Multi-objective optimization design for ultra-low-energy college dormitory buildings based on the actual occupant behavior

Building energy efficiency plays a crucial role in reducing carbon emissions and achieving sustainable development. However, during the building design stage, due to the discrepancy between idealized energy use behavior and actual conditions, the effectiveness of energy-saving measures often falls short of expectations, hindering the achievement of energy efficiency goals. To solve this issue, a multi-objective optimization framework for ultra-low-energy college dormitories based on actual energy use behavior was proposed, with building energy consumption (BEC), thermal discomfort time (TDT), and initial incremental cost (IIC) as optimization objectives. First, based on the actual energy use behavior, an EnergyPlus model was constructed for college dormitories in Changsha, China. Then, the key building design parameters were identified through global sensitivity analysis. Furthermore, an agent model was constructed using artificial neural networks to rapidly predict building performance. Finally, for different building design scenarios, the non-dominated sorting genetic algorithm II was used to perform multi-objective optimization of the design parameters. The results showed that air infiltration and air-conditioning (AC) performance were the most critical factors influencing BEC; air infiltration, south wall insulation, and north wall insulation were the primary factors affecting TDT, while AC performance, interior wall insulation, and floor insulation had the greatest impact on IIC. Compared to the base case (Standard JGJ 134-2010), seven optimal solutions reduced BEC by 9.3% to 50.7%, TDT by 1.2% to 7.8%, and increased IIC by 5.6 to 288.4 CNY/m2. This study can provide methodological guidance and data support for the construction of ultra-low-energy college dormitories in China.