Interval Optimization-Based Quantification of Adjustable Capacity in Microgrids and Its Accuracy Evaluation

In a microgrid (MG), distributed resources such as photovoltaics (PVs), energy storage systems (ESSs), electric vehicles (EVs), and flexible loads can facilitate power absorption and support for grid. However, the uncertainty in generation and load impacts the stability of MG operation, and traditional deterministic optimization methods struggle to determine the adjustable capacity of the MG to the grid. To address this, interval optimization theory is introduced into the multi-agent interaction model of the MG's PV-ESS-EV-grid-load system, optimizing to obtain the adjustable capacity interval of the MG to the grid. The day-ahead forecast interval of PVs and flexible loads is used to represent the uncertainty of generation and load. Based on the fuzzy c-means (FCM), 24 operational scenarios are clustered to combine PVs and flexible loads intervals from the perspectives of season, weather, and load. Two interval optimization models (the interval number model and the optimistic-pessimistic model) are established, and their adjustable capacity intervals are evaluated for accuracy. The case study reveals the adjustable capacity intervals of the MG in various operational scenarios throughout the year, and the coverage rate of the actual MG's exchange power with grid is used to assess its accuracy. The results indicate that the optimistic-pessimistic model has higher accuracy, reaching 94.95%.