Time-Dependent O-D Estimation of Urban Expressway with Multi-Source Data Fusion: A Case Study of Suzhou

Time-dependent origin-destination (O-D) is vital for dynamic traffic simulation and advanced traffic information systems. However, acquiring time-dependent O-D is challenging because large-scale and time-varying traffic data are required. This paper proposes a method of time-dependent O-D estimation based on multi-source data fusion. License plate recognition data and traffic flow survey data are integrated; time-dependent O-D is obtained synchronously. First, travel trajectory is classified according to actual data; the shortest path algorithm is used to obtain O-D. Then, combined with obtained O-D and main section flow survey data of the expressway ramp flow, the complete O-D of the whole expressway is obtained through O-D inversion; the time-varying distribution ratio of O-D is obtained by road section flow. Finally, a dynamic traffic simulation model of Suzhou expressway is constructed by DTALite; traffic condition data from AutoNavi are utilized to validate model performance. Results show that accuracy of the time-dependent O-D estimation model is acceptable, which can help to obtain time-dependent O-D for large-scale traffic networks.