On Thursday, Raphael Saavedra discusses two-stage robust optimization with mixed-integer recourse and its applications to unit commitment
The presentation will be held next Thursday, July 19, at 2 PM in room 406 at the Cardeal Leme building
Abstract: One of the most common approaches for both stochastic and robust unit commitment is a two-stage formulation where the commitment decisions -- comprised mostly by binary variables -- represent the first stage, while the energy dispatch represents the second one, also known as recourse. Within this kind of formulation, the recourse is generally modeled as a linear programming problem, as all variables concerning generation, power flow, and phase angles are continuous. However, it might be relevant to include binary or integer variables in the recourse -- for example, in the presence of fast units which can be committed after the uncertainty is observed, or in the case of flexible units which can assume several states. In light of this, we analyze a recently proposed methodology to solve two-stage robust optimization problems with mixed-integer recourse and its possible applications in unit commitment.