Integrated multi-stage sensitivity analysis and multi-objective optimization approach for PCM integrated residential buildings in different climate zones

The early design stage provides the greatest opportunities to achieve energy-efficient buildings; however, designers require relevant performance data to manage interdisciplinary and conflicting objectives. Thus, for the first time an integrated multi-stage sensitivity analysis and multi-objective optimization approach was developed for four different zones of arid climate in PCM integrated residential buildings. A multi-stage sensitivity analysis is proposed to identify the relative importance of early design stage parameters, which covers envelope thermophysics, building layout and energy-efficiency measures, considering energy and economic indicators. Three global sensitivity analysis approaches (Standardized rank regression coefficient, Partial rank correlation coefficient and Morris method) were employed in the first stage, while the local sensitivity analysis was employed in the second stage. A multi-objective optimization using the genetic algorithm was performed to obtain the Pareto frontier, and two sets of solutions were proposed: the most energy-efficient and the most cost-effective. Finally, multi-criteria decision making was carried out. It was found that sensitivity analysis results showed different rankings of parameters. For each climatic zone, a different set of optimal solutions existed. Compared to reference building, the best solutions reduced total energy consumption by 11.1–34.6%, while the payback period reduced by 17.6–48.5 years.