, 优化, 群集智能, 进化算法, 高维, 函数优化

Abstract: GSO(Group Search Optimizer) is a new swarm intelligence algorithm,which has a superior performance on high-dimensional multi-modal problems.An improved GSO is presented in this paper to simplify the calculation process and improve the optimal performance.The modification has two main aspects:Firstly,during the iterative process,the number of dimensions of allowed variations may change from more to less to improve the convergence speed.Secondly,random is used to determine the coefficients of individual in next gap,to avoid the ratio inclining to be static,and to increase the randomness.From the testing of benchmark functions and the comparison with other papers,it can be concluded that the algorithm has a similar search performance in comparison with GA,EP,ES,PSO,GSO for low-dimensional functions,however,it has a markedly superior performance to other EAs for high-dimensional functions.

Key words: search, optimizer, swarm intelligence, evolutionary algorithm, high-dimensional, function optimization