PROPERTIES OF UPDATING METHODS FOR THE MULTIPLIERS IN AUGMENTED LAGRANGIANS

被引:28
作者
GLAD, ST
机构
[1] Department of Automatic Control, University of Lund, Lund
关键词
augmented Lagrangians; constrained optimization; Lagrange multipliers; Nonlinear programming; penalty functions; quasi-Newton methods; rate of convergence;
D O I
10.1007/BF00933239
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
The convergence properties of different updating methods for the multipliers in augmented Lagrangians are considered. It is assumed that the updating of the multipliers takes place after each line search of a quasi-Newton method. Two of the updating methods are shown to be linearly convergent locally, while a third method has superlinear convergence locally. Modifications of the algorithms to ensure global convergence are considered. The results of a computational comparison with other methods are presented. © 1979 Plenum Publishing Corporation.
引用
收藏
页码:135 / 156
页数:22
相关论文
共 25 条
  • [1] ABADIE J, 1970, INTEGER NONLINEAR PR
  • [2] [Anonymous], 1971, COMPUTATIONAL METHOD
  • [3] COMBINED PRIMAL-DUAL AND PENALTY METHODS FOR CONSTRAINED MINIMIZATION
    BERTSEKAS, DP
    [J]. SIAM JOURNAL ON CONTROL, 1975, 13 (03): : 521 - 544
  • [4] Broyden C. G., 1970, Journal of the Institute of Mathematics and Its Applications, V6, P222
  • [5] BUYS JD, 1972, THESIS LEIDEN
  • [6] COLVILLE AR, 1968, IBM3202949 NEW YORK
  • [7] DENNIS JE, 1974, MATH COMPUT, V28, P549, DOI 10.1090/S0025-5718-1974-0343581-1
  • [8] FLETCHER R, 1970, INTEGER NONLINEAR PR
  • [9] FLETCHER R, 1974, NONLINEAR PROGRAMMIN, V2
  • [10] GLAD ST, 1973, 7323 U LUND DEP AUT