Optimality conditions and duality for robust nonsmooth multiobjective optimization problems with constraints (Q2420789)

This paper deals with robust nonsmooth multiobjective optimization problems of the form \[ (RMOP)\text{ }\min \text{ }f\left( x\right) ,\text{ s.t. }g\left( x,v\right) \in -S,\;\forall v\in V, \] where \(x\in \mathbb{R}^{n}\) (the space of decisions), the objective function \(f:\mathbb{R}^{n}\longrightarrow \mathbb{R}^{m}\) has components \(f_{k}\), \(k=1,\dots,m\), \(S\subset \mathbb{R}^{m}\) (the space of criteria) is a proper, closed and convex cone, the uncertain parameter \(v\)\ ranges in the uncertainty set \(V\), which is the cartesian product \(\prod\limits_{j=1}^{m}V_{j}\) of nonempty, convex and compact sets \(V_{j}\subset\mathbb{R}^{n_{j}}\), \(j=1,\dots,n\,,\) and the constraint function \(g:\mathbb{R}^{n+p}\longrightarrow \mathbb{R}^{m},\) with \(p=\sum\limits_{j=1}^{n}n_{j},\) has components \(g_{j}\), \(j=1,\dots,n\). Denote by \(F\) the robust feasible set, i.e., the solution set of the semi-infinite system \(\left\{ g\left(x,v\right) \in -S,\,\forall v\in V\right\}\). The paper provides robust necessary optimality conditions for weakly robust efficient solutions and properly robust efficient solutions (w.r.t. a given proper, closed and convex cone with nonempty interior \(Q\subset \mathbb{R}^{n}\)) by making use of the below Assumption \(A\) and a suitable concept of generalized convex function: \begin{itemize} \item \(g\) satisfies Assumption \(A\) whenever the following conditions hold: \begin{itemize} \item[(A1)] \(g\) is locally Lipschitz with respect to the first argument and uniformly on \(V\) w.r.t. to the second argument. \item[(A2)] The function \(v_{j}\longmapsto g_{j}\left( x,v_{j}\right) \) is concave on \(V_{j},\) \(j=1,\dots,n.\) \end{itemize} \item A function \(f:X\longrightarrow \mathbb{R}^{q},\) where\ \(X\subset\mathbb{R}^{n},\)\ is said to be \(Q\)-convexlike if \(f\left(X\right) +Q\) is conxex [\textit{J. B. G. Frenk} and \textit{G. Kassay}, J. Optim. Theory Appl. 102, No. 2, 315--343 (1999; Zbl 0937.90127)]. \end{itemize} More in detail, Theorems 3.1--3.4 provide Fritz-John-type and Kuhn-Tucker-type robust necessary optimality conditions for a given \(\overline{x}\in F\) to be robust \ (weakly, properly) efficient by assuming that \(g\) satisfies Assumption \(A\) and \(f\) is \(Q\)-convexlike. Moreover, Theorems 3.5--3.6 provide sufficient conditions for a given \(\overline{x}\in F\) to be robust \ (weakly, properly) efficient by assuming that the pair of functions \(\left( f,g\right) \) is type II \((Q,R_{+}^{m})\)-generalized convex at \(\overline{x}\) (a technical condition introduced by the authors in Definition 2.7). The same assumption allows to prove (weak, strong, converse) Mond-Weir-Type robust duality theorems (Theorems 4.1--4.5) as well as Wolfe-type duality theorems (Theorems 5.1--5.2).