In order to improve the recognition efficiency and accuracy of urban polycentric structure and provide scientific basis for urban planning, this paper uses kernel density analysis, hotspot analysis and weighted overlay analysis to explore the spatial aggregation patterns and characteristics of urban elements, and designs a spatial weighted overlay analysis method based on public dependence, spatial aggregation degree and attribute values. By using the POI data with high accuracy, fast update and full coverage, which can reflect the spatial aggregation characteristics of urban elements, the method realizes the measurement of aggregation characteristics and the polycentric recognition of Zhengzhou city. The results show that the POI elements in main urban area of Zhengzhou city present an annular spatial distribution structure. It has the characteristics of centralizing in the center and spreading along the road. The polycentric structure is unbalanced. There are some differences between the planned sub-centers and the identified ones. Because of the remote location and lack of supporting facilities, the peripheral functional centers with lower development level is far from achieving the dispersal and evacuation role they should have. Therefore, it is necessary to improve the overall planning of the industry to enhance the outward radiation and inward attraction of the sub-center through the link of transportation.