Multi-objective building design optimization considering the effects of long-term climate change. (December 2021)

• Record Type: Journal Article
• Authors: Zou, Yukai; Lou, Siwei; Xia, Dawei; Lun, Isaac Y.F.; Yin, Jun
• Is Part Of: Journal of building engineering. Volume 44(2021)
• Publication Date: 2021-12
• Subjects: Climate change -- Artificial neural networks -- Multi-objective optimization -- Genetic algorithm -- Building performance; Building -- Periodicals; 690.05
• DOI: 10.1016/j.jobe.2021.102904
• Languages: English
• ISSNs: 2352-7102

Abstract

Abstract: Building performance is heavily influenced by weather conditions. Though the climate is changing vastly, few building performance optimizations (BPO) consider global warming over the life expectancy of the buildings. This paper develops a novel multi-objective BPO framework by the using simulation-based surrogate models under the future weather conditions that are determined by morphing the typical meteorological year (TMY) data. This framework is adopted to optimize a typical classroom in a hot and humid area under the future weather scenarios of representative concentration pathways (RCP) 4.5 and 8.5. The energy, thermal comfort and daylighting performances with and without considering the climate changes in the optimizations are compared and their notable differences are discussed. Under future climate, the optimization considering the future climate change improves the building performances significantly compared to the optimization without any climate change considerations (i.e., using the historical TMY). Especially, the winter discomfort hours in RCP4.5 and 8.5 decrease by 7.4% and 13.3%, respectively, when such future climate changes are considered in the BPOs, compared to the BPOs using the historical TMY data. The results show that BPO without considering climate change effects may cause non-negligible uncertainties. The proposed method can effectively improve the building performance in a changing climate. Highlights: Future hourly weather data was generated using the 'morphing' method. Parametric simulations were conducted to create building performance databases. Surrogate models were established to save computing cost. Multi-objective optimizations were performed under future weather conditions. Climate change should be seriously considered in building performance optimization.