Grid-connected residential solar-photovoltaic (PV) and battery systems are increasingly popular types of microgrids. Determining the optimal energy management system (EMS) strategy for such microgrids depends on many factors, such as power demand, solar irradiation, and system costing. The energy flow for a residential PV-battery microgrid is thus studied in detail. Three algorithms are used, including load-levelling, peak-shifting, and an original model predictive control (MPC) EMS. PV cells, battery overpotentials and degradation are simulated with physically-meaningful models. Real data from Long Island, New York, are used to simulate the load power demand, solar irradiation, utility costs, degradation costs, and PV credits. Both load and PV forecasting error are considered. Results for the base cases demonstrate the advantage of MPC EMS. Simulation parameters are then varied to show that the simulated cost savings depend on the costing assumptions and forecasting error.
Assessing Degradation-Aware Model Predictive Control for Energy Management of a Grid-Connected PV-Battery Microgrid
2022-06-15
1774060 byte
Conference paper
Electronic Resource
English
Neuro-Fuzzy-Based Model Predictive Energy Management for Grid Connected Microgrids
BASE | 2020
|