基于关联规则挖掘的地下空间功能协同规律研究

Facing urban renewal, underground space use has exhibited a trend of diversification and complexification.  The synergy between underground spaces and the built environment has become a crucial factor influencing spatial performance.  Addressing the current lack of clarity regarding the coupling mechanisms between underground space function and built environmental, as well as the scarcity of quantitative planning tool for collaboration, this study transcends traditional experience-driven paradigm by establishing a data-driven systematic analytical framework.  First, based on underground space planning standards and point of interest (POI) classification systems, a mapping rule adapted to underground facilities is proposed to effectively extract underground point of interest (UPOI).  Second, by integrating UPOI with density-based spatial clustering of applications with noise (DBSCAN) and Convex Hull, a quantitative characterization method for underground space use is established.  Moreover, an indicator system for the built environment assessment is developed based on multi-source data.  Finally, an improved Apriori algorithm is introduced to reveal the association rules between underground and ground functions.  This study conducts empirical research using nine megacities in China, not only validating the feasibility of the proposed methods but also extracting universal laws of functional collaborative utilization through the superposition of association rules.  Targeted planning strategies are formulated, providing quantitative decision support for the refined governance of underground space in the new era.