The Penetration Effect of Connected Automated Vehicles in Urban Traffic: An Energy Impact Study (1803.05577v1)

Abstract: Earlier work has established a decentralized framework of optimally controlling connected and automated vehicles (CAVs) crossing an urban intersection without using explicit traffic signaling. The proposed solution is capable of minimizing energy consumption subject to a throughput maximization requirement. In this paper, we address the problem of optimally controlling CAVs under mixed traffic conditions where both CAVs and human-driven vehicles (non-CAVs) travel on the roads, so as to minimize energy consumption while guaranteeing safety constraints. The impact of CAVs on overall energy consumption is also investigated under different traffic scenarios. The benefit from CAV penetration (i.e., the fraction of CAVs relative to all vehicles) is validated through simulation in MATLAB and VISSIM. The results indicate that the energy efficiency improvement becomes more significant as the CAV penetration rate increases, while the significance diminishes as the traffic becomes heavier