Energy Scheduling for Residential Distributed Energy Resources with Uncertainties Using Model-based Predictive Control

This paper proposes a reliable energy scheduling framework for distributed energy resources (DER) of a residential area to achieve an appropriate daily electricity consumption with the maximum affordable demand response. Renewable and non-renewable energy resources are available to respond to customers' demands using different classes of methods to manage energy during the time. The optimal operation problem is a mixed-integer-linear-programming (MILP) investigated using model-based predictive control (MPC) to determine which dispatchable unit should be operated at what time and at what power level while satisfying practical constraints. Renewable energy sources (RES), particularly solar and wind energies recently have expanded their role in electric power systems. Although they are environment friendly and accessible, there are challenging issues regarding their performance such as dealing with the variability and uncertainties concerned with them. This research investigates the energy management of these systems in three complementary scenarios. The first and second scenarios are respectively suitable for a market with a constant and inconstant price. Additionally, the third scenario is proposed to consider the role of uncertainties in RES and it is designed to recompense the power shortage using non-renewable resources. The validity of methods is explored in a residential area for 24 hours and the results thoroughly demonstrate the competence of the proposed approach for decreasing the operation cost.