Assessing the impacts of urban morphological factors on urban building energy modeling based on spatial proximity analysis and explainable machine learning

The investigation of the relationship between urban morphology and building energy consumption on a broad scale has garnered significant scholarly interest. Particularly in the early phases of urban building design, the optimization of urban morphological factors (UMFs) has demonstrated its efficacy and cost-effectiveness in enhancing the energy efficiency of urban buildings. This paper presents a framework for exploring the relationship between urban morphology and energy consumption in urban buildings. The framework encompasses defining and quantifying UMFs using a spatial proximity analysis approach, constructing an urban building energy model, and employing explainable artificial intelligence (AI) methods to analyze the impact of each factor on energy consumption. The findings identify the potential impact zones surrounding target buildings and identify 26 UMFs related to urban buildings and the road network. The study reveals high-impact UMFs significantly influencing energy consumption and provides corresponding recommendations for urban building planning. Moreover, the impact of these factors on energy consumption is similar across different building types, although there are variations in their contributions. The research contributes to identifying influential UMFs and provides practical implications for early urban building planning. The proposed methodology can be generalized to other cities, enabling broader applications of the framework.