Information-constrained model predictive control with application to vehicle platooning

Information constraints, induced by a delayed communication between individual decision makers (DMs) pose a significant challenge for the optimal control of physically interconnected systems, as for example vehicle platoons. In this work, we address the problem of distributed state/input power-constrained optimal control of a vehicle platoon under the assumption that neighboring vehicles communicate to each other with one step delay. In order to account for sudden changes in the environment such as different speed limits, or change of the desired relative distances between vehicles, a model-predictive control (MPC) approach is adopted. Despite the information constraints and state/input constraints we prove the optimality of linear control policy. To this end, we provide an optimal structure of control law that is imposed into the MPC optimization problem, to account for information constraints induced by one-step communication delay between neighboring vehicles. The efficacy of the approach is illustrated in simulation.