Metodologia para o projeto preliminar de edifícios utilizando otimização multiobjetivo baseada na simulação de desempenho

Abstract

Buildings energy consumption has a great energetic and environmental impact worldwide. The architectural design has great potential to solve this problem because the building envoltória exerts influence on the overall system performance, but this is a task that involves many objectives and constraints. In the last two decades, optimization studies applied to energy efficiency of buildings have helped specialists to choose the best design options. However, there is still a lack of optimization approaches applied to the early design stage, which is the most influential stage for the building energy efficiency over its entire life cycle. Therefore, this article presents a multiobjective optimization model to assist designers in the preliminary building design, by means of the PAES (Pareto Archived Evolutionary Strategies) algorithm with the EnergyPlus Simulator coupled, to evaluate the solutions. The search process is executed by a binary array where the array components evolve over the generations, together with the other building components. The methodology aims to find optimal solutions with the lowest constructive cost associated with greater energy efficiency. In the case study, it was possible to simulate the process of using the optimization model and analyse the results in relation to: bioclimatic principles, economic performance, environmental performance, energy performance, usability and accuracy, proving that the tool serves as support in building design. Optimal solutions achieved energy savings and operational CO₂ emissions over typical standards, as well as short-term return on investment in four different types of weather.