String Stable Platooning Control based on Adaptive Dynamic Programming

String stability can be employed to evaluate the formation stability of vehicular platoons by checking if the velocity in the upstream direction of the vehicle platoon decreases, which is closely related to the safety of platooning systems in practice. Traditional string stable controllers are usually based on the accurate knowledge of vehicle model, which is hard to implement in practice. In this paper, we have proposed a novel data-driven control approach based on adaptive dynamic programming, $\mathcal{H}_{\infty}$ control, and zero-sum differential games, which does not rely on the model knowledge. Moreover, the platoon in closed-loop with the proposed control policy can ensure the strictly string stability. Traffic simulation results demonstrate the effectiveness of the proposed approach.