The implementation of the relevant management system makes the road-parking behavior standardized, while increasing the difficulty of temporary parking of operational vehicles such as taxis. Therefore, in order to improve the relevant management measures and promote the sustainable development of the taxi industry, it is necessary to survey the demand for taxi parking and study the layout of taxi stops. To process the GPS data of the taxis, and to extract the loading and unloading positions of the passengers from the spatial trajectory data, big data analysis technology is used. Compared with the data obtained using traditional survey means, the spatial trajectory data reflects the situation of the whole system, which can make the analysis more accurate. K-means cluster analysis was used to determine community demand. Finally, the immune optimization model was used to determine the optimal taxi stand location. The problem of taxi stand location at the level of urban network from two dimensions of quantity and spatial distribution is solved in this paper. The location of 10 taxi stands can not only meet the parking needs of regional taxis, but also reasonably allocate urban resources and promote sustainable development. This study also has a certain reference value for relevant management departments.