Distributionally robust optimal power flow with contextual information

• A novel chance-constrained OPF model that accounts for contextual information The dependence between the wind power forecast and its error is exploited robust to ambiguity in conditional distribution Compared others, our method produces reliable solutions are cheaper In this paper, we develop a distributionally formulation of Optimal Power Flow problem (OPF) whereby system operator can leverage information. For purpose, exploit an set based on probability trimmings optimal transport through which dispatch solution protected against incomplete knowledge relationship uncertainties context conveyed by sample their joint distribution. We provide tractable reformulation proposed under popular conditional-value-at-risk approximation. By way numerical experiments run modified IEEE-118 bus network with uncertainty, show how substantially benefit from taking into account well-known statistical point outputs associated prediction error. Furthermore, conducted also reveal distributional robustness conferred probability-trimmings-based approach superior bestowed alternative approaches terms expected cost reliability.