Robust Economic Dispatch Algorithm for Microgrid Based on Data-driven Expected Scenario Set Sequence

In the light of the uncertainty of new energy and load power, in this paper we propose a robust economic dispatch algorithm for microgrids based on data-driven expected scenario set sequence. The clustering method is employed to cluster a substantial amount of historical scenario data, so as to form a clustered scenario set sequence. This sequence is then reduced to an expected scenario set sequence based on probability. In the day-ahead scheduling stage, the economic dispatch problem of the microgrid is solved by the column constraint generation algorithm, with the feasibility of any scenario as the constraint condition and the probability weighted index of the microgrid operating cost corresponding to the desired scenario as the objective function. In the intra-day scheduling stage, the measurement data of new energy and loads are utilized to re-dispatch the microgrid based on the results of the day-ahead scheduling. Penalizing is conducted according to adjustments of traditional energy generation power and grid interaction power, to track the day-ahead schedule scheduling results. Thereby, the optimal microgrid equipment output is obtained to improve the economy of the microgrid. The effectiveness of the method is confirmed by a simulation case.