| dc.description.abstract | This research focused on proposing a real-time control for intersection traffic signals at different roadway levels, including isolated intersections, corridors, and grid networks. First, the real-time control algorithm (DORAS) for isolated intersections was derived, and was tested on a real-life intersection with actuated control and OPAC. The results showed that DORAS was able to significantly reduce average vehicle delay and stops. Secondly, an advanced control algorithm (MADM) for corridors and grid networks was introduced, which was developed based on DORAS and Max Pressure (MP) method by Varaiya. The algorithm was tested in simulation for corridor and network case with other methods.
The results showed that MADM outperformed the chosen real-time control methods, as well as the classic bandwidth maximization method in terms of average vehicle delay and stops. Finally, a relatively independent and isolated chapter, a mathematical algorithm that considers probabilistic vehicle presence was developed to treat the left-turn spillback at intersections given the locations of connected vehicles (CVs). Two consecutive intersections in College Station, TX were tested through simulation, in which the downstream intersection was subject to left-turn spillback. The results showed that the algorithm was able to significantly reduce the average vehicle delay at downstream intersection without impact the performance of upstream intersection. Future work includes testing the proposed real-time control algorithm on real corridors and networks, improving the computational efficiency, and comparing with other real-time control methods. | en |
