Webinar Recoring: Control of Connected Autonomous Vehicles in Mixed Traffic: Modeling and Field Experiments

Advanced connected and automated vehicle (CAV) technologies can be utilized to achieve precise vehicle trajectory control and render unprecedented opportunities to improve transportation system performance in safety, mobility and sustainability. However, it is quite challenging to fully realize these advantages from CAV in mixed traffic environment due to stochastic and uncertain human-driven vehicle (HV) behavior. This study presents several key models for CAV control in both longitudinal and latitudinal directions considering coordination with infrastructure units (e.g., traffic signals) in mixed traffic environment. The fundamental idea of these models is to use learning based methods to estimate and predict HV behavior and use fast heuristic to plan and control the AV trajectory in a near-optimal manner. These models are validated with field experiments on a large-scale testbed in mixed traffic settings. These field studies are the first of its kind involving both longitudinal and latitudinal controls in mixed traffic. The results show that the proposed control models can safely and efficiently implement key AV control functions in mixed traffic. The webinar and handouts are available at https://www.cutr.usf.edu/2019/04/cutr-webcast-control-of-connected-autonomous-vehicles/