Optimal Control of Connected Autonomous Vehicles in a Mixed Traffic Corridor

This paper investigates the potential of improving the overall traffic and energy efficiency by properly controlling a proportion of controllable connected and autonomous vehicles (CAVs) in a mixed traffic corridor. Specifically, we develop a control framework that optimizes controllable CAV trajectories taking into account other vehicles for simultaneously improving traffic throughput and reducing the total energy consumption of all vehicles. The property of the control framework is firstly analytically examined in a simplified and tractable scenario where a human-driven vehicle (HV) follows a CAV. We found that the optimal acceleration is larger if one emphasizes more on improving travel distance within the optimization horizon, or smaller when one emphasizes more on saving energy. The continuous-time optimization model formulation is then discretized, which is solved for real-time application in a model predictive control (MPC) fashion. In numerical studies, the proposed method is tested in various scenarios, e.g., with/without an intersection, under different proportions of controllable CAVs, possible vehicle permutations, and varying overall traffic intensities. Numerical results show that the normalized energy consumption can be reduced by up to 45% and the average travel time reduced by 65%, showing a significant improvement in the road throughput. Notably, even with a limited number of controllable CAVs, the proposed method can achieve a promising performance, e.g., about 20% controllable CAVs can achieve half the benefits of a fully controllable CAV environment.