吉林大学学报(理学版)

• 数学 • 上一篇    下一篇

基于梯度法的高效全局优化算法

柯贤斌1, 刘红卫1, 游海龙2   

  1. 1. 西安电子科技大学 数学与统计学院, 西安 710126; 2. 西安电子科技大学 微电子学院, 西安 710071
  • 收稿日期:2015-05-11 出版日期:2016-01-26 发布日期:2016-01-19
  • 通讯作者: 柯贤斌 E-mail:xianbin_ke@126.com

Efficient Global Optimization Algorithm Based on Gradient Method

KE Xianbin1, LIU Hongwei1, YOU Hailong2   

  1. 1. School of Mathematics and Statistics, Xidian University, Xi’an 710126, China;2. School of Microelectronics, Xidian University, Xi’an 710071, China
  • Received:2015-05-11 Online:2016-01-26 Published:2016-01-19
  • Contact: KE Xianbin E-mail:xianbin_ke@126.com

摘要:

用极大似然估计法和交替方向法估计Kriging模型参数, 提出一种基于有效集共轭梯度法的Kriging模型参数优化算法, 并在此基础上改进了高效全局优化算法. 结果表明, 利用改进的全局优化算法可解决高效全局优化算法的过早收敛问题.

关键词: Kriging模型, 极大似然估计, 高效全局优化算法, 共轭梯度法

Abstract:

Based on the maximum likelihood estimation method and the alternating direction method to estimate the parameters of Kriging model, a Kriging model parameter optimization algorithm based on active set conjugate gradient method was proposed on the basis of which the global optimization algorithm was improved. Finally, the problem of premature convergence has been solved by means of the improved global optimization algorithm.

Key words: Kriging model, maximum likelihood estimation, efficient global optimization algorithm, conjugate gradient method

中图分类号: 

  • O241.3