Data-driven distributionally robust reserve and energy scheduling over Wasserstein balls

In order to address the wind power uncertainty in the security constrained economic dispatch (SCED), this study proposes a two-stage data-driven distributionally robust reserve and energy scheduling (DDRRES) model that considers the loss of wind spillage and load shedding. This model aims to minimise the total cost while ensuring that the operating constraints are satisfied on the adjustable uncertainty set, of which the boundaries are decision variables. Unlike the previous approaches, which assumed that the underlying true probability distribution (PD) of uncertainty is known, the proposed model does not rely on the specified distribution but extracts the information from historical data directly. The ambiguity sets, i.e., Wasserstein balls are constructed to contain the possible PDs. Fixing the boundaries of adjustable uncertainty set, the operational risk is the expected loss under the worst-case PDs over Wasserstein balls. Thus the operational cost and operational risk can be balanced by adjusting the adjustable uncertainty set. After the tractable formulation of DDRRES is obtained, such a model is solved with the column-and-constraint generation method. The performance of the proposed approach is verified on a 6-bus test system and a 118-bus system.