1998
DOI: 10.1007/bf01213996
|View full text |Cite
|
Sign up to set email alerts
|

A modified trust region algorithm for hierarchical NLP

Abstract: Large-scale design optimization problems frequently require the exploitation of structure in order to obtain efficient and reliable solutions. Successful algorithms for general nonlinear programming problems with theoretical underpinnings do not usually accommodate any additional structure within the problem. In this article modifications are made to a trust region algorithm to take advantage of hierarchical structure without compromising the theoretical properties of the original algorithm. 1. Start with some… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
3
0

Year Published

2001
2001
2002
2002

Publication Types

Select...
3
1

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(3 citation statements)
references
References 14 publications
0
3
0
Order By: Relevance
“…Consider the case where the modi ed design is a scaled design ¹K ¤ , as given by Eq. (16). Then, from Eq.…”
Section: Scaling Of the Initial Stiffness Matrixmentioning
confidence: 99%
See 1 more Smart Citation
“…Consider the case where the modi ed design is a scaled design ¹K ¤ , as given by Eq. (16). Then, from Eq.…”
Section: Scaling Of the Initial Stiffness Matrixmentioning
confidence: 99%
“…where 1X L and 1X U are predetermined lower and upper limits, respectively, on the design variable changes. An alternative approach, used in trust region algorithms, 16 is to restrict the solutions to some region around X ¤ by constraints of the form approximations (such as the Taylor series), where small changes in the design variables are assumed. They are not effective for the CA method applied to nd accurate solutions for large changes in the design.…”
Section: Nearly Scaled Designsmentioning
confidence: 99%
“…An alternative approach, used in trust region algorithms, 16 is to restrict the solutions to some region around X ¤ by constraints of the form…”
Section: Nearly Scaled Designsmentioning
confidence: 99%