Modelling and Analysis of Bottleneck Breakdown on Freeways with Multiple On-Ramps: a Copula Approach

Haizhong Wang,Jia Li,Yang Yu,Shangjia Dong

2014-01-01

Abstract:Freeway traffic flow breakdown is prevalent on the highway system for known and unknown reasons. One typical occurrence is the freeway downstream traffic breakdown. It is understood that the downstream bottleneck breakdown/failure is dependent on the upstream mainline flow and the flow oncoming from the on-ramps. This is a critically important research problem as the upstream demand uncertainty and flow pattern dependency impacts the throughputs at the downstream bottleneck. This research is largely motivated by the necessity to better understand the downstream bottleneck breakdown mechanism given multiple joint input flows through on-ramps and the freeway mainline flow. Therefore, the dependency structure between the different on-ramp flows or dependency between freeway mainline is essential to understand how the multiple upstream entrance-flows can jointly contribute to the downstream flow breakdown. The dependency among multiple on-ramps acts in a somewhat non-tractable but random fashion. In this paper, a copula-based bottleneck breakdown risk model is developed. The copula modelling framework is capable to capture the random flow dependency and the downstream bottleneck breakdown risk. For example, when the upstream flow is significant, downstream breakdown tends to form. One of the benefits of this modelling framework is to provide an alternative to calculate the downstream bottleneck breakdown probability on freeway mainline. A monte carlo-based bottleneck breakdown risk model is developed to evaluate the how demand uncertainty and dependency structure among multiple on-ramp flows impacts downstream bottleneck breakdown risk. The numerical results showed the freeway mainline traffic flow bottleneck breakdown risk can be estimated through Monte Carlo simulation.