计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (24): 40-42.DOI: 10.3778/j.issn.1002-8331.2008.24.010

• 理论研究 • 上一篇    下一篇

一种新的差分进化算法

邓泽喜1,2,曹敦虔2,刘晓冀2,李 娜3   

  1. 1.毕节学院 数学系,贵州 毕节 557100
    2.广西民族大学 数学与计算机科学学院,南宁 530006
    3.德州学院 数学系,山东 德州 253023
  • 收稿日期:2007-10-29 修回日期:2008-01-02 出版日期:2008-08-21 发布日期:2008-08-21
  • 通讯作者: 邓泽喜

New differential evolution algorithm

DENG Ze-xi1,2,CAO Dun-qian2,LIU Xiao-ji2,LI Na3   

  1. 1.Department of Mathematics,Bijie College,Bijie,Guizhou 557100,China
    2.College of Mathematics and Computer Science,Guangxi University for Nationalities,Nanning 530006,China
    3.Department of Mathematics,Dezhou College,Dezhou,Shandong 253023,China
  • Received:2007-10-29 Revised:2008-01-02 Online:2008-08-21 Published:2008-08-21
  • Contact: DENG Ze-xi

摘要: 针对高维复杂函数的优化问题,提出了一种新的差分进化算法(NDE)。该算法在运行中根据迭代次数自动地调整交叉概率因子,从而在搜索的初始阶段提高种群多样性,而在搜索后期加强局部搜索能力。对几种经典函数的测试表明,新算法不仅具有很强的全局搜索能力,而且能有效避免早熟收敛问题。

关键词: 差分进化, 早熟收敛, 交叉概率

Abstract: For complex functions with high dimensions,a New Differential Evolution algorithm(NDE) based on the variance of the population’s fitness is presented.In order to balance global and local search ability,fasten convergence speed,avoid premature,the cross rate is automatically updated according to the generation.The experimental results show that the new algorithm not only has great advantage of convergence, but also can avoid the premature convergence problem effectively.

Key words: differential evolution, premature convergence, cross rate