Large-scale microscopic traffic simulation based on license plate recognition data and OpenStreetMap

Microscopic traffic simulation can provide scientific support for traffic design, traffic planning, traffic monitoring, and traffic demand management, and how to construct accurate and efficient microscopic traffic simulation is an important research direction. Current research on microscopic traffic simulation mainly focuses on the basic theory, such as the car following model and lane changing model. However, there is a lack of research on the practice and application of microscopic traffic simulation, especially for large-scale microscopic traffic simulation. In this study, we proposed a simple and efficient method for large-scale microscopic traffic simulation, and built a city-level microscopic traffic simulation system of Xiaoshan District, Hangzhou, China as an example. OpenStreetMap (OSM) data and license plate recognition (LPR) data were firstly fused, and then the road network, traffic infrastructure and travel information of vehicles were obtained based on the fused data. Next, the travel demand was obtained using the dynamic traffic assignment method and route choice algorithm. On this basis, the Simulation of Urban MObility (SUMO) platform was used for city-level microscopic traffic simulation. Finally, a calibration method was proposed to calibrate the microscopic traffic simulation system. The results show that the proposed method can simulate the traffic operation dynamics well.