With the accelerating implementation of China's dual-carbon strategy and the rapid proliferation of new energy vehicles (NEVs), the supply–demand imbalance in urban electric vehicle (EV) charging infrastructure has emerged as a critical bottleneck constraining the high-quality development of green transportation. Existing static management frameworks are fundamentally ill-suited to handle the tidal fluctuations of charging loads and the polarised spatial allocation of resources. This study takes Weiyang District, Xi'an, as a representative urban prototype and develops an integrated Smart Charging Digital Twin Command Platform (SCDTCP) encompassing perception, diagnosis, decision-making, and control functions. Multi-source heterogeneous data — comprising Gaode Maps point-of-interest (POI) coordinates, seventh national census grid data, and a 740-respondent micro-level EV owner survey — are fused and subjected to Mahalanobis-distance-based multivariate outlier removal to construct a high-fidelity data foundation. At the macro-spatial level, a three-dimensional kernel density estimation (3D-KDE) method and a Supply–Demand Matching Index (SDMI) are innovatively introduced to precisely identify charging service deserts and their attenuation boundaries. At the micro-level, an ordered logistic regression (O-Logit) model is constructed to quantitatively confirm that perceived ICE vehicle encroachment on charging bays is the primary driver of user satisfaction decline. Subsequently, a Maximum Covering Location Problem (MCLP) model and an M/M/c multi-server queueing model are deeply integrated to adaptively generate spatially targeted station-siting plans and time-of-use (TOU) pricing strategies. Finally, the digital twin platform enables closed-loop hardware–software control spanning from global situational awareness to flexible command dispatch. Empirical evaluation demonstrates that the proposed strategy effectively redistributes 69.23% of peak-hour overload, eliminates inter-street SDMI disparities, and significantly improves infrastructure throughput and user satisfaction. The study advances the paradigm shift in charging governance from ad hoc infrastructure expansion toward data-intelligent targeted management, thereby providing a replicable engineering framework for equitable charging service delivery and low-carbon transport transition.