Predictive Cruise Control with Private Vehicle-to-Vehicle Communication for Improving Fuel Consumption and Emissions

摘要

Future traffic information through vehicular communication allows connected and automated vehicles to optimize their speed trajectories and drive more safely and efficiently through predictive controllers. Sharing accurate information about the vehicle allows such controllers to perform best, but may raise privacy concerns. To improve privacy guarantee over the shared information while preserving its utility for predictive controllers, this article proposes a novel information perturbation mechanism, as opposed to the baseline of independently perturbing the data in each broadcast. Specifically, the mechanism is applied to the transmitted vehicle speed, and this perturbed data is used in an optimal speed planner to design a fuel and emissions efficient speed trajectory. Results show a deterioration of the controller performance when privacy is taken into consideration under the baseline method. With the proposed method, the controller performance is improved while providing the same privacy guarantee. It is shown that controller design is also affected by the choice of perturbation mechanism.