A revised support vector regression (SVR) ensemble model based on boosting algorithm (SVR-Boosting) is presented in this paper for electricity price forecasting in electric power market. In the light of characteristics of electricity price sequence, a new triangular-shaped 为oss function is constructed in the training of the forecasting model to inhibit the learning from abnormal data in electricity price sequence. The results from actual data indicate that, compared with the single support vector regression model, the proposed SVR-Boosting ensemble model is able to enhance the stability of the model output remarkably, acquire higher predicting accuracy, and possess comparatively satisfactory generalization capability.
A systematic study on the electrical load forecasting for large-scale iron and steel companies was made. After analyzing the electrical load′s characteristics, an algorithm framework for the load forecasting in iron and steel complex was formulated based on model combination and scheme filtration. The algorithm features data quality self-adaptation, convenient forecasting model extension, easy practical application, etc., and has been successfully applied in Baoshan Iron and Steel Co Ltd, Shanghai, China, resulting in great economic benefit.
