Cooperative control for platoon generation of vehicle-to-vehicle networks: a hierarchical nonlinear MPC algorithm

This paper develops a hierarchical nonlinear model predictive control (MPC) algorithm for the cooperative control of vehicle-to-vehicle (V2V) networks. In the hierarchical algorithm, an estimator is proposed as a single layer, namely, the estimated layer for generating sub-targets with local physical constraints. Then, based on an augmented simplified vehicle model, a nonlinear MPC technique is newly presented in the local layer for planning the desired trajectory with its optimization solved by a continuation/generalized minimal residual (C/GMRES) method. Note that, by considering the error components between the planned trajectory and the measured positions, a controller with a two-loop structure is adopted to improve tracking performances. Therefore, the control objective is obtained, namely, the vehicle platoon in the desired path is generated and the collisions among vehicles are avoided. The simulation results with comparisons are finally given to illustrate the effectiveness of the algorithm.