There is a close relationship between the traffic status of various regions in the city. It is of great reference significance to explore the spatial dependence of urban traffic activities and study the internal reasons for the change of regional traffic status for road network optimization and the analysis of residents' travel behavior. In this paper, with the help of trajectory data of taxi in Ningbo, the average driving speed of taxi in different blocks during the characteristic period was calculated, and the global Moran 'sI was further calculated based on this index, and LISA cluster graph was drawn. Finally, the spatial autocorrelation and spatial-temporal evolution trend of traffic state in central Ningbo were described quantitatively by four sets of indicators, namely migration direction, migration Angle, migration distance and area of low-value area. The results show that: 1) the real-time speed of taxi has a strong spatial correlation, and the global Moran 'sI fluctuates with the passage of time; 2) the spatial distribution of local Moran 'sI high aggregation and low aggregation is relatively stable. The high aggregation area is distributed in the edge of the city, while the low aggregation area is concentrated in the center of the main urban area, the center of Zhenhai District and the main area of Beilun District. 3) the spatio-temporal evolution law of low-value downtown area is obvious, and there is a significant difference between working days and non-working days.

CICTP