Data-driven optimal cooperative adaptive cruise control of heterogeneous vehicle platoons with unknown dynamics

This paper considers the cooperative adaptive cruise control (CACC) problem of heterogeneous vehicle platoons and proposes a data-driven optimal CACC approach for the heterogeneous platoon with unknown dynamics. To cope with the unknown dynamics of the vehicle CACC platoon system, the adaptive dynamic programming is used to design an online iteration policy for optimal CACC of the platoon. Using the predecessor-following topology, the CACC controllers are computed by employing the desired spacing errors, relative velocities, and accelerations of the vehicles. The stability of the closed-loop CACC system and the iteration algorithm are presented. Furthermore, the string stability of the platoon with the CACC system is established in terms of the acceleration transfer function between adjacent vehicles in frequent domain. Finally, the effectiveness of the proposed method is verified in two complex scenarios of varying speed cruise.