A Simulation Study for the Static Early Merge and Late Merge Controls at Freeway Work Zones

Abstract

The merge control has been proposed as a way to maximize the utilization of road capacity at the work zone area. Two of the most widely used merge control schemes are the static Early Merge (EM) control and the static Late Merge (LM) control. Although many papers adoptedfield measurements and/or microscopic traffic simulations to explore the performance of these two control schemes, they seldom cross compared them under the same work zone settings, and some of them even derived conflicting results. To extensively investigate and cross compare the traffic performance of the EM and LM controls, we conducted a simulation study using the microscopic traffic simulator VISSIM for a 2-to-1 lane closure work zone. In addition, the sensitivity analysis was adopted to determine the influential VISSIM parameters with respect to the work zone throughput.It is found that the parameters CC1 (i.e., headway time) and CC2 (i.e., longitudinal following threshold in the car-following process) have significant impacts on the simulated work zone throughput. In addition, adjusting CC1can make the throughput drop faster in the EM scenarios than in the LM scenarios. The simulations also show that when CC1 is low, the work zone capacity under the EM control is higher and the queue is shorter. On the contrary, when CC1 is high, the work zone with LM control has a higher capacity and shorter queue. As the parameter CC1 is critical to the simulationresults, it should be carefully calibrated and validated in the VISSIM-based merge control simulations. Based on the simulation results, we recommended that it is more appropriate to implement the EM control when the drivers are aggressive and the safety distance is relatively short (i.e., CC1is low); when the drivers are cautious and the safety distance is long (i.e., CC1 is high), it is better to adopt the LM control in the work zone area.