A novel prediction modeling scheme based on multiple information fusion for day-ahead electricity price

Abstract

On the basis of the information fusion idea, a novel multiple information fusion modeling method is proposed. Several artificial neural networks are used to fuse the information of data. And then the results of information fusion by ANNs will be fused again according to their performance. Using the novel multiple information fusion scheme, a new modeling approach is presented to establish the prediction model. The day-ahead electricity price prediction model is tested by the real data. The experiments demonstrate that the new prediction model established by multiple information fusion method has better performance.