Geometric programming with multiplicative slack variables

John R. McNamara

Naval Research Logistics Quarterly, 1986, vol. 33, issue 3, 501-512

Abstract: A posynomial geometric programming problem formulated so that the number of objective function terms is equal to the number of primal variables will have a zero degree of difficulty when augmented by multiplying each constraint term by a slack variable and including a surrogate constraint composed of the product of the slack variables, each raised to an undetermined negative exponent or surrogate multiplier. It is assumed that the original problem is canonical. The exponents in the constraint on the product of the slack variables must be estimated so that the associated solution to the augmented problem, obtained immediately, also solves the original problem. An iterative search procedure for finding the required exponents, thus solving the original problem, is described. The search procedure has proven quite efficient, often requiring only two or three iterations per degree of difficulty of the original problem. At each iteration the well‐known procedure for solving a geometric programming problem with a zero degree of difficulty is used and so computations are simple. The solution generated at each iteration is optimal for a problem which differs from the original problem only in the values of some of the constraint coefficients, so intermediate solutions provide useful information.

Date: 1986
References: Add references at CitEc
Citations:

Downloads: (external link)
https://doi.org/10.1002/nav.3800330314

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wly:navlog:v:33:y:1986:i:3:p:501-512

Access Statistics for this article

More articles in Naval Research Logistics Quarterly from John Wiley & Sons
Bibliographic data for series maintained by Wiley Content Delivery ().