Sensitivity analysis for models with dynamic inputs: a case study to control the heat consumption of a real passive house

Floriane Anstett-Collin (1) , Thierry A. Mara (2) , Jeanne Goffart (3) , Lilianne Denis-Vidal (4)

1 CRAN - Centre de Recherche en Automatique de Nancy
2 PIMENT - Physique et Ingénierie Mathématique pour l'Énergie, l'environnemeNt et le bâtimenT
3 USMB [Université de Savoie] [Université de Chambéry] - Université Savoie Mont Blanc
4 LMAC - Laboratoire de Mathématiques Appliquées de Compiègne

In this communication, we perform the sensitivity analysis of a building energy model. The aim is to assess the impact of the weather data on the performance of a model of a passive house, in order to better control it. The weather data are uncertain dynamic inputs to the model. To evaluate their impact, the problem of generating coherent weather data arises. To solve it, we carry out the Karhunen-Loève decomposition of the uncertain dynamic inputs. We then propose an approach for the sensitivity analysis of this kind of models. The originality for sensitivity analysis purpose is to separate the random variable of the dynamic inputs, propagated to the model response, from the deterministic spatio/temporal function. This analysis highlights the role of the solar gain on a high-insulated passive building, during winter time.

Mots clés: Global Sensitivity Analysis; Karhunen-Loève Expansion; Dynamic Inputs; Building Energy Model

Domaines: Automatique / Robotique

Floriane Anstett-Collin, Thierry A. Mara, Jeanne Goffart, Lilianne Denis-Vidal. Sensitivity analysis for models with dynamic inputs: a case study to control the heat consumption of a real passive house. 10ème Conférence Francophone de Modélisation, Optimisation et Simulation, MOSIM'14, Nov 2014, Nancy, France. ⟨hal-01088077⟩