An Online Multi-Level Energy Management System for Commercial Building Microgrids With Multiple Generation and Storage Systems

This paper presents an online multi-level energy management system for local microgrids of commercial buildings that integrate roof-top photovoltaic sources, battery storage systems, utility grids, diesel generators, supercapacitors, and commercial buildings consisting of active orchestrated loads, to solve the uncertainty problem of sources and loads, while also optimizing the local microgrid operation cost of commercial buildings and the utilization rate of local renewable energy. The energy management system includes a long-term rolling optimization level, rule-based optimization level, and load demand optimization level. At the long-term rolling optimization level, an online rolling method of data restructuring is proposed, where measurement data, short-term prediction data, and day-ahead prediction data are reconstructed to reduce the uncertainty in photovoltaic source prediction and load demand prediction. Four methods are proposed for the energy management system and simulated in MATLAB/Simulink under three typical weather conditions, cloudy, sunny, and rainy. Simulation results show that the performance of Method 3 is closest to that of Method 4, whose data conditions are ideal; Method 3 reduces the operational cost of the commercial building microgrid and improves the utilization rate of photovoltaic sources, at the slight cost of non-critical load shedding.