Modeling real-time dynamic queue length for urban traffic control systems

In this study, we chose two isolated intersections to collect traffic volume per minute, queue length, and signal timing in Taipei City traffic control network. There are two queue-length estimation equations, the system method of vehicle scanning and the regression method. In order to verify the said models, another data set was collected to compare with the traffic performance of the said models. From the results of a field test, it is found that the regression model performs better than the other one in traffic estimation efficiency based on the existing traffic control system framework in Taipei City. Therefore, this model may be employed to represent traffic congestion, to evaluate signal timing plans, to generate suitable paths in vehicle route guidance systems, to detect traffic incidents, and to act as performance index of timing alternatives in the adaptive signal control systems.