Global sensitivity analysis based on multi-objective optimization of rural tourism building performance

Abstract

Morphological design has significant impacts on the performance of rural tourism buildings (RTBs). Multi-objective optimization has been practiced for three prototypes of RTBs in Tianjin, China to explore the RTB design with better energy load, indoor daylighting and thermal comfort performance. However, making decisions from the solution set is still a challenging task for decision-makers. So post-optimization analysis is conducted in this study, and multi-objective optimization solution set of the three RTB types is investigated through global sensitivity analysis (GSA) on the platform of R studio. To begin with, correlation coefficients of the parameters are analyzed by Spearman and partial correlation coefficients methods, and strong correlation relationships are discovered among design parameters of RTBs. Afterwards, Genizi and random forest-based cforest methods are adopted for GSA towards different performance of RTBs, and comprehensive analysis is implemented for the results. Finally, it's found that the depth of frequently used space has significant impacts on annual, cooling and heating load, and daylighting of RTBs. Besides, the height of skylight is discovered to be key important parameter for both cooling and daylighting, and the width as well as the window to wall ratio of the heating space in winter have important influence on winter thermal comfort performance. Overall, the framework and methodology for GSA in post-optimization phase is put forward. The influential mechanism of morphological design parameters on energy load, daylighting and thermal comfort performance of RTBs is explored on basis of the multi-objective optimization solution set. This study not only supplies scientific suggestions for decision-making of RTBs in Tianjin, China, but also offers reference for RTB design and practice in similar regions of the world.