Urban Resident Travel Characteristics Mining And Functional Areas Identification Based On Multi-source Data In Lanzhou

With the development of science and technology,mobile positioning equipment and computer technology are gradually matured.A large number of urban spatio-temporal big data has emerged.The big data provides data support for exploring the characteristics of urban residents’ travel behavior and identifying the urban functional structure.Based on multi-source big data such as urban GPS trajectory data,POI data and mobile base station data,the travel patterns of urban residents and the distribution characteristics of traffic hotspots are studied by data mining and visualization techniques.The research results provide travel guidance for residents,and guidance for taxi drivers to choose passenger routes and search locations.The distribution characteristics of urban functional areas and their interaction laws are studied..The experimental results provide a reference for the decision-making of urban planning and management departments.The specific research content is mainly summarized in the following:(1)An urban traffic hotspot area identification algorithm is proposed based on CLARA clustering algorithm.The original GPS data is preprocessed,and the GPS trajectory is matched to the map.The study area is divided by using the urban grid method and the amount of up and down is counted in the grid.The traffic hotspots are identified by the CLARA clustering algorithm.The results show that the proposed algorithm is effective for identifying urban traffic hotspots.The traffic hotspot areas of Lanzhou are gradually expand from the core area of the city from the inside to the outside.The expansion trend is consistent with the characteristics of Lanzhou’s valley-shaped topography,and expands outward in the form of "dumbbell".(2)A method for quantifying the intensity of regional interaction based on a directed weighted complex network is established.The identified traffic hotspots are equivalent to nodes and flows between hotspots are equivalent to the edge weight of the complex network.The traffic hotspots complex network is built and the evaluation index interaction strength this defined.The intensity of interaction between urban traffic hotspots is analyzed.The result shows that temporal and spatial characteristics of the residents’ travel behavior and the urban spatial interaction characteristics are different during weekdays and weekends.(3)An identification algorithm is proposed for urban functional area using Apriori algorithm.First,the strong association rules are extracted based on Apriori algorithm by combining taxi GPS trajectory,POI data and mobile phone base station.Secondly,the urban area is divided by the city grid and the Thiessen polygon.The strong association rule matrix and the functional area identification indexes are constructed.An identification algorithm is proposed for urban functional areas.Finally,the distribution characteristics of urban functional areas are studied in Lanzhou.The effectiveness of the functional area identification algorithm is verified by combining the enrichment factor and the traffic mining based on kernel density estimation.