ISBN: 978-981-14-4787-7 DOI: 10.18178/wcse.2020.06.060
Short-term Load Forecasting Based on Wavelet Decomposition and XGBoost
Abstract— The development of intelligent power systems and the large-scale access of distributed power sources have continuously deepened the impact on the power distribution side, and have placed higher requirements on the accuracy of load forecasting. In this paper, a short-term load forecasting method based on wavelet analysis and XGBoost is proposed. First, use wavelet analysis to classify power loads in different frequency bands, and then use the XGBoost model for training and prediction of the classified loads. Finally, using real power data in a certain area as a sample, the average absolute percentage error (MAPE) and the mean squared error(MSE) are used to compare and analyze the 24th hour data predicted by XGBoost and SVM and LSTM, respectively. The results show that XGBoost has a better fit and higher accuracy for short-term power loads
Index Terms— power system; load forecasting; wavelet analysis; XGBoost
Ningning Zheng, Yunfeng Shao, Suli Zou, Zhongjing Ma
School of Automation, Beijing Institute of Technology, CHINA
Lvliang Power Supply Company, CHINA
Cite: Ningning Zheng, Yunfeng Shao, Suli Zou, Zhongjing Ma " Short-term Load Forecasting Based on Wavelet Decomposition and XGBoost " Proceedings of 2020 the 10th International Workshop on Computer Science and Engineering (WCSE 2020), pp.403-410, Shanghai, China, 19-21 June, 2020.