您的位置:首页  > 论文页面

基于共轭方向的最优化算法

发表时间:2010-01-15  浏览量:1835  下载量:785
全部作者: 潘鑫
作者单位: 天津工程师范学院理学院
摘 要: 给出了一种基于共轭方向的最优化方法。每个搜索方向在由2个共轭向量张成的子空间中生成,并保证了搜索方向的下降性。该方法可以视为特殊的迭代子空间方法,总体收敛性得到保证。可以证明这种方法无需精确线搜索就具有二次终止性。新算法运行过程中只需存储4个向量,故适用于大型优化问题。数值实验表明新算法有效。
关 键 词: 运筹学;二次终止性;共轭方向;迭代子空间;大型优化问题
Title: A new type of optimization method based on conjugate directions
Author: PAN Xin
Organization: Science School, Tianjin University of Technology and Education
Abstract: In this paper, a new type of optimization method based on conjugate directions is proposed.Each search direction is generated in the subspace spanned by two conjugate directions, and can be guaranteed descent. It can be proved that this type of method has quadratic termination property without exact line search. The new method is designed to be required only the storage of 4 vectors such that it is suitable for large scale optimization problems. Numerical experiences show that the new method is effective.
Key words: operational research; quadratic termination property; conjugate directions; iterated-subspace; large scale optimization problems
发表期数: 2010年1月第1期
引用格式: 潘鑫. 基于共轭方向的最优化算法[J]. 中国科技论文在线精品论文,2010,3(1):96-101.
 
0 评论数 0
暂无评论
友情链接