Multi-timescale optimization between distributed wind generators and electric vehicles in microgrid

Due to the low-cost and low-pollution property, distributed wind generators and electric vehicles (EVs) are becoming important supply and load types in the microgrid. However, it is still a challenge to integrate large amounts of wind power and EVs into the microgrid because of their nature of uncertainty. As the prediction accuracy of the wind power and EV parking events depends on the timescale, we propose a multi-timescale optimization model in this paper to improve the matching degree between wind supply and EV charging demand in the microgrid. We make the following contributions. First, the day-ahead planning is formulated as a Markov Decision Process to determine the optimal procurement of conventional energy from the power grid considering the uncertainties of wind power and EV parking event. Second, based on the idea of model predictive control, a real-time scheduling is considered to minimize the cost for satisfying the load balance using the improved prediction of wind power and EV parking event. Numerical results demonstrate the effectiveness of our model based on the real data.