Machine learning for energy demand model from a load curve

Hamza Mraihi (1) , Edi Assoumou (1) , Gilles Guerassimoff (1) , Valérie Roy (1)

1 CMA - Centre de Mathématiques Appliquées

The strong deployment of intermittent renewable energy in the European electric system has to be anticipated to optimize the power plant planning. A better knowledge of the correlation between the hour step load curves and the national level in the European country is crucial. Depending on the size of the future horizon we are regarding, the demand level determines the profitability of the solutions for the flexibility. The results presented in this communication represent the first step in the assessment of the forecasted electric load curve for a residential park of buildings. After a statistical analysis of the time series proposed by the European Network of Transmission System Operators for Electricity, we show some results of the application of machine learning technique for the demand projection characterization for prospective studies.

Mots clés: Residential & Tertiary; data analysis; Energy policy; Indicators

Domaines: Mathématiques [math]; Sciences de l'ingénieur [physics]

Hamza Mraihi, Edi Assoumou, Gilles Guerassimoff, Valérie Roy. Machine learning for energy demand model from a load curve. 29th European Conference on Operational Research EURO 2018, Jul 2018, Valencia, Spain. ⟨hal-01953117⟩