Big Data Driven Simulation-Based Dynamic Traffic Assignment and Empirical Studies
Date:
Chen, X., Zhang, S., Shen, K., Ye, Y. & Sun, W. (2016). Big Data Driven Simulation-Based Dynamic Traffic Assignment and Empirical Studies. The 11th Annual Conference of ITS China, Chongqing, China, November 16-18, 2016. (Best Paper Award)
Abstract: With the continuous development of China’s urbanization, a rapidly increasing amount of travel demand brings heavy pressure to transportation infrastructure, operations and traffic management. It is of significance to support decision making of traffic management agencies by understanding and evaluating the dynamic impact of implementing traffic management strategies on road networks. This paper takes Hangzhou as an example to develop a mesoscopic simulation-based dynamic traffic assignment (DTA) model for the urban road network. First, the static origin destination (OD) trip matrix is estimated by using cellular signaling data. Secondly, both supply and demand calibration of this model are conducted on the basis of remote microwave sensors and license plate recognition cameras, including the calibration of the link fundamental diagram, road impedance function estimation, and dynamic OD estimation based on DTA results. Finally, we quantitatively assess the impact of large-scale implemented work zones before the 2016 G20 Summit on the road network performance using the developed model. Results showed that the network-wide average travel time increased by 56.60%, while the network-wide average speed decreased by 36.11%. The big data driven simulation-based DTA model developed in this paper has wide prospects of applications in the evaluation and optimization of comprehensive transportation management strategies.