Centralized Distance-based MPC Strategy for Local Formation Tracking of a Multi-Robot Fleet

In this paper, a centralized control strategy is proposed to perform the formation tracking (FT) of a multi-robot fleet, without using absolute position information. The core of the approach lies in the predictive regulation of inter-agent distances using model predictive control (MPC) to improve robustness against deformations. The proposed strategy allows the formations to be efficiently maintained around the moving leader, even in GPS-denied conditions. To evaluate the advantages of using a distance-based formalism for local formation tracking, the proposed strategy is then compared with another approach inspired by the literature, in which robots maintain positions with respect to the fleet centroid. Simulation results show the efficiency of the proposed MPC framework in maintaining a local formation around the leader and highlight the benefits of using the distance-based formalism in constrained settings. To conclude the study, further discussions are made about the specification of each formalism.