Research On The Spatial Structure And Functional Area Identification Of Urban Travel Based On Taxi GPS Data

With rapid economic growth and urban expansion,large numbers of people are moving into cities,leading to a series of urban problems such as traffic congestion and tailpipe pollution.The daily travel of residents is a necessary activity to maintain the normal operation of cities.By studying the travel of large-scale people,we can reveal the inherent interaction characteristics of urban spatial structure and provide a reference basis for urban traffic management and residents’ travel.The rapid development of the Internet of Things and communication technologies has generated a huge amount of fine-grained traffic data,which provides us with an opportunity to recognize and understand the relationship between crowd travel patterns and urban spatial structure from a new perspective.Based on this,this paper uses one-week taxi trajectory data to reveal the complexity characteristics of taxi travel trajectory networks,the distribution of functional areas and the relationship between urban spatial structure from the perspective of the spatio-temporal patterns of residents’ travel,using complex networks,time series clustering and spatial analysis.The main research contents and conclusions are:(1)A quantitative analysis of residents’ travel characteristics from a spatial and temporal perspective reveals that indicators such as travel size and travel length have obvious commuting characteristics on weekdays,and residents prefer to use taxis for short and medium distance trips,and the spatial distribution of residents’ travel activities is more intensive on rest days,with a strong spatial and temporal evolution pattern for each travel characteristic.(2)Based on the trajectory data,a class of complex network diagram representing urban travel is constructed and analyzed using complex network measurement methods.At the same time,the unbalanced relationship between node and edge mobility,and the uncertainty of departure and arrival locations of the network have obvious spatio-temporal evolution patterns,and the overall node access shows obvious "core-edge" characteristics in different time periods.The stronger the change over time.(3)Based on the spatial and temporal changes of residents’ travel activities,a class of departure time sequence vectors and arrival time sequence vectors are constructed,and residents’ travel patterns are classified into clusters by using improved dynamic time regularization and clustering algorithms,and the functional attributes of land parcels are identified by combining the characteristics of residents’ travel curves and POI density and enrichment index.The results show that the different departure and arrival mode curves show different peaks in the morning peak,afternoon peak,evening peak,nighttime and early morning,and the corresponding parcels show a certain circle structure in the spatial distribution and their respective functional tendencies.It also reflects the spatial and temporal changes in the activities of different functional areas and people.