为识别城市公共充电网络的利用效率差异及时空异质性特征，支撑充电基础设施的存量优化与精细化配置，本文提出一种基于多源运营数据的充电站功能分型方法。基于7 572个公共充电站的建设、运营与订单数据，构建涵盖单枪使用频次、充电时长、能量输出及收益水平的多维评价指标体系，并结合主成分分析与无监督聚类实现充电站功能识别。进一步，结合核密度估计与城市用地分类，揭示功能类型、土地属性、时间节律之间的耦合关系。结果显示：高效型站点仅占17.52%，却承载83.74%的充电需求，而71.49%的站点长期处于低利用状态，反映出现有充电网络存在显著的结构性供需错配。不同用地类型对充电行为具有明显塑形作用：高效型站点在交通场站、加油/加气站及住宅区呈现差异化的时间节律与能量输出特征；常规型站点构成稳定的日常补能网络；低效型与闲置型站点则表现出“高容量-低需求”的空间失衡现象。研究表明，城市充电网络优化应由规模扩张转向存量重构与差异化配置。据此，从强化需求约束审批、低效站点动态腾退以及推进跨主体协同监管等方面提出了精细化的治理政策建议。本文提出的功能分型方法可为高效节点识别、低效资源治理及新增设施规划提供数据支撑与决策依据。

To identify the utilization efficiency differences and spatiotemporal heterogeneity characteristics in urban public EV(Electric Vehicle) charging networks and to support the optimization of stock and refined configuration of charging infrastructure, this study proposes a data-driven functional classification approach for charging stations using multi-source operational data. Based on the construction, operation, and order data of 7 572 public charging stations, a multidimensional evaluation index system is developed, including single-gun charging frequency, charging duration, energy output, and revenue performance. Principal component analysis and unsupervised clustering are further combined to identify functional station types. Kernel density estimation and urban land-use classification are then integrated to reveal the coupling relationships among functional types, land-use attributes, and temporal charging patterns. The results show that high-efficiency stations account for only 17.52% of all stations but serve 83.74% of the total charging demand, whereas 71.49% of stations remain underutilized for long periods, indicating a significant structural mismatch between infrastructure supply and charging demand. Land-use characteristics significantly shape charging behaviors. High-efficiency stations exhibit differentiated temporal rhythms and energy-output patterns across transport hubs, fuel stations, and residential areas. Regular stations constitute the stable daily charging network, while low-efficiency and idle stations exhibit pronounced "high capacity-low demand" spatial imbalance. The findings suggest that urban charging network optimization should shift from extensive expansion toward stock reconfiguration and differentiated allocation. Consequently, this paper proposes refined governance policy suggestions in terms of reinforcing demand-constrained approvals, implementing dynamic retirement for underutilized stations and promoting cross-sector collaborative supervision. The proposed functional classification framework provides data support and decision-making references for efficient node identification, underutilized resource management, and future charging infrastructure planning.