eduzhai > Applied Sciences > Engineering >

Neural Network Middle-Term Probabilistic Forecasting of Daily Power Consumption

  • Save

... pages left unread,continue reading

Document pages: 21 pages

Abstract: Middle-term horizon (months to a year) power consumption prediction is a mainchallenge in the energy sector, in particular when probabilistic forecasting isconsidered. We propose a new modelling approach that incorporates trend,seasonality and weather conditions, as explicative variables in a shallowNeural Network with an autoregressive feature. We obtain excellent results fordensity forecast on the one-year test set applying it to the daily powerconsumption in New England U.S.A.. The quality of the achieved powerconsumption probabilistic forecasting has been verified, on the one hand,comparing the results to other standard models for density forecasting and, onthe other hand, considering measures that are frequently used in the energysector as pinball loss and CI backtesting.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...