Responsive Signal Control Design for Urban Traffic Network

This paper investigates different design of traffic responsive signal control system. The control system aims to balance traffic density over a signalized road network through minimizing queue sizes on the links. For real time applications, the system has to operate quickly with reasonable degree of accuracy. This study investigates the fidelity of traffic model that we need for real time control operations. We focus on feedback type control, and compare two different systems developed respectively by the deterministic queuing model (DDQ) and Gazis' store-and-forward model (SFM). The difference between the two traffic models is that deterministic queuing model updates the traffic estimates every second, while the store-and-forward adopts a 'quasi-steady state' approximation and only updates the traffic estimates every signal cycle. The proposed system design will be evaluated by simulation and results are discussed. The study will provide insight on developing effective real-time urban traffic control system.