Wind Power Prediction Using EMoFS Technique

Abstract

It is crucial for the energy market to predict the generated wind power of a wind park one day ahead. However, the error rate of the wind power prediction is high due to i) the natural fact that the behaviour of wind is stochastic, non-stationary and intermittent, ii) the fact that the relation between produced wind power and instant wind speed is nonlinear. In this study, a novel approach that is utilizing EMoFS technique in wind power prediction is proposed. In order to verify the validity of the proposed approach, error rate of the predictions of the proposed approach is compared with the ones that are generated by artifical neural networks based approach. It is observed that the proposed approach generates slightly better results compared to the generic approach that utilizes artificial neural networks.