To develop an effective traffic control algorithm for an oversaturated intersection, the video detection technique is employed in this paper, to make up for the shortcomings of methods based on loop detector data. A prediction model for arrival traffic volumes in oversaturated conditions was developed based on the three-order moving average method. Then considering the impacts of oversaturation on vehicle travel time between upstream and downstream intersections, an identification model was brought forward to judge whether the intersection is in an oversaturated condition. Taking minimizing queue length during the entire oversaturated period as the objective, a cycle length optimization method was developed. Field traffic data are collected from an oversaturated intersection in Changchun city and simulations in VISSIM were conducted to test the signal timing scheme output by the optimization algorithm. Finally, the optimization control scheme was compared with the current signal timing scheme of the intersection, and then conclusions were given.
展开▼