Merrill不动点算法及其在不可微规划上的应用 |
| |
引用本文: | 姜冶,郭建,唐焕文. Merrill不动点算法及其在不可微规划上的应用[J]. 计算数学, 1988, 10(4): 361-368 |
| |
作者姓名: | 姜冶 郭建 唐焕文 |
| |
作者单位: | 大连工学院应用数学系(姜冶,郭建),大连工学院应用数学系(唐焕文) |
| |
摘 要: | 1.引言 不可微最优化(NDO)近年来很受重视,有重要的应用前景,它常使许多经典的可微优化方法失效.因此,需要研究新的有效算法. Merrill算法是以欧氏空间集值自映射为背景,寻求Kakutani不动点的一种单纯不动点算法.用其求解最优化问题时,由于它在一定条件下常具有大范围收敛的性质以及
|
MERRILL''S RESTART ALGORITHM AND ITS APPLICATION TO NONSMOOTH OPTIMIZATION |
| |
Affiliation: | Jiang Ye;Guo Jian;Tang Huan-wen Department of Applied Mathematics, Dalian Institute of Technology |
| |
Abstract: | Merrill's restart algorithm is a methods for finding numerically fixed points of uppersemicontinuous point-to-set mappings. With an analysis of the algorithm and its application tononsmooth optimization, we extend the convergence condition of this method and implement iton an IBM-PC microcomputer. Numerical results on several standard test problems indicatethat the method is a very promising algorithm for nonsmooth optimization. |
| |
Keywords: | |
本文献已被 CNKI 等数据库收录! |
| 点击此处可从《计算数学》浏览原始摘要信息 |
|
点击此处可从《计算数学》下载全文 |
|