Optimizing Energy Efficiency and Comfort in Rural Self-Built Housing: A Comparative Study of GA and EA Algorithms

With the growing global focus on sustainable building design, reducing energy consumption and carbon emissions in buildings has become a critical research topic. This study explores the variations in electricity consumption, carbon emissions, and comfort levels influenced by building design parameters, and compares the applications of Genetic Algorithm (GA) and Evolutionary Algorithm (EA) in building optimization. The research is based on a Private residence in Huzhou, Zhejiang Province, where actual data were collected, including wall-to-window ratio, heating set point, cooling set point, roof shading coefficient, equipment power density, airtightness, and thermal transmittance of external walls, to construct a detailed building model. During the optimization process, DesignBuilder software was utilized for modeling, and both GA and EA optimization algorithms were implemented. The simulation results show that GA outperforms EA in terms of convergence speed, stability, and balancing multiple objectives. Specifically, GA reduced the annual electricity consumption from 862.26 kWh to 825.32 kWh, a 4.28% decrease, with the standard deviation dropping from 287.61 to 275.43 (a 4.25% improvement in stability). Discomfort time was significantly reduced by 31.56%, from 1030.2 hours to 705.12 hours, with substantial reductions in both standard deviation and interquartile range (IQR). Carbon emissions decreased from 552.62 kg to 533.86 kg, a 3.39% reduction, highlighting the potential for emission reduction. The results indicate significant differences between the two algorithms in optimizing building energy consumption, reducing carbon emissions, and enhancing comfort levels. GA demonstrates advantages in convergence speed and result accuracy, while EA excels in solution diversity and global search capability. Ultimately, this paper presents optimization recommendations based on both algorithms, aimed at providing a reference for future sustainable building design.