The related congestion failure estimating methodology and model in transportation networks

Previous works about the probability-based transportation networks evaluation method mainly focus on the static reliability evaluation, they ascribe the stochastic of the travel time to the external long time factors (the traffic supply or the traffic demand). Under this situation, the link's travel time related relationship can be inferred, and it is efficacious for planners or engineers to make a decision for a long time. Even though some evaluation methodologies about transportation networks' real-time travel time reliability has been presented, these works assume that the link's travel time is independent. In this paper we relax this assumption. Using the Gauss copula theory, we present a new method to evaluate the transportation networks' real-time travel time reliability. The results show that it will overestimate the route or the networks' travel time reliability when not considering the links' travel time are related. Not only that, we deep the static reliability evaluation model to the dynamic, we also present the link and transportation network congestion failure evaluation model. Estimations from the model are compared to field-measured data. It shows that, under the error interval ±2 times, the link congestion failure model accuracy rate is above 90.3%, under the error interval ±0.05; the net congestion failure model accuracy rate is above 95%.