A TRUST REGION ALGORITHM FOR NONSMOOTH OPTIMIZATION

被引：8
作者：
BANNERT, T
机构：
[1] Institut für Numerische und Angewandte Mathematik, Georg-August-Universität Göttingen, Göttingen, D-37083
来源：
MATHEMATICAL PROGRAMMING | 1994年 / 67卷 / 02期
关键词：
TRUST REGION; NONSMOOTH OPTIMIZATION; EXACT DIFFERENTIABLE PENALTY FUNCTION; MARATOS EFFECT;
D O I：
10.1007/BF01582223
中图分类号：
TP31 [计算机软件];
学科分类号：
081202 ; 0835 ;
摘要：
A trust region algorithm is proposed for minimizing the nonsmooth composite function F(x) = h(f(x)), where fis smooth and h is convex. The algorithm employs a smoothing function, which is closely related to Fletcher's exact differentiable penalty functions. Global and local convergence results are given, considering convergence to a strongly unique minimizer and to a minimizer satisfying second order sufficiency conditions.
引用
收藏
页码：247 / 264
页数：18
相关论文
共 14 条
[1]  
BANNERT T, 1993, THESIS GEORG AUGUST
[2]  
Clarke F., 1987, OPTIMIZATION NONSMOO
[3]  
FLETCHER R, 1982, MATH PROGRAM STUD, V17, P67
[4]  
Fletcher R., 1973, Mathematical Programming, V5, P129, DOI 10.1007/BF01580117
[5]  
Fletcher R., 1970, Integer and nonlinear programming, P157
[6]  
Fletcher R., 1982, NUMERICAL ANAL, P85
[7]  
Fletcher R., 1987, PRACTICAL METHODS OP
[8]  
FLETCHER R, 1972, NUMERICAL METHODS NO, P371
[9]  
MADSEN K, 1975, J I MATH APPL, V16, P321
[10]   A TRUST REGION ALGORITHM FOR EQUALITY CONSTRAINED OPTIMIZATION [J].
POWELL, MJD ;
YUAN, Y .
MATHEMATICAL PROGRAMMING, 1990, 49 (02) :189-211
12