Model-based analysis of traffic flow in signalized intersections

Traffic systems are man-made systems in which the human component as a complex factor plays a significant part. The aim of the research was to offer a method which is applicable for setting the real traffic situations at a signalized intersection against the simulated ones, using a traffic model with similar statistical features. The simplest technical apparatus has been used to monitor the real traffic situations. Only one video camera is needed for data collection and it is located in the observable area. Hence time-dependent parameters have been collected which are available for the analysis of the traffic situations. In the case of both traffic models the analysis is based on a behavior model of inhomogeneous driver-populations and an estimation method. The behavior model of inhomogeneous driver-populations handles the time-dependent parameters of driver behavior as the most important predictors, describes the movement and the behavior of the inhomogeneous driver-populations approaching a signalized intersection, passing the amber signal light. The estimation method developed using the model parameters opens the door to estimating the composition of drivers and measuring the capacity of an intersection on a more sensitive scale.