New genetic algorithm-based workflow for multi-objective optimization of Net Zero Energy Buildings integrating robustness assessment

Net Zero Energy Buildings (NZEBs) play a key role to save energy. However, unpredictable scenarios and uncertainties about stability of energy production from renewables frequently cause measured performance of the building-plants system that differs from predictions. So, it is crucial to individuate the best alternatives in the early-stage design of the building and related energy systems, also considering uncertain circumstances. Hence, the aim of this paper is to enhance reliability of the building design process by providing an innovative workflow based on multi-objective optimization that encompasses robustness assessment of different design alternatives against uncertain scenarios. The standard genetic algorithm optimization routine has been deeply improved to set as a key performance indicator the robustness of energy performance of several building-HVAC system configurations, also considering the financial aspects optimization. Compliance with NZEB target of the obtained optimal solutions is then automatically verified by the algorithm. The new methodology combines open-source coding language Python and the dynamic energy simulation engine EnergyPlus. The robustness of each solution is evaluated through Taguchi method. First obtained results are related to energy-efficient solutions for the building envelope. Regarding the HVAC system, ground source (instead of air source) heat pumps and variable refrigerating flow systems are preferable. Finally, medium–high size of photovoltaic panels is preferred, although with high investment costs.