Conjugate duality in vector optimization (Q1197430): Difference between revisions

This paper considers a conjugate duality theory in vector optimization. Conjugate duality was fully developed in scalar optimization by Rockafellar and provides a unified framework for the duality theory. In 1980, the author and Sawaragi extended it to the case of multiobjective optimization. In 1982, Kawasaki refined their results and obtained a reciprocal duality by introducing the concept of ``supremum set'' in an extended Euclidean space. His supremum is based on the weak efficiency of the closure of a set in an extended sense. In 1988, the author newly defined the supremum of a set in the extended multidimensional Euclidean space. In this paper, the author extends his concept to a general partially ordered linear topological space. Based on this definition of supremum, conjugate maps and subgradients are introduced for vector-valued and set- valued maps. These concepts enable us to develop conjugate duality in vector optimization. The paper consists of four parts. Section 2 has a preliminary character with some basic definitions and a survey of the important results of the author's previous work. Section 3 is concerned with conjugate maps and includes an extension of Fenchel's Inequality. Section 4 is concerned with subgradients for vector-valued and set-valued maps. Finally, conjugate duality in vector optimization is explored in Section 5. Although these results are quite similar to the earlier works by the author and Sawaragi or Kawasaki, the new approach in this paper makes the proofs much easier and more understandable, and provides a better understanding of conjugate duality. In particular, it is very interesting that the convex assumptions for perturbation maps assures their subdifferentiability and hence the reflexive relation between a set- valued map and its biconjugate.