A quasi-newton quadratic penalty method for minimization subject to nonlinear equality constraints

被引:6
|
作者
Coleman, TF [1 ]
Liu, JG
Yuan, W
机构
[1] Cornell Univ, Dept Comp Sci, Ithaca, NY 14850 USA
[2] Cornell Univ, Cornell Theory Ctr, Ithaca, NY 14850 USA
[3] Univ N Texas, Dept Math, Denton, TX 76203 USA
[4] Cornell Univ, Ctr Appl Math, Ithaca, NY 14853 USA
基金
美国国家科学基金会;
关键词
nonlinearly constrained optimization; equality constraints; quasi-Newton methods; BFGS; quadratic penalty function; reduced Hessian approximation;
D O I
10.1023/A:1008730909894
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
We present a modified quadratic penalty function method for equality constrained optimization problems. The pivotal feature of our algorithm is that at every iterate we invoke a special change of variables to improve the ability of the algorithm to follow the constraint level sets. This change of variables gives rise to a suitable block diagonal approximation to the Hessian which is then used to construct a quasi-Newton method. We show that the complete algorithm is globally convergent. Preliminary computational results are reported.
引用
收藏
页码:103 / 123
页数:21
相关论文
共 50 条