Traffic Impacts of Optimal Eco-Driving in Realistic Traffic Microsimulations

Tyler Ard, Clemson University

Abstract This talk covers the fundamental energy and traffic impact of a proposed Connected and Anticipative Cruise Controller in a PTV VISSIM microsimulation environment. We dissect our controller into two parts: 1. the anticipative mode, more immediately beneficial when automated vehicle fleet penetration is low, and 2. the connected mode, beneficial in coordinated platooning scenarios and high automated vehicle penetrations appropriate for autonomous vehicle specific applications. In-horizon and terminal constraints handle safety considerations, and vehicle constraints for acceleration capabilities are implicitly understood through the use of powertrain maps. Real traffic scenarios are then modeled using time headway distributions from traffic data. To study impact over a range of demands, we vary input vehicle volume from low to high and then vary CAV penetration from low to high. We find that our automated vehicles maintain traffic flow as compared to the all-human driving cases, where a decrease in network traffic flow was observed. Furthermore, at the highest input volume examined when shockwaves occurred and reduced vehicle throughput in the simulation, the introduction of CAVs improved traffic flow and removed shockwave effects. Finally, we find that our vehicles perform at a 10% - 20% higher fuel efficiency over human drivers. Due to secondary effects of smoothing traffic flow, fuel benefits are even greater when considering a fleet of human and automated vehicles, where humans were found to drive with up to 10% more fuel efficiency under high-volume scenarios.