Short-Term Wind Energy Forecasting

摘要

As the world strives to identify and develop sustainable alternatives to fossil fuels a fundamental obstacle to successful integration has been the inability to accurately estimate the potential yield from sustainable energy sources. The most difficult of all sustainable resources is Wind, due to its dynamic nature. Many approaches exist to generate forecasts for wind regime. In general, these models can be classified as either involving a Numerical Weather Prediction model (NWP) or not. The past fifteen years have been a very intensive period for forecasting research and development. However, it is also clearly demonstrated that this work is being carried out from the perspective of the large utility. Whilst this is acceptable in the Danish, Swedish and American markets, were the majority of wind parks are owned and operated by such organizations, it does not hold true for developing markets where the majority of wind parks are under the operational control of smaller organizations, which do not have sufficient expertise or budgetary ability to pursue such ambitious projects. This study presents a number of methodologies for forecast generation and compares the approaches to the industry standard across a variety of forecast horizons.