OPTIMIZATION OF A GLAZED BUILDING USING A MULTI-OBJECTIVE GENETIC ALGORITHM TO REDUCE ENERGY DEMAND

Abstract

To avoid an increase in air conditioning energy demand, it is necessary to consider the best combinations between different design variables. Adopting multi-objective optimization in computer simulations can speed up the selection of these combinations. The objective of this pilot study was to present a multi-objective optimization method to reduce the energy demand of artificial air conditioning of a building, with predominantly glazed facades and inserted in the Bioclimatic Zone 3 of Brazil. The three-dimensional modeling of this building was carried out using the Rhinoceros 7 software, while its parameterization and simulation were performed using the Ladybug Tools plugin. The multi-objective optimization was performed through the Octopus plugin, using the SPEA-2 algorithm. To obtain the lowest energy demand for air conditioning, the objective functions analyzed were the lowest energy demand for heating and cooling. The modified variables were thermal transmittance of the glass, and depth and distance of the brises. The modified variables were thermal transmittance of the glass, and depth and distance of the brises. The best solution reduced the cooling energy demand by 8.9%, increased the heating energy demand by 97.9%, and reduced the total energy demand by 8.8%. Of the 500 combinations that took four days to complete, 20 were considered optimal. These results indicate that applying multi-objective optimization in thermal energy performance analysis in buildings with predominantly glazed facades is extremely efficient, due to the potential to automatically combine different variables and identify better combinations for parameters that negatively influence each other.