Short-term power prediction and downtime classification

Short-term wind forecasting has seen an increase in interest in recent times. While the commercial use of such short-term forecasts is often investigated in detail, it is only recently that the quality of prediction is getting significant importance. This importance is expected to continue to increase over time. The choice of different models for prediction, methods of comparison, markers of comparison, and the quality of prediction are often major research question that has a direct impact on the availability of such power. The increase in data availability and popularity of data analytics has also led to an increase in such research and its need. On the other hand, early detection and classification of downtime based on measured data can lead to better operations and maintenance of such farms, leading to better-managed availability of wind energy. This is also significantly linked with the available data and the quality of analysis and similar questions around the choice of analysis methods and comparison of such methods exist. This chapter provides an overview of such approaches and along with some guidelines around choice and comparison of methods. Wind energy data from Ireland is considered in this regard.