A Robust and Model Predictive Control Based Energy Management Scheme for Grid-Connected Microgrids

摘要

As the number and scale of microgrids that connected to the main grid have grown obviously in recent years, and furthermore as the output of renewable energy and the level of user load in microgrids are usually uncertain, how to maintain the economic operation of microgrids while reducing microgrids' potential impact on the main grid has become a challenging problem. An energy management scheme which combines the robust optimization approach and model predictive control approach is proposed to cope with this problem. The proposed scheme divides the energy management work into two stages: a day-ahead stage and an intraday stage. In the day-ahead stage an energy management model is proposed to get the interaction plan between the main grid and the microgrid, and the robust optimization approach is utilized to deal with the uncertainties of renewable energy and user load. In the intra-day stage the model predictive control approach is adopted to cope with the updated predictive information in order to keep the real-time power balance of microgrids and reach this system's economic operational goal, in which the deviation between the interaction plan and the actual result of interaction is especially penalized to reduce the negative impact on the main grid. At last, the effectiveness of the proposed scheme of energy management for grid-connected microgrids is validated by simulation results, and comparison has also been conducted with other two schemes.