Nonsmooth semi-infinite multiobjective optimization problems (Q2251568): Difference between revisions

\noindent This paper deals with a semi-infinite multiobjective optimization problem (SIMOP) involving nondifferentiable functions, whose constraints range on an arbitrary (possibly infinite) index set without a topological structure. Necessary conditions for (weakly) efficient solutions of a SIMOP are established. The tools include the approximate extremal principle, the fuzzy sum rule for the Fréchet subdifferential, the sum rule for the limiting/Mordukhovich subdifferential, and certain scalarization formulas of the coderivatives. Sufficient conditions for (weakly) efficient solutions of a SIMOP are provided by means of introducing the concepts of (strictly) generalized convex functions, defined in terms of the limiting subdifferential of locally Lipschitz functions. The authors propose types of Wolfe and Mond-Weir dual problems, and explore weak/strong dual relations under assumptions of (strictly) generalized convexity.