Comparison of models to identify thermal characteristics of multi-layer building walls using inverse methods

The objective of this study was to evaluate inverse techniques for estimating the resistance and the capacity of walls in real weather conditions. The inverse techniques were applied on temperatures measured by sensors located in the wall. The performance of four different models were compared. We demonstrated a high correlation with some parameters. A solution was proposed to limit the number of parameters identified, while allowing the identification of the thermal resistance and thermal capacity of the walls. The high uncertainty applied to some experimental data led to the exclusion of these data from the inverse method to improve its convergence. Most of the models were highly accurate in estimating thermal resistance and capacity of a lightweight wall using a set of experimental data.