One-hour-ahead wind speed prediction using a Bayesian methodology

The contribution of wind power in market-driven power systems together with the uncertain nature of the wind resource have led to many research efforts on methodologies to predict future wind speed/power production. Applications such as the operational balancing market in the UK would benefit from accurate one-hour-ahead forecasts of the available power from all generators, wind being no exception. This paper focuses on one-hour-ahead wind speed prediction using a Bayesian approach to characterise the wind resource. To test the approach, two years of wind speed data from a weather station were modelled as an autoregressive process. In this paper, the methodology used is described together with the model employed and prediction results are presented and compared to the persistence method. The results obtained indicate that Bayesian inferencing can be a useful tool in wind speed/power prediction, particularly due to the flexibility inherent to the methodology

I. Introduction

The growing participation of wind power in modern market-driven power systems is becoming a significant problem due to the uncertain and variable nature of the wind resource. This has motivated R&D into techniques for better characterisation and prediction of the wind resource and the power produced by wind farms.