On meteorological forecasts for energy management and large historical data: A first look

Cyril Voyant (1, 2) , Cédric Join (3, 4, 5) , Michel Fliess (6) , Marie-Laure Nivet (2) , Marc Muselli (2) , Christophe Paoli (7)

1 Centre de Radiothérapie
2 SPE - Sciences pour l'environnement
3 NON-A - Non-Asymptotic estimation for online systems
4 CRAN - Centre de Recherche en Automatique de Nancy
5 ALIEN
6 LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau]
7 GSU - Université Galatasaray

This communication is devoted to a comparison between various meteorological forecasts, for the purpose of energy management, via different time series techniques. The first group of methods necessitates a large number of historical data. The second one does not and is much easier to implement, although its performances are today only slightly inferior. Theoretical justifications are related to methods stemming from a new approach to time series, artificial neural networks, computational intelligence and machine learning. Several numerical simulations are provided and discussed.

Mots clés: scaled persistence; multilayer perceptron; meteorology; time series; big data; quick fluctuations; persistence; forecasting; trend

Domaines: Ingénierie de l'environnement; Apprentissage [cs.LG]; Base de données [cs.DB]; Réseau de neurones [cs.NE]; Machine Learning [stat.ML]; Logique [math.LO]; Informatique ubiquitaire; Automatique / Robotique

Cyril Voyant, Cédric Join, Michel Fliess, Marie-Laure Nivet, Marc Muselli, et al.. On meteorological forecasts for energy management and large historical data: A first look. International Conference on Renewable Energies and Power Quality, ICREPQ'15, Mar 2015, La Coruña, Spain. ⟨hal-01093635v2⟩