Forecasting Wind Power Generation

Wind energy is increasingly recognized as a necessary feature of modern and future energy portfolios. However, its actual integration in the electricity networks and markets presents a complex challenge. The main reason for that is that, contrary to conventional types of energies like gas, coal or hydropower, wind power generation is a direct function of the meteorological conditions which humans have no control over. An essential contribution to an optimal integration of this renewable energy into power systems and markets is then the forecasting of its power output at lead times ranging from a few minutes up to 10 days ahead.

The book introduces the general mathematical and statistical framework for the problem of wind power forecasting, and also considers a range of practical aspects ranging from the evaluation of forecasts, to the communication of forecast uncertainty and it use in decision-making. The authors start by covering the essential aspects of the data used as input (of both electrical and meteorological nature), then laying down the general models and methods of wind power forecasting with emphasis on statistical methods. The complementary approaches of point forecasting, probabilistic forecasting and forecast scenarios are also dealt with. This book is a valuable tool for scientists and engineers working in wind energy related R&D and industry, as well as graduate students and non-specialist researchers with interest in the fields of forecasting, statistical science, meteorology and renewable energy.