Centralized and distributed model predictive control for consensus of non-linear multi-agent systems with time-varying obstacle avoidance

In this paper centralized and distributed model predictive control methods are used for the consensus problem of nonlinear multi-agent systems in the presence of fixed and time-varying obstacles. The distributed method is implemented in two ways: serial and parallel. Contractive constraints are applied to establish stability, and the convergence of agents to a consensus point is guaranteed. Obstacles are considered as the constraints of the optimization control problem. Using the proposed methods, the need for designing terminal ingredients that are essential parts of designing the conventional MPC for establishing stability is eliminated. Terminal ingredients are usually hard to design and have been used for low-order systems but the proposed methods reduce the complexity of the design and can be applied for higher-order systems. To show the effectiveness of the proposed methods, the simulation is provided for the consensus of a group of wheeled mobile robots for both centralized and distributed methods and the results are reported.