Sag curves, road segments where the gradient changes from downwards to upwards, generally reduce the roadway capacity and cause congestion. This results from a change in longitudinal driving behavior when entering a sag curve as drivers tend to reduce speeds or increase headways as vehicles reach the uphill section. In this research, a control strategy is investigated through manipulating the speed of connected vehicles (CVs) in the upstream of the sag curve to avoid the formation of bottlenecks caused by the change in driver behavior. Traffic flow along a sag curve is simulated using the intelligent driver model (IDM), a time-continuous car-following model. A feedback control algorithm is developed for adjusting the approach speeds of CVs so that the throughput of the sag curve is maximized. Depending on the traffic density at the sag curve, adjustments are made for the speeds of the CVs. A simulation-based optimization method using a meta-heuristic algorithm is employed to determine the critical control parameters. Various market penetration rates for CVs are also considered in the simulations. Even at relatively low market penetration rates (e.g., 5–10%), significant improvements in travel times and throughput are observed.
Nonlinear Analysis of a New Car-Following Model Based on Internet-Connected Vehicles
