Geometric Programming with Slack Constraints and Degrees of Difficulty

Abstract

An algorithm has been developed capable of solving the generalized signomial geometric programming problem, subject to linear or nonlinear constraints of any type, which are either tight or nonbinding at optimality. The algorithm was specifically developed to deal with those problems having degrees of difficulty. Necessary (and sometimes sufficient) conditions for optimality were obtained without increasing the degrees of difficulty in the problem, or using approximation techniques.