In this paper, a GPS/GIS integrated system for urban traffic flow analysis is proposed. Urban GIS-T data are used to construct the GIS map of urban area when the system begins its work. Afterwards, real-time GPS data of probe vehicles are periodically collected to implement the location amendment. And then, location-amended GPS data are dynamically fitted with the adaptive traffic flow model. By using fitted traffic flow models, states of traffic flow are ultimately estimated along rolling time periods. Compared with conventional works, the proposed system successfully preserves the continuity of vehicle travels without additional data sources. Experiments based on GIS-T data of Shanghai and GPS signals of taxis in Shanghai indicate that the new system is both efficient and robust.
展开▼
