Multi-objective optimization based on surrogate models for sustainable building design: A systematic literature review

Abstract

Surrogate models can overcome the building performance simulation complexity and replace time-consuming simulation engines within multi-objective optimization studies. Although this approach can leverage the development of sustainable building design, just a few studies integrate these concepts collectively. Therefore, this paper explores simulation studies that combine these techniques through a Systematic Literature Review (SLR). Specifically, this study illustrates the state-of-the-art literature, main challenges, and opportunities regarding multi-objective optimization based on surrogate models. The three significant contributions resulting from this study are: 1) a list of research gaps with issues partially addressed, 2) a framework for future evaluations, and finally, 3) directions for future research in the field. Based on 54 documents analyzed, there are research areas that demand further attention and present opportunities for significant contributions. For instance, there has been limited exploration of data mining and transfer learning techniques, a lack of optimizations considering uncertainty parameters that could enhance the resilience of the projects, and a deficiency in building decarbonization discussions. Furthermore, developing user-friendly tools or web applications could streamline the replication of the optimization approach, thus promoting more sustainable building design. Simultaneously, ethical standards could support a research environment that is transparent, accountable, and trustworthy. Finally, this SLR consolidates previous scientific achievements, acting as a valuable resource to facilitate the adoption of multi-objective optimization based on surrogate models for future researchers, thereby serving as a functional and reference guide and allowing the design of more sustainable buildings.