A Comparative Study of Short-Term Load Forecasting Methods for Improved Energy Management in Microgrids with Battery Storage

This study presents a comparative analysis of short-term load forecasting methods aimed at enhancing energy management algorithms in off-grid microgrids with battery storage. The evaluated methods include load forecasting using Markov chains, extended Kalman filters, and autoregressive integrated moving average (ARIMA). The microgrid under consideration comprises a geothermal generator, a biomass generator, a photovoltaic (PV) system, and a battery storage system. The predicted data, combined with weather forecast data, are used to improve energy management algorithms, optimize energy costs, and prolong the lifespan of the battery storage system.