Estimating traffic operations at multi-lane roundabouts: a case study

This paper addresses traffic modeling issues at urban multi-lane roundabouts where, despite circulating vehicles have priority, negotiation of the right-of-way can occur between antagonist traffic flows, as a result of minor drivers’ failing to obey the nominal operating rule (stop or yield control). Existing models for the estimation of operational performances have the shortcoming of not representing the interdependencies between entering and circulating vehicles at multi-lane roundabouts. An analytical capacity model derived from field observations was developed for this kind of intersections in a previous study. The complexity of the model lies in the difficulty of observing the behavioral parameters which are needed to implement the model. A procedure to get unknown behavioral parameters from traffic surveys is here proposed. This concerns saturation headways, often eluding direct observations due to rare occurrences of traffic conditions in which they can be observed. The unknown parameters were estimated through a regression model using on field data collected at a multi-lane roundabout. The presence of data correlation within a cluster of observations required the estimation of the regression parameters through a generalized estimating equation model. Results gave insight into the analysis of operations at multi-lane roundabouts, containing evidence to support assumptions made for the estimation of unobservable parameters.