Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Civil and Environmental Engineering

Degree Program

Civil Engineering, PhD

Committee Chair

Li, Zhixia

Committee Co-Chair (if applicable)

Sun, Zhihui

Committee Member

Sun, Zhihui

Committee Member

French, Mark

Committee Member

Nasraoui, Olfa

Author's Keywords

Connected and autonomous vehicles; reservation-based intersection; scheduling; dynamic programming; time complexity


Reservation-based intersection control has been evaluated with better performance over traditional signal controls in terms of intersection safety, efficiency, and emission. Controlling connected and autonomous vehicles (CAVs) at a reservation-based intersection in terms of improving intersection efficiency is performed via two factors: trajectory (speed profile) and arrival time of CAVs at the intersection. In an early stage of the reservation-based intersection control, an intersection controller at the intersection may fail to find a feasible solution for both the trajectory and arrival time for a CAV at a certain planning horizon. Leveraging a deeper understanding of the control problem, reservation-based intersection control methods are able to optimize both trajectory and arrival time simultaneously while overcoming the infeasible condition. Furthermore, in order to achieve real-time control at the reservation-based intersection, a scheduling problem of CAV crossing the intersection has been widely modeled to optimize the intersection efficiency. Efficient solution algorithms have been proposed to overcome the curse of dimensionality. However, a control methodology consisting of trajectory planning and arrival time scheduling that can overcome the infeasible condition has not been explicitly explained and defined. Furthermore, an optimal control framework for joint control of the trajectory planning and arrival time scheduling in terms of global intersection efficiency has not been theoretically established and numerically validated; and mechanisms of how to reduce the time complexity meanwhile solving the scheduling problem to an optimal solution are not fully understood and rigorously defined. In this dissertation, a control method that eliminates the infeasible problem at any planning horizon is first explicitly explained and defined based on a time-speed-independent trajectory planning and scheduling model. Secondly, this dissertation theoretically defines the optimal control framework via analyzing various control methods in terms of intersection capacity, throughput and delay. Furthermore, this dissertation theoretically analyzes the mechanism of the scheduling problem and designs an exact algorithm to further reduce the time complexity. Through theoretical analyses of the properties of the scheduling problem, the reasons that the time complexity can be reduced are fundamentally explained. The results first validate that the defined control framework can adapt to extremely high traffic demand scenarios with feasible solutions at any planning horizon for all CAVs. Under extensive sensitivity analyses, the theoretical definition of the optimal control framework is validated in terms of maximizing the intersection efficiency. Moreover, numerical examples validate that a proposed scheduling algorithm finds an optimal solution with lower computation time and time complexity.