A Comparative Analysis of XGBoost and Temporal Convolutional Network Models for Wind Power Forecasting
Quoc‐Thang Phan, Yuan‐Kang Wu, Quốc Dũng Phan
Abstract
Forecasting renewable power generation becomes more important for the operation of modern power systems. However, the largest challenge for integrating wind power into power systems is its characteristics of stochastic and intermittent. Fortunately, various machine learning or deep learning methods have been proposed in recent years. The structures of deep neural networks, such as temporal convolutional networks (TCN), have the ability to capture patterns in time series data. On the other hand, XGBoost is a tree-based method, which is commonly applied to supervised learning problems. This work used the two methods to predict short-term wind power generation. Based on the forecasting results, the used XGBoost model outperforms traditional artificial neural network (ANN), long short-term memory (LSTM) recurrent neural network, and TCN models.