Non-Iterative Distributed Model Predictive Control For Flexible Vehicle Platooning Of Connected Vehicles

This work proposes a non-iterative distributed model predictive control (DMPC) scheme for flexible vehicle platooning with state and input constraints. The proposed DMPC scheme partitions the multi-vehicle system into alternate clusters belonging to two categories, and distributively solves constrained optimal control problems (COCPs) in a two-step manner at each sampling instant. At the first step, the scheme solves COCPs over one category of vehicle clusters that are disjoint. Then, at the second step, COCPs are solved over the other category with predicted states of common neighbors transmitted from the first step. Instead of applying compatibility constraints on all vehicles, only compatibility constraints of common neighbors of two neighboring clusters are required in the proposed scheme to ensure feasibility and closed-loop stability. The proposed DMPC law can be obtained by solving a semidefinite program (SDP). A flexible vehicle platooning example is simulated to illustrate the effectiveness of the proposed DMPC scheme.