On optimization over a polyhedral set and Augmented Lagrangians.

Abstract

Linear constraints have a long tradition in optimization problems. It means that the feasible set is a polyhedral set or can be described as a polytope. In Global Optimization,we use the characteristics to derive specific algorithms. In our contribution,wewill focus on the so-called Augmented Lagrangian approach where constraints are captured in Lagrangian terms and penalties.We show some first numerical analysis with the easiest case of having a linear objective.