Distributed Stochastic Model Predictive Control for Heterogeneous Vehicle Platoons Subject to Modeling Uncertainties

This article proposes a distributed stochastic model predictive control (DSMPC) algorithm for vehicle platoons in which every vehicle is subject to modeling uncertainties. When the distributions of the uncertainties are available, DSMPC enables a less conservative treatment of the uncertainties than is commonly used for distributed robust MPC. The DSMPC problem is transformed into a deterministic one by constraint tightening. Terminal costs and constraints are carefully designed so that the recursive feasibility of the DSMPC problem is achieved. Moreover, a control update policy combined with the terminal costs design ensures the asymptotic average convergence of the vehicle states. Analysis results are presented and simulation outcomes are provided to verify the effectiveness of the proposed DSMPC method for platoon systems.