Machine learning for wind power prediction
University of New Brunswick
Wind power prediction is of great importance in the utilization of renewable wind power. A lot of research has been done attempting to improve the accuracy of wind power predictions and has achieved desirable performance. However, there was no complete performance evaluation of machine learning methods. This thesis presented an extensive empirical study of machine learning methods for wind power predictions. Nine various models were considered in this study, which also included the application and evaluation of deep learning techniques. The experimental data consisted of seven datasets based on wind farms in Ontario, Canada. The results indicated that SVM, followed by ANN, had the best overall performance, and that k-NN method was suitable for longer ahead predictions. Despite the findings that deep learning failed to give improvement in basic predictions, it showed the potential for more abstract tasks, such as spatial correlation predictions.