Accession Number:

ADA109528

Title:

Computational Advances in Large-Scale Nonlinear Optimization.

Descriptive Note:

Master's thesis,

Corporate Author:

NAVAL POSTGRADUATE SCHOOL MONTEREY CA

Personal Author(s):

Report Date:

1981-09-01

Pagination or Media Count:

135.0

Abstract:

This is a comparison of two state-of-the-art large-scale nonlinear optimization systems exhibiting unprecedented problem solution capabilities both in size of problem handled and method of solution. These codes are MINOS, developed by B.A. Murtagh and M.A. Saunders, and XS, developed by G. G. Brown and G. W. Graves. The codes are evaluated with respect to their problem solving capabilities and potential for practical application by analysts. Computational results are presented for thirteen nonlinear and nonlinear mixed integer test problems with from two to 793 variables 12 to 100 integer variables and one to 401 constraints. Portions of this work were presented at the CORSORSATIMS joint meeting in Toronto, May 1981. Author

Subject Categories:

  • Theoretical Mathematics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE