A Markov chain model for traffic queue evolution at an intersection

The model describes the probabilistic behavior of traffic queues in the vicinity of an intersection, taking into account their spatial extension. Vehicle arrivals and departures are recorded by vehicle detecting devices. The posterior distribution of the vehicle concentration at road segments is shown to admit a recursive formulation, which makes the actual implementation of queue filters and predictors feasible. Examples are given which illustrate the model behavior concerning the queue build-up and discharge mechanism.