Parameterization applied to building energy performance

Abstract

The parameterization of energy efficiency measures is used in thermal and energy simulation processes in which gradual evaluations through simulation aim to obtain solutions for the building design. The procedures were improved through the decades towards the optimization of solutions, which optimized parameters have interrelated impacts on the building performance. Optimization methods involve complex sensitivity analysis, which adopts complex methods as Latin Hipercube for sampling or Neural Networks for evaluations, but also simplified analysis using linear regression equations. The Brazilian Labeling Program of Inmetro uses regression parametric equations of to identify the energy efficiency level of housing units. This study investigates shape parameters of a housing model using those regression algorithms. The variation of two shape parameters (width and depth of interior rooms), affected several labeling parameters, which results indicate the direction to introduce a shape grammar for energy efficiency of dwellings.