Stochastic energy scheduling in microgrids considering the uncertainties in both supply and demand

Microgrids have emerged as a promising paradigm to integrate the renewable generation units, energy storage systems, and dispersed loads. However, the inherent fluctuation of renewable generations adds significant uncertainty to the supply side of microgrid. On the other hand, due to the small scale of the microgrid, the load demand in microgrid also has higher uncertainty than that observed in utility grids. Furthermore, the supply uncertainty and demand uncertainty may have different distribution characteristics, e.g., Gaussian and non-Gaussian uncertainties may exist simultaneously. These uncertainties pose significant challenges to the microgrid energy management. To address these challenges, this paper establishes a new stochastic energy scheduling scheme for microgrids. In this scheme, the energy scheduling is formulated as a stochastic model predictive control problem, which incorporates the uncertainties in both sides of supply and demand. Using machine-learning techniques, the corresponding stochastic optimization problem is converted to a standard convex quadratic programming, and thus can be solved efficiently. A key feature of this scheme is that it handles the coexistence of Gaussian and non-Gaussian uncertainties. Simulation results validate the effectiveness of the proposed scheme.