The relationship between environmental factors and the indoor physical environment is very close, and external shading is considered an effective way to adjust the interaction between the indoor and outdoor environment. However, determining how to set up an external shading system remains a notable issue. In the early design stage, architects have adopted the process of designing the form and function first and then checking whether those characteristics meet the energy-saving specifications. However, this process involves a great deal of repetitive and inefficient work and cannot meet the requirements of energy savings and emission reductions in a global context. Therefore, it is particularly important to seek a design method that combines energy-saving design with form-based design. This paper takes a construction project in Northwest China as its research object. In this study, typical parametric models for external shading are designed. Furthermore, indoor performance objectives based on light environment analysis are proposed, and Ladybug Tools and the genetic algorithm (GA) are used for optimization and verification. The optimization results show that the adaptive shading system can significantly reduce the total cooling energy consumption per unit area in summer by 20% and 15%, respectively. The comfort level throughout the year improved by 14.8% (air conditioning on) and 4.7% (air conditioning off). This study proposes a fast and effective shading parametric design method for architects in the early stage, improving the efficiency and accuracy of performance-based design.

Although it is well established that thermal environments significantly influence travel behavior, the synergistic effects of points of interest (POI) and thermal environments on behavior remain unclear. This study developed a vision-based outdoor evaluation model aimed at uncovering the driving factors behind human behavior in outdoor spaces. First, Yolo v5 and questionnaires were employed to obtain crowd activity intensity and preference levels. Subsequently, target detection and clustering algorithms were used to derive variables such as POI attractiveness and POI distance, while a validated environmental simulator was utilized to simulate outdoor thermal comfort distributions across different times. Finally, multiple classification models were compared to establish the mapping relationships between POI, thermal environment variables, and crowd preferences, with SHAP analysis used to examine the contribution of each variable. The results indicate that XGBoost achieved the best predictive performance (accuracy = 0.95), with shadow proportion (|SHAP| = 0.24) and POI distance (|SHAP| = 0.12) identified as the most significant factors influencing crowd preferences. By extrapolation, this classification model can provide valuable insights for optimizing community environments and enhancing vitality in areas with similar climatic and cultural contexts.