Optimality Conditions and Duality in Nonsmooth Adjustable Robust Optimization Problems
摘要
In this paper, we consider nonsmooth adjustable robust optimization problems and necessary/sufficient conditions based on qualification conditions and generalized convexity concepts. Verifiable sufficient conditions and useful relations for these qualification conditions are provided. Our analysis covers a wide range of uncertain sets commonly used in adjustable robust optimization, and we employ a dual approach to address non-reformulated cases of uncertain set, while also providing a direct approach to these problems through reformulation.