Short-term load forecasting for microgrids based on fuzzy logic

The ubiquitous penetration and integration of distributed and renewable generation has necessitated load forecasting in microgrids. Generation, supply and storage of energy must be suitably planned with respect to load demand on the microgrid. In this study, a short-term load forecasting (STLF) technique, which is based on fuzzy logic, is developed and analyzed. The developed fuzzy logic (FL) based STLF method is used to forecast load demand supported by an 11 kilo-volt (kV) substation in Providence, Rhode Island. Power consumption or load demand is predicted over an interval of one hour. Fuzzy rules which are based on previous year load demand, temperature and time of the day parameters are prepared. Four fuzzy inference systems (FIS) or four models for the sake of simplicity, one for each of the four seasons in Providence are modeled in Matlab using the Fuzzy Logic Toolbox (FLT). Subsequently, using Simulink Editor of the Simulink software four Simulink models are built. Real-life data of the 11 kV substation in Providence and temperature values (i.e weather data) for Providence are used at the simulation stage which is done using the Simulink software in Matlab. The mean absolute percentage error (MAPE) is then calculated to evaluate the performance of each of the four models.