Computational generation of energy-efficient rural residential floor plans under hot summer and cold winter climates based on segmentation algorithm model

Building sector accounts for a large energy consumption, while the “Digital Intelligence” technology has a promising perspective in the energy-efficient building design. The functional layout and the spatial geometry have significant impacts on the building energy consumption. Compared to the current research status, the novelty of this work is that a segmentation algorithm model with capacities of generative design and energy consumption optimization was proposed to the rural residential buildings. The segmentation algorithm model was used to obtain the design building parameters, optimize the functional space layout and generate the design schemes. Among the design schemes of floor plans generated from the proposed algorithm model, the most energy-efficient one was exported, by assigning the large energy demand rooms to the low energy consumption intensity locations, and selecting the minimum sum of the ratios of room perimeter to the total area of floor plan. A comparison study between the base case and the generated scheme was conducted, under the Hangzhou’s climate, and a difference of 792 kW·h between the two buildings was realized, with energy saving rate of 6.4 %. Therefore, through the proposed segmentation algorithm model, the design schemes of the rural residential buildings could be automatically optimized and generated.