Thermoenergetic optimization of a school building: discussion of the performance of four multiobjective evolutionary algorithms

The low thermoenergetic performance of buildings leads to increasing final consumption of electricity in the construction sector. The demand for air conditioning has contributed significantly to this increase and may be directly related to inadequate architectural planning. Through the establishment of a multi-objective optimization strategy, it is possible to identify performance improvement parameters for these buildings. This study discusses the performance of four multi-objective optimization algorithms applied to a standard project on early childhood education, developed by the Brazilian government program Proinfância, implemented in bioclimatic zone 2. Energy use intensity (IUE) for heating and cooling were defined as objective functions and five variables were changed: the thickness of the thermal insulation of the external walls and the floor, the absorptance of the external walls and roof, and solar orientation. The SPEA2 algorithm showed the best permance, achieving a solution capable of reducing IUE by approximately 92% in heating and 52% in cooling, when compared to the reference case.

Keywords:
Computer simulation; Simulation-based optimization; Multi-objective optimization; Energy efficiency