Load and wind power forecasting methods for the optimal management of isolated power systems with high wind penetration

International audienceLoad demand forecasting is essential for the management of autonomous power systems. Forecasts are required for the optimal scheduling of power units for the next 2-48 hours depending on the power system size. Where such systems are subject to increasing renewable energy penetration, this renewable generation capacity is often treated as a 'negative load' due to its inherent unpredictability. A range of methods are presented and evaluated for the efficient forecasting of load and wind power generation in isolated power systems. The best of these methods have been implemented within an on-line advanced control advice system within the dispatch centre of the island of Crete. Load demand forecasting is essential for the management of autonomous power systems. Forecasts are required for the optimal scheduling of power units for the next 2-48 hours depending on the power system size. Where such systems are subject to increasing renewable energy penetration, this renewable generation capacity is often treated as a `negative load' due to its inherent unpredictability. A range of methods are presented and evaluated for the efficient forecasting of load and wind power generation in isolated power systems. The best of these methods have been implemented within an on-line advanced control advice system within the dispatch centre of the island of Crete