A data-driven model predictive control framework for robust cooperative adaptive cruise control using mobile sensing data from connected vehicles
Department of Civil, Environmental, and Geospatial Engineering; Center for Cyber-Physical Systems
This paper proposes a data-driven Model Predictive Control (MPC) framework to improve a robust Cooperative Adaptive Cruise Control (CACC) system by optimally controlling a Connected and Automated Vehicle (CAV) platoon under uncertainty. Speed variation, spacing variability, and driving comfort are considered in the objective function of the proposed MPC model. Although vehicle connectivity technologies are available in traffic networks, a CACC system should be represented as a stochastic system due to uncertainties in traffic flow dynamics. These uncertainties will affect vehicle automation of the CACC platoon. To deal with these uncertainties, the authors formulate a data-driven MPC model by solving a likelihood robust optimization (LRO) problem in a rolling horizon framework to consider the worst-case scenario for robust connected and automated driving using mobile sensing observations of the leading CAV in the CACC platoon. The authors also explicitly address the constrained driving situation (where a preceding vehicle is present in front of the CACC platoon) in the proposed MPC model formulation through predicting the driving speed and trajectory of the preceding vehicle using real-time mobile sensing data from other connected vehicles in front of the CACC platoon by forecasting the forecasts of others. This data-driven MPC model can be applied to both constrained and unconstrained driving conditions for robust connected and automated driving. The authors test the data-driven MPC model on a congested freeway segment on I-94 in Ann Arbor, Michigan using the microscopic simulator VISSIM.
Transportation Research Board 97th Annual Meeting
A data-driven model predictive control framework for robust cooperative adaptive cruise control using mobile sensing data from connected vehicles.
Transportation Research Board 97th Annual Meeting.
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/702
This paper was sponsored by TRB committee AHB45 Standing Committee on Traffic Flow Theory and Characteristics.