Performance analysis of traffic control system based on stochastic timed Petri net models

We present a new modeling method to evaluate the performance indices of traffic control systems where each intersection has two or four phases. The method is based on the stochastic timed Petri net models of isolated intersections. It considers one direction of traffic along a street each time, and the interactions between traffic of different directions are partially approximated by statistical models. By so doing, it dramatically reduces the computing complexity that other Petri net-based methods suffer due to the consideration of detailed interactions among all directions of traffic. An example is presented to show the potential application of the new method