A two-stage robust optimal scheduling model for microgrids accounting for the uncertainties in wind turbine output

Abstract

To effectively address the impact of uncertainties in wind turbine output on the safe and economical operation of microgrids， this paper proposes a two-stage distributed robust optimal scheduling model for microgrids that accounts for these uncertainties. First， an improved deep neural network （DNN） is used to forecast the output of wind farms， and the forecasting error is calculated； then， a non-parametric kernel density estimation method is employed to analyze the forecasting error data， and a box type uncertainty set is constructed using the cumulative probability density curve of the error. Next， a two-stage robust optimal scheduling model with a min-max-min structure for microgrids is established. Finally， strong duality theory is applied to decompose the original problem into a master problem and subproblem， both of which are formulated as mixed-integer linear programming problems， and the optimal scheduling scheme is obtained through iterative solving. The case study results show that the proposed model demonstrates stronger robustness compared to traditional two-stage robust optimal models.

Keywords

• two-stage robust optimal
• non-parametric kernel density estimation
• box type uncertainty set
• economical microgrid dispatching