Research
I work on forecasting problems in the energy sector, with a particular focus on the growing footprint of electric vehicles on the grid.
Electrical load forecasting
At EDF R&D, I model short- and medium-term electricity consumption for a range of internal stakeholders, using statistical and machine learning methods — generalised additive models, splines, wavelets, and deep learning.
EV charging demand
My PhD modelled the load created by electric vehicle charging using point processes and machine learning, addressing the specific volatility and sparsity of charging data.
Time series & forecasting methods
I supervise two PhD students extending this work: Guillaume Principato, on hierarchical conformal predictions applied to electric vehicle charging demand (since Dec. 2023), and Eloi Campagne, on graph neural networks applied to net demand forecasting (since Feb. 2024, after first joining me for an internship). I also collaborate with Kaoutar Bouaachra (Ecole Polytechnique de Paris / EDF Lab Paris-Saclay) on spatio-temporal modelling of EV charging demand.