The aim of the European Project ANEMOS is to develop accurate and robust models that substantially outperform current state-of-the-art methods, for onshore and offshore wind power forecasting. Advanced statistical, physical and combined modelling approaches were developed for this purpose. Priority was given to methods for on-line uncertainty and prediction risk assessment. An integrated software platform, 'ANEMOS', was developed to host the various models. This system is installed by several end-users for on-line operation and evaluation at a local, regional and national scale. Finally, the project demonstrates the value of wind forecasts for the power system management and market integration of wind power. Keywords: Wind power, short-term forecasting, numerical weather predictions, on-line software, tools for wind integration.

This paper presents the objectives and the research work carried out in the frame of the ANEMOS project on short-term wind power forecasting. The aim of the project is to develop accurate models that substantially outperform current state-of-the-art methods, for onshore and offshore wind power forecasting, exploiting both statistical and physical modeling approaches. The project focus on prediction horizons up to 48 hours ahead and investigates predictability of wind for higher horizons up to 7 days ahead useful i.e. for maintenance scheduling. Emphasis is given on the integration of highresolution meteorological forecasts. For the offshore case, marine meteorology is considered as well as information by satellite-radar images. An integrated software platform, ‘ANEMOS', is developed to host the various models. This system will be installed by several utilities for on-line operation at onshore and offshore wind farms for prediction at a local, regional and national scale. The applications include different terrain types and wind climates, on- and offshore cases, and interconnected or island grids. The on-line operation by the utilities will allow validation of the models and an analysis of the value of wind prediction for a competitive integration of wind energy in the developing liberalized electricity markets in the EU.

This paper presents the objectives and the research work carried out in the frame of the ANEMOS project on short-term wind power forecasting. The aim of the project is to develop accurate models that substantially outperform current state-of-the-art methods, for onshore and offshore wind power forecasting, exploiting both statistical and physical modeling approaches. The project focus on prediction horizons up to 48 hours ahead and investigates predictability of wind for higher horizons up to 7 days ahead useful i.e. for maintenance scheduling. Emphasis is given on the integration of high- resolution meteorological forecasts. For the offshore case, marine meteorology is considered as well as information by satellite-radar images. Specific modules are also developed for on-line uncertainty and prediction risk estimation. An integrated software platform, 'ANEMOS', is developed to host the various models. This system will be installed by several end-users for on-line operation at onshore and offshore wind farms for prediction at a local, regional and national scale. The applications include different terrain types and wind climates, on- and offshore cases, and interconnected or island grids. The on-line operation by the utilities will allow validation of the models and an analysis of the value of wind prediction for a competitive integration of wind energy in the developing liberalized electricity markets in the EU.