IN-PERSON London. Section B

Abstract

Traditional village communities are often formed from the bottom-up, lacking unified planning and management, where the relationship between people and the environment is characterized by continual compromise. In this process, differences in regional meteorological elements also shape people’s tolerance to extreme climates. This paper aims to innovate traditional planning and design methods through a performance and data-driven design process while incorporating traditional construction experience. The steps are as follows: First, basic information about the courtyard site, such as form and scale, is extracted and inputted as independent variables into the Grasshopper Ladybug Tools platform. Using the Wallacei_x plugin based on Performance-driven multi-objective generative design to optimize the best solutions for different types of courtyards. The initial analysis focuses on indoor and outdoor thermal comfort, illumination, and energy consumption. Secondly, combination design studies are conducted based on selected courtyards. The merits and demerits of different courtyard organizations are compared from the perspective of outdoor wind comfort, and optimized combination models are chosen. Thirdly, the data set about the courtyard parameters generated in the first step will be given corresponding performance labels based on the iterations screening of the Pareto front solution set and the secondary judgment of comfort. Input these corresponding data sets into the XGBoost algorithm program and generate a prediction equation. Finally, the optimal solution sets generated in the first and second steps are mapped into the scheme design process. Adjustments are made as per actual requirements during this period, and several schemes generated are evaluated for performance using the predictive equations from the third stage. Ultimately, an optimized design scheme is achieved. This article aims to explore new ideas for village planning and design through this research.