Environmental performance-driven optimization of rural guesthouse clusters in South China: a human-AI collaborative decision-making workflow

Abstract

To address the environmental exigencies of rural built environments in Jiangmen City, Guangdong Province (112°E, 22°N), this study investigates a synergistic pathway for morphological optimization and microclimatic mitigation under South China’s hot-humid conditions. A human-AI collaborative framework is established, integrating parametric multi-objective optimization (MOO) with an AI-generated content (AIGC)-enabled perceptual feedback loop, thereby reconciling quantitative performance simulation with qualitative vernacular spatial design. The optimized spatial configurations demonstrate significant environmental dividends, characterized by an 88.47 kWh/m2 reduction in annual average heat radiation and a 0.21°C decrement in the average Universal Thermal Climate Index (UTCI). Critically, these physical enhancements correlate with a 12.5% reduction in annual “Strong Heat” exposure hours, providing a robust thermal buffer to attenuate cooling load requirements. Furthermore, the implementation of AIGC as a perceptual filter facilitates the resolution of inherent conflicts between technical performance and vernacular aesthetics, yielding a 35% augmentation in decision-making efficacy. This work delineates a transferable methodological trajectory for sustainable rural development, aligning computational tool innovation with the United Nations Sustainable Development Goals, particularly SDG 7 (Affordable and Clean Energy), SDG 11 (Sustainable Cities and Communities), and SDG 13 (Climate Action).