On Lagrangian duality in vector optimization: applications to the linear case

We recall a general scheme for vector problems based on separation arguments and alternative theorems, and then, this approach is exploited to study Lagrangian duality in vector optimization. We show that the vector linear duality theory due to Isermann can be embedded in this separation approach. The theoretical part of this paper serves the purpose of introducing two possible applications. Some well-known classical applications in economics are the minimization of costs and the maximization of profit for a firm. We extend these two examples to the multiobjective framework in the linear case, exploiting the duality theory of Isermann. For the former, we consider the minimization of costs and of pollution as two different and conflicting goals; for the latter, we introduce as second objective function the profit for a competitor firm. This allows us to study the relationships between the shadow prices referred to the two different goals and to introduce a new representation of the feasible region of the dual problem.