Development of a parsimony analysis methodology for urban energy simulation ; Développement d'une méthodologie d’analyse de la parcimonie pour la simulation énergétique urbaine

Research has to address global energy issues in order to reduce energy consumption and greenhouse gas emissions to limit climate change. This research includes the development of new urban simulation tools, called UBEM (Urban Building Energy Modelling), to help the collectivities, study offices and other energy transition stakeholders reduce energy consumption in the building sector. These UBEMs are composed of models that have to deal with the problems of lack of parametrization data and calculation costs related to urban simulation. Many models exist with different levels of detail, to simulate all the physical phenomena linked to the building, energetic systems or external solicitation while respecting these constraints. Due to the wide variety of models available, the user/modeler may find himself in situations where the selection of the most suitable models for his study can be tedious and complex. Thus, a methodology for carrying out a "parsimonious" simulation is proposed in this thesis. This parsimony of simulation aims at finding the right level of modelling by determining the trade-off point between: the number of input parameters and their uncertainty, the level of detail of the model with its simplifying hypotheses, the precision obtained and the simulation time, all for a given output and context. Key guidance indicators (KGI) are created from district characteristics and used to determine threshold values for the selection of the model complexity to be used. This contribution allows to advise and guide various users and modelers in their urban modelling, so as to promote a simulation which is not the most accurate, but the most suitable for the case study. Thus, the thesis proposes a methodology for the development of more parsimonious and therefore more effective decision support tools, allowing to make the various stakeholders aware of all the parsimony challenges: move from the logic of always more to just the right thing. ; La recherche doit répondre aux enjeux énergétiques globaux afin de ...