Distributed Model Predictive Formation Control of Micro Multirotor UAVs With Virtual Tube

Abstract

A distributed approach based on virtual tube is proposed in this article to address the formation exploration of micro multirotor unmanned aerial vehicles (UAVs). A generic mathematical model for the virtual tube is established. Then, a path planning method for a formation cluster based on distributed model predictive control and the consensus theory is introduced. This method incorporates tube constraints by modeling the problem as a multiobjective optimization with linear constraints, then the considered problem is ultimately transformed into a quadratic programming problem that can be deployed on real UAV's onboard computers with limited computational resources. In addition, an assumed state constructor based on neighbor prediction information is proposed to enhance algorithm stability, thus eliminating the need for a priori knowledge of target points for all UAVs. Finally, the effectiveness and practicality of the algorithm are validated through simulations in three different tube environments and a real-world experiment involving six multirotor UAVs.