Day-ahead electricity price prediction based on multiple ELM

Abstract

Aiming at the characters of day-ahead electricity price, an electricity price prediction model is established by combining the intelligent modeling methods. A new neural network method ELM is selected for its better performance to establish the basic day-ahead electricity price prediction model. Using the information fusion and ensemble ideas, a multiple ELM modeling approach is proposed 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 the new method has better performance.