Development and testing of a micro-grid excess power production forecasting algorithms

Abstract Traditional electricity grids lack flexibility in power generation and load operation in contrast to smart-micro grids that form semi-autonomous entities with energy management capabilities. Load forecasting is invaluable to smart micro-grids towards assisting the implementation of energy management schedules for cost-efficient and secure operation. In the present paper is examined the 24h forecasting of excess production in an existing micro-grid. Alternative input parameters are considered for achieving an accurate prediction. The prediction can be used for scheduling the charging process of a thermal storage during weekends based on excess power production levels.