Identification and optimization of recurrent congestion driven by multisource data on highway

As an important part of the urban transportation system, the comprehensive operational efficiency of highway is crucial for the smooth and sustainable development of the entire transportation network. In this paper, a shortest path traffic flow assignment method under the constraint of license plate recognition (LPR) points is proposed by integrating multi-source heterogeneous traffic sensor data, and the assigned traffic flows are used to calculate the road saturation of road links. The congestion condition of highways in different time periods is calculated by GPS trajectory Dataset. Then the two indices of road saturation and congestion condition are combined to identify recurrent congestion points by visual analysis. Finally, the optimization strategy is empirically studied by analyzing the traffic flow changes before and after the road network structure adjustment. The research results provide theoretical and data support for Hubei Provincial Traffic Planning Department to solve the recurrent highway congestion.