A Distributional Interpretation of Uncertainty Sets in Unit Commitment Under Uncertain Wind Power

This paper proposes a new distributionally robust optimization (DRO) framework for unit commitment under uncertain wind power. The proposed framework minimizes the worst-case expected total cost over an ambiguity set of possible probability distributions. Unlike the other DRO models that typically exploits variance and covariance data of random variables, this framework uses a distributional interpretation of uncertainty sets to construct the ambiguity set, and it can be solved as an equivalent problem that resembles a conventional two-stage robust linear program. Case studies demonstrate that the proposed model may effectively capture ambiguous distribution information and improve system performance.