Distributed Model Predictive Control for Cooperative and Flexible Vehicle Platooning

This paper investigates the cooperative and flexible vehicle platooning problem in automated highway systems. We formulate the platoon into a dynamically decoupled system and address the flexible platooning problem using distributed model predictive control (DMPC) techniques. A two-step non-iterative DMPC strategy is proposed that sequentially solves local components of a constrained optimal control problem over two batches of vehicle clusters based on intervehicle communication. By introducing the clustering procedure, the proposed DMPC scheme only requires compatibility constraints of common neighbors of two adjacent vehicle clusters. With the two-step DMPC scheme, stability of the overall closed-loop system is guaranteed by a series of linear matrix inequalities that can be efficiently solved in practical applications. Collision-free properties are addressed using coupled state constraints together with terminal sets during a cooperation procedure. Simulations of flexible platooning are conducted for both joining and leaving events to show the effectiveness of the two-step DMPC scheme.