Traffic Signal Timing Optimization on Superstreet Using Ecology-Based Optimization

Abstract

A Superstreet (also called Restricted Crossing U-turn Intersection or RCUT) is a relatively new research concept of non-traditional intersection with an alternative geometrical structure. This type of intersection became very popular due to reducing control delays and improving traffic safety because movements on the main road are carried out without opposing flows. In this paper, we treat the environmental impact of traffic on Superstreet and present the optimization of an ecology-based performance index. The mathematical formulation based on the Genetic Algorithms (GA) for controlling Superstreet in fixed-time control mode is developed. The number of stops on the Superstreet approach is weighted with a different penalty factor (K), which depends on traffic and geometric parameters for each movement. The geometry of a field-like Superstreet has been coded into the VISSIM simulation model and used as the approach tool for the evaluation of GA and Webster signal plans. Webster's method, for calculation of the standard signal plans at the signalized intersections, is a common method applied many times in practice. We suppose a set of traffic demand patterns and carry out all necessary calculations. The results show improvements in evaluation parameters and reveal the applicability of the proposed methodology.