Unified Models and MILP-Based Scheduling Method for Home DC Microgrids: A Novel Approach to Overloading

DC microgrids (MGs) have emerged as an alternative interconnection method for DC-type loads and distributed energy resources (DERs). Owing to the vulnerability of grid-connected converters (GCCs) to overloading, it is crucial to effectively manage both the controllable devices within DC MGs and the power exchange through GCCs to ensure the stable and economical operation of DC MGs. Among the various types of DC MGs, this paper proposes a novel scheduling method specifically for home DC MGs, which are small-scale DC MGs, to account for various controllable devices and mitigate overloading at GCCs more effectively. The proposed method primarily addresses three limitations in the previous research: the use of device-specific models, GCC power limits expressed by hard constraints, and the absence of effective strategies for reducing the risk associated with prediction errors. Initially, controllable devices are classified into three types based on their operational characteristics: power-controllable, state-controllable, and time-controllable devices. For each of these types, a unified model is developed, which will then be used in the scheduling problem. For stable and economical operation, three prioritized strategies are proposed: 1) minimizing overloading at the GCC, i.e., GCC limits are represented by soft constraints, 2) minimizing operational costs, and 3) addressing the required power demand as quickly as possible, while delaying the use of stored energy as late as possible, thus reducing load shedding possibility due to prediction errors. The scheduling problem in which to implement these strategies is formulated as a mixed-integer nonlinear programming (MINLP) problem, for which finding optimal solutions can be challenging. Therefore, a method to relax the MINLP problem into a mixed-integer linear programming (MILP) problem is also proposed. Finally, the effectiveness of the proposed method is validated through various case studies using MATLAB.