Charging station planning is critical in the implementation of public transport electrification. However, in cities, there is insufficient experience and knowledge of the factors that influence the utilization of charging stations, particularly charging stations aimed at serving a fleet of electric taxis. Shenzhen is one of the pioneers in promoting electric taxis. In this paper, we collect large-scale datasets from Shenzhen and provide a data-driven space–time analysis of the relationship between charging station utilization and urban form and demand for taxi services. We use a Random Forest Regression model to explore these relationships and apply a Shapley value method to interpret the results. We find that demand for taxi services, measured as hourly pick-up and drop-off densities, have a non-linear relationship with utilization. Metro station density positively correlates with utilization, whereas the relationships between population density, land-use entropy, road density, and bus station density are more complicated.

充电站规划对于公共交通电气化的实施至关重要。然而，在城市中，对于影响充电站使用的因素，尤其是旨在为电动出租车车队服务的充电站，缺乏足够的经验和认识。深圳是推广电动出租车的先行者之一。在本文中，我们收集了深圳的大规模数据集，并提供了充电站利用率与城市形态和出租车服务需求之间关系的数据驱动时空分析。我们使用随机森林回归模型来探索这些关系，并应用 Shapley 值方法来解释结果。我们发现，以每小时接送密度衡量的出租车服务需求与利用率呈非线性关系。