A Multi-timescale Coordinated Optimization Framework for Economic Dispatch of Micro-Energy Grid Considering Prediction Error

The integration of a large number of renewable energy sources (RES) into the micro-energy grid (MEG) brings certain challenges to its optimal dispatching, especially the impact of RES and load prediction error on optimal dispatching. Therefore, this paper proposes a multi-timescale coordinated optimization framework for economic dispatch of MEG considering prediction error, including three stages: day-ahead economic dispatch, intra-day rolling optimization and real-time adjustment. In the day-ahead and intra-day stages, the device's output is solved with the objective of optimal system economy. The MEG is divided into three subsystems in the real-time adjustment stage using distributed model predictive control (DMPC). The subsystems cooperate to correct the intra-day device output, which reduces the complexity of solving system optimization problems. Meanwhile, considering that the prediction horizon in DMPC usually remains constant during the optimization, which affects the control performance of DMPC. This paper analyzes the impact of the prediction horizon and control horizon on the control performance of DMPC. Then a DMPC method with adaptive adjustment of the prediction horizon according to the prediction errors of RES and loads power is proposed, which can balance the stability and prediction accuracy of the system. Finally, the case studies verify the feasibility of the proposed strategy.