This study presents a methodology to estimate reliable vehicle trajectories using data collected from the camera and lidar sensors of autonomous vehicles by combining BEV transformations, 3D object recognition followed by vehicle following models and probabilistic deep learning techniques.
Integrating a time-dependent max-pressure signal control policy considering intersection demand every hour and a queue-based GLOSA algorithm.
This study suggests a bi-level calibration framework for Intelligent Driver Model using dynamic traffic flow states attained from clustering a trajectory data.
This study focuses on improving the responsive time of emergency vehicles at signalized intersections by using advanced traffic control methods.
This study proposes a dynamic lane allocation framework that enables efficient utilization of road space and congestion reduction by responding to changes in demand for different modes of transportation within a complex transportation network.
This study analyzed the effect of introducing SKIP-STOP operation method on the improvement of the level of service during the morning peak hours on the Gimpo Goldline, which has extreme congestion.