A preliminary analysis of some parameters' Importance in building Energy Efficiency assessment

Abstract

The objective of this work is to implement a meta-modelling method for sensitivity analysis in both building energy analysis models and prioritizing energy saving measures. So, a hybrid approach based on artificial neural network (ANN) and Monte Carlo (MC) simulation is proposed using information from a buildings' energy certification database. A detailed study of interactions between parameters is performed together with a variance-based method aiming at obtaining the Sobol indices for sensitivity analysis. The sensitivities of energy performance outcomes relatively to building parameters are obtained and the importance of each parameter is outlined.